ChessTempo User Guide

The Chesstempo chess board is central to many features. Below is a summary of features common to all boards used on the site.

Play Position vs Computer The Play Position vs Computer Button launches the play versus computer feature with the current board position set as the starting position.

Books can be purchased at the buy chess books page. Books can be purchased by both free and paying users. Some books include training features that would normally require premium membership such as spaced repetition. While training with a purchased book, such features are provided free of charge, and do not require an active membership to make use of. However premium membership does give access to a wider range of features while using a purchased book, such as cloud engine analysis. Creating custom sets beyond the default training sets provided with the book will also require a premium membership, and some advanced stats output may also require premium membership. However, the books are designed to be very usable without premium membership, and premium membership should be seen as a useful, but strictly optional supplement to the book.

Each book purchase comes with a browseable version of the book that can be read in a similar manner to a physical paper version. The book reader includes a convenient analysis board that can be used to view the position at any point in a variation, and allow you to easily browse through the variations without having to use an external board to keep track of the position. Arbitrary moves can be made on the viewing board if side lines not included in the book material need to be investigated. Premium members also have access to cloud engines to request analysis in any position.

See the your books page to see a list of your currently owned books (including those you have created yourself).

The left edge of the reading view shows a selection of icons and buttons that provide actions useful while reading. Some of these actions are specific to the type of content available in each book. Available actions include:

The chess content available in the book may also have associated action buttons:

There are several ways a book can be used for training:

Study mode allows you to step through the content of the book concentrating on one piece of content at a time. The book is broken up into study modules, and one module is shown at a time. Modules may be pure text from a chapter introduction, or game material such as a full game or chess problem. The top right of the study view shows a study browser button on the left which lets you see a list of the available modules (chapters, sections, games, problems etc), and navigation buttons to move forwards and backwards within the study modules.

Study mode can be entered by either clicking the 'Study' button in the left column in the reading view, or by clicking on the 'Study' button attached to games or problems which will jump into study mode with that item selected as the current learning module. If you use the left column Study button , study mode will be started with the last module you visited loaded (or the start of the book if this is the first time using study mode on this book).

Chess games can be played through within the study view, or if the game is available as a Guess the Move game, you can click the 'play' button to launch the game in GTM mode. After playing through the game, you’ll be shown your moves and move assessment scores alongside the annotations of the game from the book, allowing you to see author commentary for many of the positions from the game.

Chess problem modules show the problem in interactive solving mode. In interactive mode comments from the author may be used as prompts during solving, and you are given a 'solution' button to show the solution move for the current position via board arrows. You can move the pieces to make your move, and if there was a comment to display connected with that position, you’ll have to click the next move button to play the next opponent move. No rating point adjustments are made for solving done in study interactive mode.

Chesstempo provides a book creation and editing interface. There are three main uses for the book editor interface:

To create a book, use the Chess Books → Create book menu action. Choose a title , a summary, description etc, and any appropriate advanced options. You can start a book from scratch, or import material from PGN. See the book editing section of the User Guide for details on how to edit a book after creation. For details on how PGN is treated, and methods of creating book specific PGN games for import, see the creating books from PGN section for further details.

The book editor allows you to create chapters, headings, move list variation and diagrams, basically all the required contents of a chess book.

The insert menu at the top of the book editor view allows you to insert several types of chess and non-chess content:

The following section describes the PGN tags and structures that Chesstempo supports when converting an imported PGN file to a book structure. The importer will do its best to deal with any PGN in a sensible manner, but structuring your PGN to match some of the explicitly supported Chesstempo book tags will usually provide a better initial import, and reduce the amount of subsequent editing required.

Chesstempo tactics training is aimed at improving two important aspects of tactical ability, pattern recognition, and calculation skills. Blitz and Standard tactic solving modes target both these areas. Mixed mode allows you to merge both these skills, with a solve time rating adjusment more closely reflecting time available in longer over the board games, and includes positions where the best move may leave the position even, but is the only move which avoids entering a lost or losing position.

Some tactical positions have moves, which are good, but are not deemed to be the 'best'. When this occurs, you will receive a 'try again' message, and you’ll be given another chance to find the 'best' move.

The reason the problem set includes positions with these alternative winning moves is because it is important to provide training in as many real game situations as possible. Real games often have positions where you need to choose between multiple good moves, and so it is important that training includes such positions. Positions where two moves are given the same computer evaluation will not be used, so one should always be better than all others (at least from a computer evaluation perspective).

The Chesstempo rating system for tactics and endgames is based on the Glicko Rating System. Problems and Users are both given ratings and the user and problem rating are updated in a manner similar to the updates made after two players have played a game against each other. Generally, if you get a problem correct, the problem rating goes down , and your own rating goes up, and the opposite will occur if you get a problem wrong.

The tactic session panels shows stats for the current solving session. The session starts when the page loads or after the End session button is used. The current session time, a graph of the last 20 problem attempts (green dots = correct, red = incorrect), and some summary stats for the current session are shown.

There are three groups of problem types, rated, unrated , personal and from my games. Rated includes the officially rated sets such as Standard and Blitz sets. Unrated includes any of the default custom sets such as Mate in 1, Fork etc. Personal is where any custom sets you create are placed. From my games sets are sets targeting tactics extracted from your games.

Benchmark mode allows you to play any move that keeps the win, even if it would take longer to mate. You play moves until the position is deemed completed (usually by either mating or typically for problems with more pieces on the board, after taking a piece or promoting, converting to a more easily won endgame). Playing a move which loses or draws will fail the problem.

In Benchmark mode , your rating adjustment for successful solutions will be based on how many moves you take compared to the average number of moves spent by all previous solvers who have solved the problem correctly.

Theory mode forces you to play the fastest line to mate, if you play a move that is not one of the moves that leads most quickly to mate (and there may be more than one), then you will be asked to try again. This will continue until you find one of the fastest mating moves, or you play a move which loses or draws, at which point the problem will end and you will be marked incorrect. This process will continue until you have either failed, or reached the end of the problem. The main purpose of theory mode is to be shown the optimal mating line, and as such is useful for endgames where you may not know how to mate at all.

Endgame practice mode is similar to Benchmark mode , in that it does not force you to follow the optimal line. You can play any move you like, including moves which dramatically increase the number of moves required to mate. The difference between practice mode and benchmark mode is that like Theory mode, Practice mode compares the number of moves you took to the optimal number of moves to mate to calculate your rating adjustment.

The rating system for endgames is similar to that used in the tactic problems, the Glicko rating system is used to rate both endgame problems and players. For failed attempts rating adjustments are done based on a straightforward 0-1 result. Your rating adjustment for successful rating attempts depends on how many moves you played before completing the problem.

For benchmark mode if you take the same number of moves or less than the average number of moves on correct attempts for all users, you will receive full rating points. As the number of moves increases beyond the average, you’ll start to get reduced rating points. If you end up taking exactly twice the number of moves compared to average, the result will be treated like a draw, so against a problem rated at your current rating level, you will receive a zero points rating adjustment. If you use more than twice as many moves compared to average, you will start to lose rating points against a problem rated at your own rating level.

Note that in benchmark mode, some problems may have common paths that prune quickly when converting to another endgame such as taking a pawn instead of opting to mate in the current position. In this situation, the average move length played by all users may be shorter than than the fastest mate in the current position. This could lead to unfair situations where a mate in 6 is played with no extra moves, but marked down due to being considerably longer than the more common move sequence that leads to a quick prune (but a longer mate). To avoid this unfairness, Benchmark scoring on correct solution will also check depth to mate where available as a fall back score adjustment method. This means if you would score highly based on how quickly you mated, but the average move length would punish you, your score adjustment will be the highest of the depth to mate vs the user average move length scoring methods.

Benchmark’s rating method allows you to play lines that may not be the fastest method to mate , but still exhibit good endgame technique without worrying about being punished too much, due to using other solvers move lengths as the benchmark for scoring rather than number of moves to mate.

For practice mode, successful rating adjustments are based on how much longer your number of moves was compared to the fastest number of moves. If you play the optimal move at each point then you will receive a full 1-0 result. However

Theory mode rating adjustments for successful attempts are similar to practice mode, except that instead of total moves played impacting the outcome, it is the extra moves that would have been used if the longer moves had been allowed. For positions where you attempt to play more than one sub-optimal move, the move with the longest length to mate is used for rating purposes.

The list of played moves in endgame problems serves several purposes. It allows you to see the path you have followed while the problem is active, but it also provides extra details on the moves you played after the problem is finished. You can also play through different lines after the problem is over, and see the variations added to the move list.

The annotations on the move list after the problem is over show moves that were not the best move in the position. For example M10(M8) shows you played a move that was mate in 10 where the best move was mate in 8. L10(M2) shows that you played a move that leads you to lose (i.e. be mated) in 10 moves, where the best move would have had you mate your opponent in 2 moves. D(M3) indicates a move that draws when you could have mated in 3.

The legal move list appears after you have finished solving an endgame (or when viewing the problem in problem viewing mode). For each position it shows a list of all the legal moves and the number of moves to mate (or to be mated) when a win or loss is the optimal outcome, or 'draw' when best play would lead to a draw.

We also provide a distance to zero (DTZ) number for each move, which is the number of moves (in ply) before the 50 move counter would reset (so before a pawn move or capture occurs).

Note that the scoring is 50 move rule aware, so there are three types of draws, the normal draw, a cursed win draw (i.e. a position that allows mate, but not before the 50 move rule), and a blessed loss (a position that would lead to the player getting mated, but not before the 50 move rule would finish the game as a draw).

