# What is the typical move prediction accuracy of human chess grandmasters?

I couldn't find any data about the percentage of moves that a typical human grandmaster would be able to predict in games of other human grandmasters. To clarify my definition of accuracy: The predicting player would predict the next half-move for a position from a game he or she has never seen before, only being able to see past positions from the same game (preferably without even seeing the names of the players or when the game was played). To limit the influence of blunders, only games with classical time control should be used and the predicting player would have a limited time per move to make the prediction (let's say 1 minute per move). A correctly predicted move counts as 100%, an incorrectly predicted move as 0% and these numbers would be averaged over a large number of positions.

Since I couldn't find anything, I tried it myself and achieved 42 out of 119 (35.3%) in a randomly chosen grandmaster game (Alekhine vs. Erik Lundin, 1941, Round 7), but I'm quite amateurish, not a grandmaster. Randomly guessing legal moves would have achieved less than 5% in that specific game. My conditions do not have to be followed strictly, but I'm not really interested in accuracy results for blitz games, games of weaker players or "immediate", intuition-based predictions (i.e. without giving the predicting player time to think).

I'm interested in this because I want to know how accuracy varies with playing strength of the predicting player.

• I would expect this value to fluctuate wildly with whatever game you have. In a short game that follows a known forced theoretical line the prediction accuracy might be high. On the other hand, in a long game it would likely be moch lower. Also I can think of many situations where there are several equivalent moves. A GM's guess is then as good as anybody's. My suspicion is that such cases outnumber the few cases where the GM's strength does show through. May 28, 2020 at 7:52
• Are the names of the players visible to the guessing person? May 28, 2020 at 7:53
• I updated the question. Seeing the names might give you too much of a hint, so I'd prefer predictions where you have essentially no information about the game other than the moves played so far. And about the fluctuating value: That's true, but that's why I said the accuracy should be averaged over a large number of positions (from many different games if possible). 100% is of course impossible to achieve if the number of positions is large enough. May 28, 2020 at 13:52
• OK, but I still believe that there is no such thing as a typical move prediction accuracy in general as it would depend on the games you base it on (even with averaging). Instead the following questions could make more sense: 1. What is the move prediction accuracy of a GM for [specified set of games]? or 2. How much better is the move prediction accuracy of a GM relative to an amateur if given the same set of games. The latter question might not have a universal answer and might also depend on the set of games. May 28, 2020 at 14:03
• My concern is not that it is different for different GM, but different depending on the chosen set of problems. For instance, GMs would have a hard time to guess moves in patzer games or in endgames with many equivalent moves. May 29, 2020 at 6:12

0.2785

It depends a lot on the opening as each would have different preferences and pet lines.

Middle games would be a bit better as there are fewer strategies to choose from for long term effort.

End games should be best as the simple positions are easier to predict moves although many end games could offer more variation to try for a win than middle games.

Overall the answer is 0.2785 based on a sample of one. More samples should yield a more accurate statistic.

Although I doubt anyone really cares enough to do a study to see.

• 27.85% for a grandmaster? Could you give us the name of that player and the game that was used for the predictions? Jun 24, 2020 at 2:03