Those of you who play on Chess(dot)com are aware of the post-game analysis tool, which, uh, analyzes your game. The first few moves are deemed “book moves” (as in, “Nc3 is the last book move”). From the context, I’m guessing it means “moves in known lines of known openings that have been consistently played for the last 300 years.” Or something . But does “book” here refer to a metaphorical “book,” or perhaps the website’s own openings database? Or is it referring to a known chess reference, such as Modern Chess Openings?
You have it basically right. Chess.com is referring to their own “book” of reference opening moves. The term has existed for a long time, coming from the idea that a move could be found in an opening book like MCO.
Is that “book” database static or dynamic? If two players played exactly the same game twice, would the second one be labeled as all “book moves”, since they’d been played before? Or is the “book” only for sequences that have shown up some threshold number of times before?
They’re moves found in standard chess references, like MCO etc
Yes, ‘book’ used to mean found in a standard openings reference book (like Modern Chess Openings.)
Now that computers store vast numbers of games and it’s easy to reference them, ‘book’ means a move that has been played several times before (usually by titled players.)
Here is their own utterly unhelpful definition.
Opening theory refers to known opening moves that have been published in books, databases, magazines, etc. If two players play a game whose first 10 moves have been played before, then those first 10 moves are considered book moves or theory.
Is it that bad of a definition? Seems fine for a layman’s two-sentence definition.
To the OP: You might notice that when you use the “Review” feature, assuming you’re on a browser (not the app), there is a tab at the right side of the screen labeled “Openings”. That pulls up the opening database. It’s populated by master-level games (usually down to any master title, which is a very large spread in ratings, but still all highly skilled players) taken both from over-the-board (IRL) games and online games. Other playing sites will have a different corpus of online games in their database and may have differences in the OTB games included, but not at a level that you will care about. (Some services strive for meticulous completeness.)
Anyway, in the openings tab, it shows for the current position a list of various moves that appear in the database and the number of times the resulting position has been reached (possibly by transposition). For each move option, it shows the fraction of games with white win / draw / black win; you can click on different moves to explore the tree; and you can click on the game count next to a move to pull up a list of the actual games reaching that position, sorted by average player rating (useful to see how masters follow-up on the opening ideas in that line).
If you follow your game’s moves up until the last “book move”, you’ll see in this openings tab the games in the database. If your next move was not a “book move”, it might still appear in the database, but it’s just not a move or a game that the site counts as official “book”, even though it’s in the database. This can be because it was just in some online blitz games, or it’s too rare a move, or it’s too new. For the “too new” case, I thought to check a couple games from the 2018 World Chess Championship (Carlsen vs. Caruana). These are five years old now. There are many early moves that everyone would absolutely consider solid chess theory now – they were played a little bit previously, they were selected and played by the best two players in the world at the time, they have been played at the top levels MANY times since given the exposure in 2018, and they certainly appear in real-life books now – but they aren’t tagged as “book” in the chess(dot)com database.
Obviously this level of detail is irrelevant for patzers like us, but it might be interesting all the same.