You can click on any of the moves in the legal move list to play that move on the board.

Not that clicking on a move in the endgame legal move list or moving forward in the played move list will play only one side’s move. However if you play the move by interacting with the pieces, the opponent move will be played in reply, allow you to play out lines against the position.

Some endgames can be quite easy to play to mate, but very difficult to play the fastest mating sequence, these problems can be particularly annoying in modes that score you based on the optimal mate length as it can be easy to lose rating points on these type of problem. The "Blitz" button is designed to avoid some of this frustration.

The "Blitz" button allows you to take the option of solving endgame problems under the clock. The "Blitz" button is intended for use on endgames which you feel may be easy to mate , but not easy to play perfectly. After clicking the "Blitz" button, the clock changes to a count down timer and you have until the clock reaches zero to solve the problem.

Similar to Benchmark and Practice modes (irrespective of whether you were in Theory or Practice or Benchmark mode to start with), you are able to make any move you like (as long as it is still winning). You will not be punished for moves that take longer than the shortest path to mate, and in benchmark mode you will not be compared to the average number of moves to solve.

If you solve the problem within the time limit, you will get full points for getting the problem correct. There are no rewards for fast times, you will be marked the same if you have 1 second left on the clock or 50 seconds. If you run out of time, play a losing or drawing move then you will marked incorrect. You can only choose to blitz a problem before your first move. The number of seconds you have to solve the problem is defined by the function MAXIMUM(60,DTM*3.0+10) where DTM = Depth to mate. The time used before you press the blitz button is also included in the time allowed, so the longer you take to press the blitz button, the less time you will have to solve the problem once the clock starts counting down.

When using the blitz button you are required to mate the opponent, the usual endgame pruning is disabled. However, to avoid frustration from having to play out endgames against a lone king when blitzing after having already converted from a harder endgame, there are specific conditions under which a blitzed endgame will prune when you reach a position where your opponent has a lone king. If you reach such a position that would normally be pruned in a non-blitzed endgame, and have not used so much time already that you would likely not have time to mate, and have never reached within 2 moves of checkmating up to this point, then the problem will finish without you having to mate the opponent king. Sufficient time to mate is defined by the formula MAXIMUM(10 seconds,DTM*1 seconds). Or in other words if you have at least 10 seconds or 1 second per move to mate left when you reach a lone king position for your opponent, then you will be deemed to have enough time left to mate, and the early pruning check will be made.

The "Play Best" button plays the best move for you in the current position. However there will be a large penalty for using the button, so be careful how often you use it. Each time you use "Play Best" there will be a certain number of moves added to your move count (used in calculating your rating if you succeed). The penalty for using "Play Best" will be a minimum of 10 moves. In Practice and Theory mode, if you have made any previous attempts then the penalty will be the longest attempt you had made already for the current position plus an extra 5 move penalty (or 10 moves if this number is still less than 10). Benchmark mode will add a flat 10 moves on to the number of moves taken each time the play best button is used. The purpose of "Play Best" is to provide an alternative to completely giving up, and thus to allow you a chance to learn more about a position which you may otherwise feel completely lost in.

The Opening Explorer provides a way of studying the popularity and performance of different opening moves. The opening explorer allows you to play through opening lines, with statistics provided on all the moves, indicating how they have performed when played by others.

The figure below shows an example of the opening explorer in action:

You can interact with the opening explorer by making moves on the board, or by clicking one of the candidate moves. If you are playing through the moves in an actual game, the explorer will also be active, and will show the stats for the current position in the game.

Game search serves two purposes, it allows you to find games meeting a particular criteria, but it also allows you to control which games are processed when displaying the opening explorer stats.

The Games for Position list shows all the games that contain the current position on the board. If any search criteria was entered, then the games for the position will also be filtered by that criteria (excepting material search criteria which is not included in the games for position filter).

The Games for Position list, is particularly useful in conjunction with the opening explorer, and can be used to study example master games that followed the current opening line.

Like the game search results table, most columns can be sorted by clicking on the relevant column header. See the Results List section of the user guide for more information on the columns returned in the Games for Position results.

The Player List page allows you to search for a particular player. It can also be used to identify the top players, as well as the top players by country, gender, rating range, or a particular period in history.

The Openings List allows you to browse the openings stored in the database, searching and sorting the openings by a number of different criteria.

The player details page shows information relevant to a particular player. Each player who has played a game stored in the database has a player details page. There are several ways of reaching a player page, by clicking a player link on the player page, by clicking on a player link in a game search results or games for position list, or by going directly to the web address which links to the player page via a bookmark, or link posted on another web page.

The opening details page shows information relevant to a particular opening. When the opening details page loads, the move list is initialized with the moves in the main line of the opening. The board is shown in the state that exists after the last move in the opening is played, and a game search is automatically performed to show all the games that reached that position.

The game details shown next to each board on the game database page shows details relevant to either a particular game, or a sequence of moves entered by the user.

The Database selector allows you to conveniently select a subset of games to be used. Database subsets are based on the minimum rating of both players involved in each game. The default subset is 2200+, and subsets up to SuperGM+ (2700+) can be selected. To see the full database select the "All Games" option.

The database subset will effect the game search, the games for position list and the results shown in the opening explorer. The database subset will be ignored if you enter a rating based search in the advanced or quick search.

The filter impact selector allows you to choose which elements of the user interface will be impacted by the game search filter. By default the filter impacts both the games for position list, and the opening explorer stats. You can choose to change this to impact only the position list, the opening explorer stats, or neither of the two elements.

The board shown on the database page has the following extra feature:

To create a new comment , click on the 'Add' comment button. Please read all existing comments before adding a new one, as someone may have already made the same point, or may have answered a question you plan to ask.

Not all comments made are useful or informative. If you think a comment is low quality, then you can use the down vote button ( ) to vote a comment down. Similarly high quality comments can be voted up by using the vote up button ( ).

Comments can include mini-boards which either show a static board position, or a board with a list of moves and variations than can be played through. By default, boards need to be shown by clicking on the 'Show' button, this is to prevent slowing down older browsers on pages with a large number of boards. If you are using a new browser on a fast machine, you may wish to set the 'Automatically show comment boards' option on the 'Other Settings' preferences tab.

Boards can be inserted into a comment using the analysis board or 'Copy Fen' button which has a 'Copy to Comment' option. The 'Copy Fen' option creates a static board position represented by the current position on the board.

Boards can also be inserted manually using the following syntax:

If your comment is the most recent comment made, then you are allowed to either delete or edit the comment using the Edit or Delete buttons. The Edit button will place your existing comment into the add comment box, and once you have finished editing, the Submit button saves the edited comment.

The 'My Comments' tab on the Problems Page provides a convenient way of viewing replies on problems you have commented on in the past. The 'My Comments' tab shows a list of problems you have commented on, and all the comments on those problems. The problems are ordered such that the problems that have been most recently commented on appear first. This means you can regularly read the list to view new comments on problems you have previously commented on. Your own comments are highlighted, and a colour bar separates the comments of one problem from another.

Basic search allows you to choose a rating range and rating type (Blitz/Standard/Mixed for tactics and Theory/Practice for endgames). After clicking the 'Search' button, problems matching your chosen rating range in the chosen rating type will be returned in the problem list. By default the problem list returns all tactics problems, and any sorting is based on the problem number. You can sort by other columns by clicking on the column header.

The problem search results includes the following columns:

Advanced search allows premium users to have greater control over the problem search, it includes criteria for problem specific features, including criteria related to previous attempts on the problems. These criteria can all be saved as custom problem sets for solving against the matching problems on the tactics or endgame training pages.

Sorted/looping custom sets go through the problems that match the custom set criteria in a particular order, and provide the ability to choose the sorting criteria and the direction of the sort. When you reach the end of a sorted set, the set is reset to start from the beginning, and you can continue to solve any problems left in the set that still match your custom set criteria.

Custom sets of this kind can be used to implement some popular tactics training methods such as the seven circles method described by Michael de la Maza, or the woodpecker method introduced by Axel Smith and Hans Tikkanen. Under these methods, a set of problems is gone through multiple times, with the aim of achieving faster solving rates and increased accuracy during each loop. The final goal being to improve pattern recognition by repeated exposure to a set of tactical patterns.

To help track your progress as you pass through each loop of your sorted set, some stats are available for accuracy, performance rating, time taken etc for each loop. Click on the icon next to the loop counter while solving to see these details.

If you’d like to have the contents of the first loop through a sorted set decide problems considered for subsequent loops, then enable the "First loop through set decides problems considered for subsequent loops" option when creating your sorted set (like the sort direction and sort criteria option, this will be shown after selecting "Sorted/Looping" as your problem selection method). To make use of this option, create a custom set criteria that includes a number of problems which may be greater than the total number you plan to train on each loop. Select the "first loop through" option before creating the set, then start solving using your new custom set. When you get to the point where you have done the amount of problems you want to train in each subsequent loop, then click on the 'looping set details icon' to the right of the loop counter in the solving interface, and click the 'Finish Loop' button. This will start "loop 2" from the beginning of the sort order, and only consider problems seen in loop 1. When you reach the end of those problems, the set will roll over into loop 3 , and continue on again with the problems you saw in loop 1.

Keep in mind that if you have dynamic criteria in your set such as "problems I always got wrong" and you get them right in loop 1, they will still drop out for loop 2. So in this example, only the problems you got wrong during the first loop will be considered in subsequent loops when using this method, rather than the entire set of problems you’ve previously got wrong, which would be the behaviour without the 'first loop through' option being used.

You can use the 'finish loop' button after the first loop, but it will simply increment the loop counter and start the sort order from the beginning for the next loop. The number of problems seen in the first loop will still remain the relevant problems for remaining loops. The 'finish loop' button can also be used in non-'first loop through' sorted sets, and in that situation, it will again simply increment the loop counter and reset to the start position for the next loop, and the entire set of relevant problems matching the custom set criteria will be relevant for all loops, irrespective of the use of the 'finish loop' button during the first loop.

Note, that if you are using a "first loop through" set and you simply reach the end of the available problems in loop 1, there is no need to use the 'finish loop' button. The subsequent loops in the set will still only include items done in the first loop, so if you have dynamic criteria such as "problems I got wrong", problems you get wrong in other sets after loop 1 was completed will not become available in subsequent loops (which is what would happen in a non-"first loop through" set).

If you’d like to make sure that dynamic criteria you have set, such as rating ranges or quality filters don’t lead items to drop out of your set as they fall out of the matching range, then you could edit your custom set criteria of a "first loop through" set after you have finished the first loop. For example if you set the rating range to say 1000-1100, and end up doing 500 problems during your first loop, and want to make sure they stay there for subsequent loops, irrespective of whether their problem ratings move outside of the 1000-1100 range, then edit the rating range of the set. To do this, use the manage sets button, select the set, and change the rating range criteria to cover the full rating range. This will not include any new problems that were not already seen during loop 1, but will make sure problems in loop 1 will not drop out if their range shifts over time to be outside the range. This is useful if you want to ensure you do only a fixed number of problems each cycle, but still want to use search criteria that might change over time. Tag votes and quality votes can also change over time, and you can use the same trick with those. For example, if you create a set that looks for only pins, remove the pin criteria from the set after loop 1 is completed to have a fixed set of 'pins' that doesn’t change in the face of tagging votes from other users.

The looping set details panel also has a 'reset' button, which resets the entire set back to the default situation. This means the loop counter goes back to 1, the sort order resets to the start, and the looping stats for this set will be cleared. For sets using the "First loop through" option, this will also reset the 'first loop' problem selection, and you can choose the size of the subsequent loops again by using the 'finish loop' when you reach the appropriate position in your first loop of the now reset custom set.

When creating training sets where you loop through a fixed set of problems multiple times in order to improve pattern recognition, we recommend choosing a reasonably large set size, and problems that are not too calculation heavy for your current level of skill. The reason for this suggestion is that if the set is too small, you will likely find yourself rote learning problems/solution pairings rather than acquiring the underlying patterns involved. While both outcomes can occur in parallel, you probably want to focus primarily on a more generalised internalisation of the patterns, rather than a literal memory of particular problem solutions. Ideally, you’d like the literal memory of a position to have faded by the time you see it again, and for recognition of the general pattern, rather than the literal position/solution pair to be the overriding method of solving each time. The larger the set size, the more the focus will be on generalising patterns rather than robotically repeating rote memorised problem/solution pairs.

Similarly, for positions that are complex enough to be largely exercises in calculation, the value of repeated learning of these structures is probably less valuable than trying to focus on simpler patterns. While it is possible to rote learn the solutions to complicated calculation problems - even for problems with difficulty well outside your range if you spend enough time - this time may be better spent on acquiring simpler patterns. This isn’t to say calculation practice is not useful, but repeating the same calculation exercise over and over again may not be the best way to train this aspect of the game.

There are also some potential optimisations you may want to consider for these types of looping sets. One downside of using this type of set is that you can waste time continually repeating a pattern you may already know very well. This time could be better spent on problems holding patterns you’ve not yet acquired. To help counter this, you could filter out problems you’ve got correct more than N times using the advanced history filter option. Alternatively, you could drop problems from the set by setting a minimal solve time under which items are not included, so that as soon as you can see a problem solution in under 5 seconds (for example), that problem would be automatically removed from the set.

If you want more manual control over what is included or excluded, you could create a new personal tag, such as "Patterns I know", and then use the custom set matcher to create a "not" match on the "patterns I know" tag for your custom set. That way, any time you vote up the "Patterns I know" tag on a problem in your set, it will be permanently dropped from all future loops through the set. You could extend this idea further by creating a review set that only includes the problems you’ve tagged with your "Patterns I know" tag, and use that for later review, voting down your tag on any item you got wrong, or felt you didn’t know well enough during the review, thus including it back into future loops in your looping set.

Spaced repetition is a memory learning method that allows you to maximise recall, while minimising learning time. Essentially it tries to present you with problems you are having trouble with more often that those that you have already shown you can do well. You can read more about spaced repetition at the Wikipedia Spaced Repetition Article.

You can create a spaced repetition set by enabling the "Spaced repetition" option for the problem selection type (instead of the default "Random Selection"). The selector can be found just above the problem set ordering options.

For spaced repetition sets, you get two new numbers next to the problem set count at the bottom of the problem board, for example: Problem Set: TestSR ( 91 in set) ( Learning: 8, Scheduled: 4 )

"Learning" is the number of problems in the set that we are currently actively learning.

"Scheduled" is the number of problems that are currently due for review.

When "Scheduled" = 0, each time you do a problem, a new problem will be added to the set for learning. When "Learning" is the same as the number of problems in the set , trying to do problems when Scheduled = 0 will lead to problems being done in order of the next one that would be scheduled (i.e. you’ll be solving problems "early", before they are due for review). If you can’t be given problems early without repeating the last scheduled problem, then you’ll be told you need to wait for the next scheduled problem to be due.

If you get a problem wrong, it will be scheduled for serving again in 5 minutes. The first time you get a problem correct, it is scheduled for review in 1 day from when you solved it. After the second correct attempt in a row, the next problem serving time is determined by the gap between the last two times you saw the problem, and a factor estimating how difficult the problem was for you. The difficulty factor is currently decided by the ratio of your most recent solve time on the problem compared to the average solve time for all users. By default, the problem will be scheduled to be served again at a time that is two times longer than the last time gap between solving it. If you solve it under the average time, then you will not see it again for up to 3 times longer (depending on how much quicker you solved it compared to average). If you solve it slower than the average time, then the gap to the next view will be reduced to as little as 1.3 times the last gap between solutions (again depending on exactly how much slower).

Over time, the problems you know well will get increasingly large gaps between serving times, those you are having trouble with will get served more often, until they too become "well known".

You can use most of the problem search/custom set options with spaced repetition sets, although there are some exceptions. Ordered sets, are not compatible with spaced repetition, as the spaced repetition scheduling will override the sort order of the set. Also some sets such as "problems I’ve never seen before" don’t make sense with SR sets, as they will lead problems to drop out of the set, as soon as you’ve done them the first time. Merged sets are also currently incompatible with spaced repetition sets, so merged sets where one or more subsets are spaced repetition set types will not work as expected.

It can be useful to create spaced repetition sets based on a personal tag as the search criteria , so you can add specific problems to your spaced repetition set by tagging the problems. You can also use other search criteria such as "all forks from 1200-1400" etc to define set membership.

The first step in creating a custom problem set is to create search and ordering criteria that you would like to have used in your custom set. Once the search and ordering criteria are set, you can use the "Search" button to check which problems will be included in your set, and the order of those problems (where ordering has been requested). If you are satisfied with the results shown, you can then use the "Create Set" button to turn the search into a custom set. Choose a name for the set (preferably something that helps you to remember the criteria for the set), and click the "Create" button to confirm the creation of the set.

Once you have created a problem set you need to choose it as your current problem set in order to use it. You can choose to set it as the current upon creation or you can change the set on the solving page or in your preferences.

the Managing sets button launches the manage problem sets panel from where you can delete, merge, view or edit existing custom sets. After selecting a set and clicking show you will be given the search results based on the criteria of the problem set. If you wish to modify the set, you can change any of the options and chose the Save set button to save the new settings over the old set.create a new custom problem set with the modified criteria (the original problem set will not be changed). If you click multiple problem sets, you can also click the Merge button, to merge multiple sets into one "super" set. Please see the Creating Merged Problem Sets section for more details on merging sets.

Merged custom sets allow you to merge several existing custom sets into one merged set. You can select the weights of each subset to control how many problems from each subset will be served. For example you could create three subsets Fork, Pin and Discovered Attack, asking for 70% forks, 20% pins and 10% discovered attacks. Another example would be a merged subset of mate and non-mate sets with say 80% non-mates and 20% mates, you could also merge a larger number of sets, and control mates versus non-mates based on mate length, so for example 80% non-mates, 15% mate in ones, and 5% mate in twos. Problem attempt based sets can also be merged, for example, two sets for problems you have seen before and problems you have not seen before could be merged with say a 90% weight on new problems and 10% on old problems, allowing you to review old problems on average 1 in every 10 attempts. Any existing user created custom sets can be merged as long as all subsets are of the same problem type, i.e. you cannot merge tactic and endgame sets.

To merge sets you need to go to the manage custom problem sets panel (see Managing Custom Problem Sets section of the manual). You can then select the two, or more problem sets you wish to merge, and then click the Merge button. You will then be prompted to enter a name for the new merged set, ad will need to provide percentage weights for each subset. Weights will need to total 100%.

Once you have chosen your merged set as your current problem set, problems will be served up based on the percentage weights you have provided.

Tags that have 2 or more votes in favour than against are shown in bright green, and are considered active. Tags which have at least as many votes against as votes for, are are shown in red and are considered inactive. Tags with only one more vote for than against are shown in drab green, these are "pending" tags. Tags that have been overwhelmingly voted against are not displayed at all.

The number of votes for and against each tag is shown when hovering over the tags. Tags modifiers such as "Check" or the type of piece moved will be shown in the tag name, for example a knight fork that gives check would be shown as Fork[Check,Knight].

Tag voting adds up all the votes for and against a particular tag on a problem and if there is more votes for the tag than against then the problem will be marked as having that tag. This means that if you already see a problem with a tag that you agree with, then it would be useful for you to add your vote for the tag even though the problem was already marked with that tag. The reason for this is that if there was only one vote for the tag and someone else incorrectly voted against the tag, then the tag would become inactive (coloured red), however if you add your vote for the tag then it would take two votes against the tag to disable it.

Users can only vote for a particular tag once on a problem, however users CAN vote for multiple different tags per problems, so if you see more than one appropriate tag for a problem, please vote for all of them.

The number of votes for and against a tag are shown when hovering over the tag. You can click on the + or - votes to add for or against votes to an existing tag, or you can use the + button beneath the tags to vote for new tags, and the available tags to select from will be displayed.

On the list of tags already applied, you can click on the tag name to see a description of the tag if you want to check your understanding of the tag definition before voting. If you haven’t voted on a tag before or haven’t voted recently, you’ll also be shown the description view, even when clicking directly on the up/down vote scores to add your vote.

When choosing a new tag, the tags are layed out in a folder structure in the tag selection panel. Higher level folders such as 'Non-mate Motifs' can’t be voted on directly, however some folders are also voteable tags, for example inside the 'Non-mate Motifs' folder you’ll find a 'Discovery' folder. You can vote directly on the 'Discovery' folder (click on the name instead of the expand/collapse icon to vote on a folder tag). Voting on child tags of such a voteable folder tag will also vote on any voteable parents that lead to the child tag being voted on. For example there is a 'Discover Checks' tag inside the 'Discovered Attack' folder that is inside the 'Discovery' folder. So the 'Discover Checks' tag has the full folder path of 'Tactical Motifs → Non-mate Motifs → Discovery → Discovered Attack → Discoverer Checks'. If you vote on 'Discoverer Checks' then 'Discovered Attack' and 'Discovery' will also have your vote applied. This allows more flexibility in the custom set search, as it means by voting for 'Discoverer Checks' people will also be able to find items you’ve tagged in this way via the higher level motifs of 'Discovery' or 'Discovered Attack'. Down votes work in the same way as up votes in the folder hierachy, so a down vote for 'Discoverer Checks' will also down vote its relevant parents.

Needs more moves and Different Opponent Move tags allow you to add suggestions on how to improve a problem.

Needs more move tags allow you to suggest a move to continue the line when the problem may finish too early to show the point of the tactic. Only the first move in the follow up line should be entered, and when processed later, the engine will try to further extend the line from that move.

Needs different opponent move tags let you suggest a different move for the opponent that might allow the point of the position to be shown more clearly. Only a single move should be entered, and it should replace an existing move by the opponent in the main line of the tactic.

Personal tags allow you to create your own tags and tag problems with those new tags.

Personal tags are most powerful when combined with custom problem sets and problem search. For example a personal tag can act as a bookmarking system where you can create a 'favourites' tag, and tag problems with that tag that you can search for in the problem search/custom sets page. Another example is a 'review' tag that you use to tag problems you feel need reviewing at a later stage. You then create a custom set that matches against problems with your 'review' tag. As you go through the set, you can remove the reviewed problem from the set by voting down the tag. This could be combined with the 'Problems I always got wrong' criteria to produce a review custom set that keeps giving you the review problems until you’ve successfully got them correct, avoiding the need to manually remove the tag yourself.

To create a new personal tag, click the Manage Personal Tags icon in the tagging panel, and then click the Create Tag button. After creating the new tag, you can start adding it to problems in the same way you add existing tactical motif tags.

Tag modifiers allow more specific versions of tags to be represented without creating a massive number of new top level tags. For example it may be useful to recognize that a motif occurs with check, or that a particular piece was moved during the tactic. For example a Fork with check could be presented by Fork[Check,Knight] indicating that a Knight delivered a fork with check.

Modifiers can only be added when creating a new tag. You can’t add a modifier to a tag that has already been added to a problem. This means if there is an existing Fork tag with no modifiers, and you want to add a Fork[Check,Knight] tag, you would need to add a new tag using the '+' button , and then on the add tag panel, choose the Check and Knight modifiers. We also want to avoid an acculumation of the same tag with different sets of modifiers , so if you believe your new tag with modifiers is the most specific available, then you should also vote down Fork tags without modifiers. The tag without modifiers is not wrong, but your new tag is more specific, so we would prefer that it becomes the primary way of tagging the Fork on the problem. People using tag searches based on 'Fork' will still find the modifier version, with the added bonus of also being able to search for problems with more specific versions involving check or the specific piece involved. If other taggers agree with your modifiers, they can click directly on your modifier version of the tag to vote it up without having to add their own modifier version, and they can also vote down the less specific version.

We currently support three types of modifiers, the Check modifier, the attacking piece modifier (Pawn, Bishop, Knight, Rook, Queen , King) and the Prophylaxis modifer. The Check modifier is to be used when the motif involves checking the opponent king. The attacking piece modifier designates the piece(s) involved in the attack. The Prophylaxis modifier provides a way of indicating that a move is required in order to protect against a potential attack from your opponent. It is most commonly applicable on mixed mode defensive problems, but may also be relevant in non-defensive tactics.

Please note that piece modifiers (Pawn, Bishop, Knight, Rook, Queen , King) only refer to the moving piece, please don’t use these to denote the type of pieces attacked. So for example a Fork[King] is a fork made by a King on two pieces, NOT a fork on a king. Similarly Pin[Rook] is pin where a rook is pinning another piece, NOT a pin on a rook. You can use the piece modifier for opponent pieces when combined with the Prophylaxis modifier, indicating the opponent piece is the piece capable of making the potential threat. Similarly if a move is a prophylactic move that defends against check from your opponent, then tagging with both Prophylaxis and Check as modifiers is also fine.

For a small number of tags , modifiers will be redundant , as we decided to have a specific tag that overlaps with a modifier. For example 'Discoverer Checks' could be 'Discovered Attack[Check]'. To avoid clutter it is best to avoid marking 'Discoverer Checks' with the Check modifier, given the name already holds this information.

Initially, positional motifs were only available for tagging positions in the Guess the move feature. However we now allow positional tags to be used in chess problems. When deciding if a positional tag could be used on a problem, please try to limit tagging to features that were relevant to the problem solution. For example, if a tactical position has connected pawns, but they don’t feature in the tactic, then they should not be tagged. If however, the tactic involves utilising an advanced pawn to gain material, it is possible the connected nature of your pawns was relevant, and if that is the case, they should be tagged. Some positional motifs will tend to be more commonly associated with tactics than others. For example motifs in the positional list like 'Rook On 7th Rank', 'Pawn Storm', 'Rook Lift' will tend to be more commonly featured in tactics problems than some of the less aggressive positional tags.

This panel offers options to control how problems are created from games you have played on chesstempo. Hover over the ? button next to each option for details. Click the x button to reset the value to the default.

Click on the 'manage your students' button to manage the list of students.

See Student coach section on coach nomination for more details.

Parent controls allow you to specify another chesstempo account as the 'parent' of the current account, and thus providing parent control options that only the parent can turn on and off.

To enable the parent control options, enter the username and password of another chesstempo account to be used as the parent account set to control this account. Once set, the parent password must be entered in order to view and change parent control options.

Playing preferences covers settings for play online feature, and are broken down into sub-tabs.

Account preferences controls settings specific to your Chesstempo user account, such as email address and password.

The stats home view provides an overview of the user’s performance across a variety of areas. If there is a panel on one of the other pages that you use often you may want to use the 'add stats' button to insert the panel into the stats home view.

The default panels available on the stats home view are:

The tactics stats shows problem solving stats specific to tactics problems such as rating graphs, and rating type summary data for Standard, Blitz and Mixed modes (including FIDE estimates based on performance on each of these rating types) and many other stats and graphs.

The currently available panels are discussed in the next section.

The endgame stats page provides most of the stats panels available on the Tactics Stats page, except that it does not include the Tactical Motif performance table (replacing it with an endgame types performance table), and does not have a tactics mistakes heatmap. Please see the Tactics Stats section for descriptions on the equivalent endgame panels.

The playing stats view provides a breakdown of the performance in games played on chesstempo across the available range of time control/rating types (i.e Bullet, Blitz, Rapid, Long and Correspondence).

Some of the playing stats provided overlap with the old 'Your games' view on the old site game database. These stats are not meant to be a complete replacement for the game database view of your games, and a new site version of the game database will also include a version of the current 'Your games' view, which will likely borrow some of the functionality of the stats provided here.

Repertoire stats are available to provide details on activity related to the opening trainer. Several stats are already available directly on the opening trainer page, and for now these will remain there.

The book stats show stats related to activity associated with chesstempo books. Note that some of the other panels can also be used for visualising performance related to book related content by using the specific book filter, the general 'only from books' filter or by filtering for problem sets associated with a book.

The achievement stats page shows the list of completed and incomplete achievements. By default the completed achievements are shown, click on the 'incomplete achievements' button to view the not yet complete achievements.

The activity stats panel summarises the underlying data collected in order to provide the achievements data, and provides activity specific statistics for a wide range of side activities, such as current and best activity streaks, correct/won streaks, records for number correct or games won over different periods of time, total time spent on activity etc.

The student stats panel shows data related to the students who have nominated you as coach. See the student/coach relationships section of the user guide for more details on the information available here.

The guess the move stats section holds stats panels specific to moves you have made in the guess the move feature.

The GTM rating summary and rating graph shows the same data shown in the equivalent tactics stats panels.

You can see your completed achievement badges by going to the achievements tab on your stats page. Click on the incomplete acheivements tab to see a list of all the incomplete achievements and an indication of any progress that has been made towards each badges completion.

If you’d like to see details activity stats that the achievements are based on , you can see the details on the activity stats tab of your stats page.

The achievement points leaderboard allows you to see a list of the top point scorers for the week, month, year and all time. Your current rank and number of points for the period are also shown.

The engine analysis panel shows engine analysis results, and allows you to start and stop analysis, control the number of lines analysed, and initiate full game annotations.

The analysis panel has the following buttons:

The engine lines display shows an evaluation and best line for the top moves found by the engine in the current position. The moves are ordered from best move to worst move, and the maximum number of moves shown is determined by the currently selected number of lines. The order of display of each line is, Analysis Depth, Evaluation, Move, Best Line following move.

The evaluation shown before each line indicates the best material situation that the engine thinks will result from playing the particular move (i.e. the evaluation of the position at the end of the best line shown). The evaluation shown is in pawn units. So for example +1.00 would show the move leads to a position that is 1 pawn ahead for the player playing the move, -1.00 would indicate the move leads to a position that is 1 pawn behind for the person playing the move. +3.40 would show a position where the player to move is up a piece worth of material, plus some extra evaluation attributed to some positional advantage in the current position. By default, evaluations are shown from the player to move’s point of view, this means a positive evaluation indicates a move that leads to a position where the player to move is in front, and a negative evaluation indicates a resulting position where the player to move would be behind. Many chess engine interfaces have an alternative form of display where the evaluation is always shown from white’s perspective. For example if it was black’s turn to move a +1.00 score would not be a 1 pawn advantage to black, but instead a 1 pawn advantage to white, with black being behind. If you’d prefer to use this method of evaluationd display select the 'Evaluation from white’s view' option available on the "Chess Engines" tab of your user preferences (not the engine settings page).

Moves that would lead the player to move to be ahead of their opponent are marked green, and moves that would lead the player to be behind are marked red, moves deemed to lead to even or drawn positions are marked blue. Note that 'green' moves are not necessarily 'good' moves, for example if the player is a rook up and a move might lead to being only 1 pawn up, that move will be marked green as the player is still in front, but it would not be a good move.

The list of moves in each engine line will be coloured orange up to the point that they match the line played in the current list. You can click on any move in the best line display and the moves upto and including the clicked on move will be inserted into the move list for the current board.

During analysis, stats on the progress of the engine are displayed above the engine lines. The current time spent analysing the current position is displayed, along with the current move being examined (and the number of this move out of the list of candidate moves for the position). The Nodes show the total number of positions examined, and the Nodes per Second (NPS) shows the number of positions the engine is examining each second. The NPS will vary depending on the speed of your machine, how busy your machine is with other tasks, and the nature of the position being analysed.

The annotate feature allows you to use your engine to automatically annotate all the moves in a game. For example after using the Explore button on the play versus computer page to examine the game in the database, you can use the annotate feature to ask for a full game analysis from the engine, with blunders and the improved lines inserted into the move list. You can also paste games you have played elsewhere into the game database move list using the 'paste' button beneath the board, and analyse those games inside the Chesstempo web page.

To initiate game analysis click on the Annotate button which will launch a dialog that allows you to select game annotation settings. The following settings are available:

To use the new coaching features, students must nominate their coach (or coaches) on the "Coaches" tab in their preference settings. More than one coach can be nominated, by entering a comma separated list of coach usernames, for example "john,bill". To use these features, both the coach and the students must have active gold or diamond memberships.

Once a coach is nominated, the coach must confirm the nomination. To do this, the coach needs to go to the 'Coaches' tab on their settings, where they should see a "Manage your students" button. Click this to see a list of all students, and select the students marked as requiring confirmation , then click the 'Confirm' button (or 'Reject' if you don’t want to add them to your list of students). The same panel can also be used to remove students that you no longer which to coach, again select the students you wish to remove and then click the 'Drop' button.

The Students tab on the coaches' stats page has a list of all their current students. The list shows a summary of recent accuracy, and also contains a link to each student’s stats page. Coach’s can access several stats usually only available to the user themself, including:

The student list has a preferences button next to each student which allows the coach to edit the student’s settings. Clicking on a student’s preferences button brings up the preferences panel for the user. Any user preference can be changed (other than the Student’s list of coaches). Once you have completed the changes, click the save button to finish. Commonly changed settings would be difficulty settings, and current problem set.

Coaches can assign problem sets in two ways. Firstly, when creating a new problem set, coaches with students are given the option of assigning a problem set to a user. The assignment choice is made after using the Create Set button. Sets assigned in this manner belong to the student, rather than the coach. Problem sets can also be assigned via the preferences button next to each student in the coach’s student list. The current problem set can be changed and saved, just as any other preference, and this allows you to assign previously created sets to the user.

When the playing page first loads, the presentation will differ depending on whether you have any games currently active or not. If you have one or more currently active games, the game with the most urgent move requirement will be loaded first. If you have more games where it is your turn to move, the next button will navigate to those games, or you can bring up the My Games view via the blue mode menu in the top left of the page, or via the g key, and select an individual game to jump to that game.

If you have no games to be played the challenge list display will be shown. The Create challenge button will create a challenge that others can respond to. Responding to an existing challenge by clicking the challenge, will lead the new game board to appear. You can toggle back and forth between the game board view and the challenge view with the c key.

There are several ways of starting a new game:

Please see the Challenge Dialog, Challenge List,User List, and Friends List sections of the user guide for more details.

The challenge dialog is used to issue challenges to other users. Game challenges can be direct challenges, sent to a particular user (by entering their name in the opponent name field), or an open challenge available to any user who matches the challenge parameters.

You can issue a challenge by clicking any of the time control buttons in the Challenge Dialog, and then clicking the Ok button to confirm. Change the time control type to see default time controls for Bullet/Blitz/Rapid/Long/Correspondence.

If you want to create a challenge using the same parameters as your most recently created challenge, then click the 'Ok' button immediately after the challenge dialog displays (the Ok button label will show the current time control that will be used). To avoid confirming the challenge with the 'Ok' button, you can double click on the challenge button which will automatically create a challenge of that time control.

If a challenge created by another user matches your game challenge parameters, you will be immediately matched up to play, otherwise your challenge will be added to the list of current challenges, and a new game will start when someone either explicitly accepts your challenge, or creates their own challenge with matching parameters. Automatic challenge matching will only occur on challenges created by human players, challenges created by bots need to be accepted manually, and will not be used to match your own challenges.

When you logout (or disconnect by closing your browser or losing your network connection), any pending non-correspondence challenges will be automatically cancelled. Correspondence challenges will survive logout, and will persist in the system until you either cancel them, or they are accepted by another user.

The custom time control type allows you to create challenges with a time control not supported in the quick select panel.

The history view shows a list of your games, with the most recent shown first. The history view can be accessed via the blue action menu in the top level of the playing page, or by using the h key. You can select individual games for viewing by clicking on the game board which loads the game into the game view. By scrolling down the page more game boards from your history will be loaded into view.

For more sophisticated history options, and the ability explore your games in the opening explorer go to the Database › Your games top level menu option.

The player list panel shows the list of currently online players, and their rating for each rating type. To access the player list panel choose the Playing view menu: › User list option from the blue action menu in the top left of the page.

You can challenge a player to a game by clicking on their Challenge button. Other player specific actions, such as friend requests or bringing up the player’s stats page can be accessed by clicking on their username.

The player list can be filtered by username or rating range by clicking on the filter icon in the top left corner of the player list panel. To filter by rating range, you will need to select the rating type the range applies to. If you’d like to ignore the rating range filter, choose the 'Ignore' type for the rating type. The username filter is a case sensitive substring search, so for example "ern" would not match "Ernest", but "Ern" would.

The challenge list shows a list of all the currently available game challenges. Only the challenges that have a filter matching the current user will be shown, so you only see challenges you are able to accept. Click on any challenge in the list to accept that challenge. You will need to confirm your choice to avoid accidental challenge accepts, you can avoid the confirmation step by double clicking instead of single click on the challenge.

Your own challenges will be shown at the top of the list, and can be cancelled by clicking on the challenge.

Challenges can be filtered by rating range and whether they are rated or unrated by clicking on the filter icon in the top right corner of the challenge list panel. You can also sort the challenge list by clicking on the column header of the criteria you wish to sort by.

The friends list provides a view of your list of friends, and shows if any of them are currently online. Online friends will appear in the top half of the list, and will have a Challenge button next to their names that can be used to send them a direct game challenge. Friends can be added by clicking on any username on the playing page and from the player actions menu selecting the Player actions › Friend request option. The friend candidate must accept your request to move from pending to full friend status.

The All Games list lets you see all the games currently being played. You can click on any of the game rows to start viewing the game.

The playing list can be filtered by username substring search, and rating type.

The my games panel allows you to see all your currently active games. Click on any of the game boards to load the game.

Games are ordered first by those where it is your turn to play first, and then by the urgency based on time left to play.

As moves are played, both by you , and your opponent, the games will re-sort the games so that the most urgent games where it is your turn are always sorted first.

The Chesstempo online play interface is made up of the board and clocks, the move list, in game chat, and navigation and action buttons. The board is resizable by dragging the drag handle on the right of the board. Resize settings are remembered via a browser cookie, so will be used when you next use the site from the same machine and web browser.

The Vacation/Sleep feature allows you to take time out from correspondence time control games, freezing your game clocks during the period of the vacation or sleep. Clicking the Vacation/Sleep button in the top level playing menu (only available when already on the playing page) will show a dialog indicating the amount of sleep and vacation time you currently have available, indicating whether you are currently "alseep" or on vacation, and allowing you to return from, or go into sleep/vacation mode.

The ranked list of all online players is available from the "CT players" item on the top level Playing menu. You can search for users by username (substring search), rating range , and activity level. Activity level is defined by the user’s RD, the lower the RD, the higher their activity level. Activity level filtering applies to the most active rating in any of the rating types the user has played games in. The player list can be sorted by clicking on the relevant table headings.

Most of the places showing player names in the online play user interface can be clicked to show a menu of actions, one of which is to display the user’s stats page. Clicking on the "Stats" option will open a new window with the players game archive and stats page. Users can access their own page by clicking on the "My Chess Games" option in the Chess Database menu. This page is where the player’s previous games can be accessed, and stats related to each of the rating types is shown. Clicking on the details button for a rating type shows extra stats and a rating graph for that rating type. The game archive and opening explorer can be searched/filtered using most of the tools and options available in the main game database. This allows you to do things like examine your opening explorer stats against a particular player, or against players in a particular rating range, or look at your opening explorer stats for games played more than a year ago, and compare them to your most recent games. For more details on the generic database functionality please see the Chess Database chapter.

The opening explorer on the player game archive page has a Rating Type filter above it which allows you to filter games based on the rating type of the game. This selector also applies to any game archive searches, as well as the opening explorer stats, allowing you to see opening stats specific to each rating type.

The game archive page also has two "Games for position" tabs, one shows only games from the player’s own archive (labelled as "Games for position: USERNAME"), and the other shows games from the main chesstempo database. By switching between both while traversing the moves of one of their own games, you can look at either your own previous games that followed the current line, or look for master level games that also followed your line, and these may provide hints on improvements that could be applied next time you play, or give you an idea on the kind of middle or endgames that arise from your openings when played by masters.

Chesstempo uses the following rating types to help differentiate performance on games with different time controls:

Estimated time for games with no increment is simply the initial time on the clock. Time for games with an increment are calculated by assuming the game will go for 40 moves. For example, 5 minutes start time with a 12 second increment would lead to an estimate time of 5*60+12*40 which equals 780 seconds, or 13 minutes, leading to a classification as a blitz time control. Increasing the increment to 18 seconds would produce 5*60+18*40, giving an estimate of 17 minutes, which would now be classified as a rapid time control.

The challenge (or seek) formula allows you to create a description on the kind of challenges you will accept from direct challenges sent by other players. Challenges not matching your formula will have a challenge rejection message automatically sent to your opponent indicating that your formula does not currently match their challenge request.

Challenge formulas can be very expressive, and can combine a number of logical and mathematical operators to produce complex challenge preferences such as : (playing_time_left>5mins or (from="bjork" and time=60mins)) or correspondence would only accept challenges that are either correspondence challenges or non-correspondence challenges where you either have more than 5 minutes left in the current game you are playing, or the challenge is from a person with the username "bjork" and she is asking for a game with initial time set to 60 minutes.

You can change your challenge formula using the Challenge Formula option in the playing menu at the top of the page (only available when on the playing page). The default challenge formula looks like this: ` !playing or (playing_time_left > 15 minutes and (correspondence or playing_correspondence)) `

This only allows challenges if you are either not actively playing a game in the current game tab, or the game you are actively playing has more than 15 minutes left on your clock, and either the incoming challenge request is for a correspondence game, or the game you are playing is a correspondence game.

This default rule prevents you receiving challenge requests for any type of game if you are low on time in an actively played game, and prevents receiving challenges for live games when you are already playing a live game. If you are not low on time, then if you are playing a live game you can still receive correspondence challenges, or if you are playing a correspondence game then you can receive challenges of any type.

To start opening training you first need to create a new repertoire. Click on the New repertoire button to create your first repertoire. Repertoires can be either black or white repertoires, so choose a repertoire name and colour and click the Create button. If you already have a repertoire in PGN format, you can also import your repertoire from PGN by clicking on the Import PGN button.

To add new moves to your repertoire simply make the move on the board and it will be added to the currently selected repertoire. When you have some moves entered, you can enter training mode by clicking on the Train or Train branch buttons. Train branch will only train moves from under the currently selected move in the opening tree panel. See the training mode section for more details on training.

You may also want to explore the different settings that can be used to control training and viewing your repertoire. Settings can be viewed by clicking on the blue circle menu under the board and then clicking on the settings icon.

When creating a new repertoire, you can use an existing PGN file as the import source. If there are multiple games in your PGN file , and the file is not the result of a Chesstempo opening repertoire export, then each game will be interpreted as multiple lines for the same imported repertoire.

By default, any comments from your repertoire will be imported as private comments. If you’d like your PGN comments to be visible to others or ignored completely, you can choose those options on the create repertoire dialog. NOTE: If you are using a PGN with comments made by someone else, such as a chess book author, please do not make the comments public. If you are an author and find any of your annotations being used in comments on Chesstempo, please contact us to have them removed.

By default, any alternative moves at the same position for the colour of your repertoire will be imported disabled, but you can choose to have them imported enabled as alternative lines to train or not imported at all.

There is no size limit on the PGN, but there are quotas for the total number of positions you are able to have across all your repertoires. For details on the quotas which differ by membership type , please see the Chesstempo memberships page.

The quota for free accounts is fairly generous (currently 20000 positions can be stored). However you should be careful when importing repertoires that you are not trying to import a large collection of games. Loading repertoires made up of entire games, including the middle game and endgame parts of each game will eat up your quota very quickly with positions that are not useful for training purposes.

You can enter a new move for a repertoire any time you are not in training mode by playing the move on the board. Played moves are automatically stored in your repertoire, and saved on the server. Moves will be attached to the currently selected repertoire which is shown in the repertoire tree to the left of the board.

To delete moves (or branches of moves) right click on the move in the opening tree and choose the delete option. Deleting a non-final move in a line will also remove all sub-moves and variations that appear after that move.

Training mode is where you practice the moves in your repertoire. Training mode can be entered by clicking on the Train or Train branch buttons. Train branch will only train moves from under the currently selected move in the opening tree.

You can choose to add opening explorer statistics to your repertoire, which , as well as provide extra information on each move, enables several other features, such as common positions first learning prioritisation, colouring options in the opening tree and opening sunburst visualisations, and extra sorting and filtering options in the move list table.

To initiate the opening explorer sync, click on the blue circle menu launcher under the board, and select the sync explorer option. Several options are available for the sync:

Using the sync evals feature you can fill in pre-processed engine evaluations for your repertoire moves. Having evals synced with your moves allows extra colouring options for opening tree colouring and opening sunburst visualisation colouring , as well as sorting options on the move list table and a bulk disable based on engine evals.

Chesstempo has a large cache of pre-cached engine evaluation for many opening positions of varying depths. We have been using a portion of our idle cloud engine servers to analyse the most common opening positions at high depth (depth 50 using Stockfish). We’ve been running this high depth analysis for over a year, but such depth takes around half a day for each position, so even with many engines processing these requests in parallel there is still a lot of work to be done to cover all common positions. We also have opening position evaluations cached from some of the cloud engine analysis done by users of our cloud engine features, that is often of lower depth (although occasionally higher depth analysis is also available). We have also recently released a feature for diamond members that allows them to request bulk analysis of their repertoires at depth 25, and once cached this analysis is also available to other members.

To start the engine eval sync, click on the blue circle menu launcher under the board, and select the sync evals option. Several options are available for the sync: Only add evals to moves without existing evaluations:: Use this option to preserve any existing evaluations. Because you can manually add engine evals or export them via PGN when importing a repertoire from PGN you may want to use this option to make sure pre-cached evaluations do not override your hand chosen evaluations. The option is on by default, if you want to overwrite existing evals during the sync, turn this option off. Minimum engine depth to use:: Only sync pre-cached engine evaluations that were analysed to at least this depth.

The sync evals feature provides repertoire engine data for positions that already have pre-existing engine analysis. Diamond members also have access to a feature that provides engine analysis for all moves in a repertoire, filling in the gaps for moves that don’t already have pre-cached analysis. To use the engine analysis feature right click on any move in the opening tree panel and choose the 'analyse line' option. If you want to analyse the entire repertoire, click on the top level move to analyse the entire tree. In order to manage load on our cloud engine servers, there is a 5000 daily quota on moves analysed. Analysis is done with the Stockfish engine to depth 25. Note that the quota only applies to moves that don’t already have at least depth 25 evaluations cached, and with average sized repertoires, 5000 moves may be enough to cover all moves without prior evaluation in one day. Over time, as more moves are cached, we expect the 5000 limit to be sufficient to analyse entire repertoires in the vast majority of cases, with only those with an unusually large number of novel positions needing to spread the analysis out over more than one day.

If you do need to analyse over several days, you can analyse again from the top level move in your repertoire, and the system knows which moves have already been analysed and will only look at the yet-to-analysed moves in subsequent analysis requests (i.e you don’t need to remember which part of the tree you have already asked for analysis on). If you want to monitor how the analysis is going, you can request another analysis, and you’ll get feedback on how many items are still in the queue waiting to be processed (as well as details on the remaining quota for the day).

Once all positions have been analysed (requesting a new analysis will tell you if that is the case), you need to use the sync evals feature to initiate the final sync of the processed moves with your repertoire, after a page reload you will then be able to visualise the opening tree and opening sunburst using the eval colouring method, and sort the move table based on evaluations.

The opening sunburst provides a method of visualising your opening repertoire and various metrics such as engine evaluations , opening explorer stats and spaced repetition learning metrics. It lays out your repertoire in a radial tree layout which allows a more compact overview than can be provided by the opening tree. To launch the sunburst click on the eye icon in the header of the opening tree panel.

If your repertoire is of a reasonable size only a sample of the entire repertoire will be displayed, with priority given to moves whose sub-branch is a larger percentage of the total repertoire. You can click on any move in the sunburst to make that move the centre of the sunburst, thus drilling down into the move’s sub-lines. Clicking on any move will also jump to the position in the opening tree.

The board in the centre of the sunburst shows the position after the move currently hovered over with the mouse. The full line for the hovered move is shown below the board, along with the opening name, and opening explorer stats and engine eval if available. The spaced repetition based statistics for moves you’ve attempted at least once will be shown towards the top of the sunburst when hovering over the move.

If you’d like to return to the previous set of data after drilling into a move by clicking on it, then you can zoom back out by clicking on the board in the centre of the sunburst. Subsequent clicks on the board will continue to zoom back out until you reach the initial position again. If you’ve drilled several moves down and want to jump back to the start position quickly, use shift-click to achieve that. There is a resize handle in the bottom right of the sunburst if you need to resize, please note that you may have trouble making out the moves when a smaller size is used, if that’s the case you may want to change the "Maximum rings in sunburst" setting to limit the total rings displayed at a time.

You can click on the settings icon in the top left corner on the sunburst to configure the sunburst evaluation, available settings are:

There is also an opening trainer wide setting on the opening trainer advanced settings that controls the opening explorer based stats, and can be used to set a minimum number of games before a stat will be used, this can be useful for the win/rate, and opening performance rating colouring methods as smaller game numbers can produce misleading colouring due to small sample sizes, so considering using this setting (which is off by default) if you find low-sample size colours confusing.

The move list table provides a linear list of all the moves in your repertoire. Providing data on each move such as the engine evaluation and opening explorer stats (where available) and the spaced repetition learning metrics for each move. The move list table can be launched by first selecting the start of the branch you want to exam (use the top level move if you want to see all moves in your repertoire) and then click on the list icon in the header panel of the opening tree. Once launched click on the headers of the table to sort by those metrics, and click on the settings icon in the top left of the table to access settings and filters for the table. The forward/back arrows next to the settings button moves forward and backwards one move at a time in the table, selecting the move in the opening tree (or you can click directly on a move in the table to jump to that position in the tree).

By combining sorting (via clicking on column headers) and the available filters the move list can be used to identify important aspects of your repertoire. For example if you want to review all the bad moves (according to Stockfish) in your repertoire click on the eval header to toggle between sorting by increasing or decreasing eval. This allows you to have the worst moves shown first in the list. You may also want select the 'only show our moves' setting, to focus on moves of your colour that the engine thinks are poor. You can then tarverse through the list of bad moves and decide if you want to keep them in your repertoire, or replace them with better options. Similar filtering can be done with opening explorer based settings. You can also zero in on move you are performing poorly at in spaced repetition by sorting by those columns.

Another example of using the move list table is to find moves where you might get the most benefit from extending the line further. If you enable the "only show end-of-line" moves option, and then sort by the "opex #" column (i.e the number of times the move appeared in the opening explorer data that was synced with your repertoire) then you can see which lines ended in positions that were still very common according to the opening explorer data, and these positions may provide good extension points, as they are more likely to come up in actual play, so extending the line further is more likely to provide a return on your learning time investment than extending a line that rarely comes up. When using explorer based stats, you may also want to filter based on the minimum explorer games setting to completely avoid seeing moves with less than minimum number of moves.

By default the move list table will also show disabled moves, but you can hide those by enabling the 'hide disabled moves' option in the move list table settings.

If you have already synced engine evals and/or opening explorer stats with your repertoire (and subsequently reloaded the page) you will be able to disable moves in your repertoire that do not meet a minimum engine eval target or number of games played in the opening explorer stats.

To initiate the bulk disable first click on the first move in opening tree of the branch in your repertoire you want to disable (click on the first move if you want to consider the entire repertoire for disabling). Then click on the blue circle menu launcher under the board, and select the bulk disable option. Choose the pawn evaluation threshold and/or the minimum number of games the move has been played in for the disabling criteria (any move worse than the stated eval threshold or played less number of times in the opening explorer stats will be disabled). You can also choose to disable moves of either colour or just your own moves. Only moves from the currently selected repertoire will be considered for disabling.

This option can be useful for reducing a very large repertoire down to the essential lines for memorisation based on how likely you are to see them in your actual games. Disabled moves remain in the tree for instrucitonal purposes, but are not included in training.

When paused or not in training mode, details about the move just played are shown under the board, this includes: - Accuracy - Last time the move was seen - Number of times the move was wrong or correct in a row. - When it is next scheduled - The mistake history for the move, which shows all the incorrect moves that have been made in the current position and their frequency and most recent date played. Note: This only applies to moves made after the 25th of April 2020, and currently only collects mistakes made on the desktop version, the apps will also track mistakes after the next update. - Engine evaluation of the move if available (see the syncing engine evals section). - Opening explorer stats for the move if available (see the syncing opening explorer stats section).

Any comments you have made on a position or move in the comment panel, or comments imported automatically from a PGN import will be made visible in the personal comments panel. You can control when these items will be visible via the 'Show our move comments' setting in the opening trainer settings panel.

You can draw arrows and highlight squares on the main opening trainer board using right click (square highlight) or right click and drag (draw arrow), combined with Ctrl/Shift/Alt key combos for choosing the colour of the square or arrow.

These board annotations can be permanently saved on the move for this position by pressing the 's' key after you’ve finished making the required graphical annotations, and will be available when traversing the move list and included any PGN exports.

When you have one or more opening repertoires created, they will be used after each game you play on Chesstempo to provide feedback on where the game deviates from your repertoire. A deviation will be either where you play a move that was different than the one recommended in your repertoire, or your opponent plays a move that you don’t have in your repertoire yet. You can choose to add new moves via the opening explorer without leaving the playing page, or launch the opening training page with the relevant repertoire selected.

To launch the settings panel, click on the blue circle menu launcher under the board, and select the settings option.

The opening tree shows all the moves for the currently selected repertoire (or repertoires when a global repertoire like all white repertoires is selected). Clicking on any move will jump to that move on the board, with the moves required to lead to that position shown in the move list. You can start training moves below any position in the tree by clicking on a move and then clicking on the Train branch button.

Moves can be deleted by right clicking on any move in the tree and selecting the delete option. That move and all moves under it will be deleted. On mobile devices right click can be simulated by pressing and holding on the relevant move in the tree.

Parts of the opening tree can also be disabled or enabled by right clicking on the move, and selecting the enable or disable line option. Disabling allows you to keep lines in your repertoire that will not be trained, and can be useful for example when you may want to list moves for your own colour that are poor, with comments explaining why the move should be avoided. You may also want to temporarily disable lines during training if you are not currently planning to play them, but may want to do so in the future.

Lines can be copied from one repertoire to another by again right clicking on the move to start the copy from, and then choosing the Copy line option. The destination repertoire will be prompted for after the option is selected, after the copy the line will exist in both the original and destination repertoires. When copying, it is the sub-tree under the selected move that will be copied. In addition, the moves from the selected move up to the root of the opening tree will also be copied to ensure the destination repertoire has a branch on which to attach the copied sub-tree.

The repertoire tree holds all your white and black repertoires, as well as the global repertoires such as the resistant repertoires. Clicking on the white or black repertoires folder will fill the opening tree with all moves from all repertoires of that colour, and allow you to train across all the repertoires.

If you’ve created any shortcuts, they will be displayed under their parent repertoires. Clicking on a shortcut will load the parent repertoire into the opening tree (if not already loaded) and jump to the shortcut position.

Each repertoire with positions currently in learning state will have either a number or time shown to the right of the repertoire name. A number indicates the current number of positions due for presentation via the spaced repetition algorithm. If none are due it will be a time until the earliest item is scheduled due.

Note that these scheduling stats are not updated dynamically so unless you visit the repertoire, the number may say Caro-Kann (7) for example, but 7 is the number scheduled when the page loads, so you may find when you click on the repertoire the number scheduled has grown if positions have moved into being scheduled since you loaded the page. Similarly, if it says Caro-Kann (20 minutes) to indicate the next item is due in 20 minutes , while the 20 minutes will dynamically update, when it gets to 'now' it will not change to 1 but rather start to show how long ago the most recent position was scheduled, e.g Caro-Kann (5 minutes ago).

Repertoires are currently sorted alphabetically, and we don’t have way of controlling the order, until we implement explicit ordering control, you can force an ordering by utilising the alphabetical ordering. For example by using numbers in repertoire names. "1. Repertoire 1" will sort before "2. Repertoire 2" or "AA Repertoire name" will appear before "BB Repertoire name". You can right click on a repertoire name in the repertoire tree to access rename and delete commands.

With a large repertoire you may want to bookmark parts of the repertoire to allow you to quickly jump to sub sections with a single click rather than having to move through the tree or hunt for the move in the tree. Shortcuts allow you to create bookmarks for any position in the repertoire. To create a shortcut, use the add shortcut button in the action menu under the board. In the repertoire tree the shortcut will be added under the folder holding the currently selected repertoire. Clicking on the shortcut name will jump to the repertoire and position. To delete a shortcut right click on the shortcut name, and select the delete shortcut option.

If you’d like to organise a repertoire’s shortcuts into sub-folders, choose the name of the sub-folder to place the shortcut in when creating the shortcut. If a sub-folder name is not selected, the shortcut will be added directly under the repertoire in the repertoire tree. Currently you can’t reorganise shortcuts once created, so the only way to move a shortcut into a sub-folder is to delete the old shortcut and create it again using a sub-folder name. Shortcuts are currently sorted alphabetically, so if you want to control the order you need to do so via naming, for example 'A) Variation' will appear before 'B) Variation'. We hope to improve shortcut organisation capabilities in a future update.

The opening explorer provides a view of the popularity and performance of moves in the current position. If you are considering adding a new move to the repertoire, it can be used as a guide to see what master level players are choosing to play in the position, and their success when playing each move. You can also turn on the option to show the most popular move in the explorer as an arrow on the main opening trainer board.

The candidate moves list tells you which moves are available in the current position. Usually this panel is hidden during training, however you can choose to unhide it if you feel you need the hints. When outside training mode, you can use the candidate moves to traverse your repertoire. Clicking on a move will play that move. The up and down arrow keys will change the currently selected move, and the forward arrow will play the currently selected move. The currently selected move will default to the next move in the move list if you are currently traversing a move list. Moves in the candidate move list are labeled with the repertoire they come from. If you play a move that appears in a different repertoire from the one you have currently selected, that repertoire will become the current repertoire, and the opening tree view will be updated to reflect the change of repertoire.

For premium members, the engine panel provides a way of asking for an opinion from the engine on the current position. If you’d like to add the start move of a line shown by the engine to your current repertoire, click on the move and it will be added to the currently selected repertoire. The engine evaluation will be stored with the move, and will be available in any exported PGN. If you’d like to add an engine evaluation to moves not added from the engine analysis, you can do that via the 'E' key when the move to be assigned an evaluation is selected.

If you press the 'a' key or launch the annotation window from the move list blue action menu , the annotation panel will load , allowing you to add annotation symbols to moves such as '!', '?', '+=' etc. Like engine move evaluations , these are permanently stored and available in any PGN exports.

Comments can be made on either the move or position level. A move comment should be specific to the move just made, whereas a position comment applies to the current position, and is not necessarily tied to a particular move (although it may mention one or more moves in the position).

You can choose to have your comment public or private. If you are making comments from a book you own, and are copying the analysis thoughts of the book author in a way which would be considered plagiarism, please keep the comment private.

The comments for the opening trainer are displayed in a slightly different way to the comments on other parts of the site. Instead of always ordering the comments in chronological order, the system chooses what it believes to be the best comment, and then displays the other comments in chronological order after that. Any comment you’ve made yourself will be considered to be the 'best' comment when you are reading comments, if you have not commented on a position then the system chooses the comment with the highest vote sum. If a tie break is required, then the comment makers with the highest problem rating (in any problem solving mode) is chosen.

Due to the above ranking method, comment voting in opening training should be approached in a slightly different manner to problem comments. Because we want the most useful comment to be at the top, if you see a comment that you believe is better than the best comment , even if you think the best comment is ok, then you should down vote the current best comment, and up vote the comment that you believe to be better.

If you’d like to see the current 'best' comment quickly without opening the comments panel, you can do so by pressing the 'c' key which will show a pop-up panel with the best comment. Press 'c' again to quickly hide the panel.

There is also an option to show your own comments in a separate panel below the move list which you can turn on in your settings (by default your own comments will be shown while viewing moves in non-training mode).

To access the graphs click on the blue action menu under the board and select either the global stats or repertoire stats option.

The move score is shown after each move you make, and is displayed above the move list. Moves are scored from 0-12, with 10 being a move that matches the master move, and 0 being a potentially game losing blunder. A bonus score of up to 2 is given when you play a move that is considerably better than the move played by the master. If you played a move that matched a relatively main line opening move, you are also given the same score that you would have been, had you played the master move, irrespective of whether the master played the same opening move. See the Score Calculation section for more details on how the score is calculated.

In addition to the score, you are told the engine evaluation for your move, the master move, and the move the engine considered the best move if it is different to the master move or the move you played. The best engine move itself is not shown if it is different from the played or master moves , as often the best move can remain the best move for several moves in a row, and providing the actual engine move would provide unwarranted hints, hence only the evaluation is shown.

The score from 0-12 is fed into a rating calculation (using the glicko rating system) which takes into account the rating of the current position, and your current rating to calculate a rating adjustment. If you are playing a game in unrated mode, the rating adjustments will be zero, however the rating adjustment that would have been made in rated mode is shown in square brackets e.g. User: 1830.2 (0 [+4.1]) indicates your rating adjustment was 0, but if solving in rated mode it would have gone up 4.1 points.

Rating each position in a game provides the advantage of being able to produce a user rating based not only on the matchup rate of your moves, but also factors in how easy it was to find those moves. A system which only scores based on comparison to the master or best engine move ends up giving the same result for playing the master move in a simple obvious retake as it would in a move requiring long calculation, or subtle positional evaluation.

A consequence of the rating system is that easy positions (compared to your current rating) will require a better move to gain rating points compared to a harder position. For example, playing a move that scores 5 on a very easy position that is well below your rating may lead to a rating point loss, but may be enough to gain rating points on a much harder problem. One way of interpreting the score and the impact on the rating outcome is that a 0 score is equivalent to a rating outcome that would arise if you lost a game, 10 leads to a rating adjustment equivalent to winning a game, and 5 is equivalent to drawing a game. A score of 10 will always gain at least some points, no matter how easy the position is. However, as mentioned in the previous example, whether you gain or lose points with scores below 10 will depend on how the position rating compares to your own rating.

The score given for a particular move depends on a number of factors:

The progressive scoring display to the left of the board shows a list of statisticss showing your progress on the current game, the following stats are followed:

Guess the move supports per move comments and tags. As you move around in the game, the current move and move number is shown in the position comment/tag section title, and any comment or tag made or viewed will be relevant to the current move. You can also make general comments/tags that are relevant to the entire game by jumping to the start of the game, before the first move is played, and the comment/tag title bar changes to "Entire Game", at which point comments/tags made or viewed are referring to the entire game rather than a particular move.

In addition to the per move position comment view, there is also an "All Comments" view which is a covenient way of reading over all comments made on a game. You can jump to any position in the "All Comments" view by clicking on the move header for the relevant move. The "All Comments" view also allows adding an additional comment to already commented moves.

The game selector is shown when the GTM page is loaded. It allows you to choose to start playing a new game, resume a game in progress, or replay an already played game. The replay button is shown next to games you’ve played before, and allows you to choose between replaying the game in unrated mode, or reviewing your most recent attempt on the game. Once you’ve selected a game for play or review, the game selector is hidden, and can only be accessed again by either reloading the page , or reaching the end of the game, and clicking on the 'Select Game' button shown under the board.

The game selector also shows some statistics related to the problem, such as the average move score you achieved last time you played the game (in brackets next to the ast date played column), and the current rating of the problem (which is an average of the rating of all of the positions in the game).

Review Mode allows you to look through the game, viewing the engine lines for your move, the master move and - where it differs from the master and played moves - the best engine move. As you move through the move list, you are also shown the scoring details for each move. Review mode is automatically entered after you reach the end of a game, or you can enter review mode by clicking 'replay' next to a previously played game, and choosing 'review' when asked to choose being replaying unrated, or reviewing the previous game.

In the review mode move list , a move you played is marked as "PLAYED", a move the master played is marked with "MASTER" and the best engine move is marked as "ENGINE". So for example a move that you played, that was also the master move AND the best engine move will be labeled "MASTER/PLAYED/ENGINE". For the moves that you played, they will also be coloured based on the score, good moves are green, bad moves red, and moves in between the two as khaki.

The variation folding and hiding buttons below the move list allows for some of the detail to be hidden in order to provide a more compact move list. The plus or minus expanders can be used to manually open and close the variation lines when folding is turned on. Please see the PGN Viewer Key Bindings section for more details on controlling the folding and traversing the move list via the keyboard.

When a game is first loaded, you can choose to skip playing the opening part of the game by clicking the 'Skip Opening' button shown under the board. This is only an option before you have played a move, and the button is hidden once the first move is played. The opening skip is based on the opening explorer stats, and jumps to a point where the total number of games in the position is relatively low.

When in review mode, the UCI engine can be started by clicking on the 'Start Engine' button shown under the move list. This allows you to use your own engine running on your local machine to further analyse any position in the game.

The PGN viewer supports some special markup that allows arrows, square highlighting , mini-board insertion and figurine notation to be rendered on the board or move list.

Variation folding allows side variations to be expanded and collapsed. Folding can make it easier to concentrate on the main line or important side lines when a game is annotated with many side variations. Folding is turned on in the annotation window, or can be turned on by using the 'f' or 'F' key (fold/unfold current variation or fold all/unfold all variation keys). Once on, you can click on the variation toggle button in front of each variation to open and close any particular variation (or use the 'f' key to toggle the variation fold when inside a variation).

The viewer annotation window allows the manipulation of comments, annotation symbols and side variations. It can be opened by using the 'a' or 'A' keys, or by clicking the annotation button beneath the board.

If you’d like to display games on your own website, you can include a few HTML snippets in your page to allow you to load the viewer on external web sites.