https://novachess.ai/articles/chess_tactical_analysis.html

Here's a bit on the methodology:

For all the games, each position (for each color) from moves 12-25 was considered. The metrics used were:

– Total point value of pieces that can be captured on any turn, showing how many threats/tactical opportunities exist

– How many legal moves each side has on their turn (excluding positions when a player is in check), as piece mobility tends to be higher in tactical positions

– How much material was captured by move 25, as tactical games tend to have more captures (as a general rule)

I think it's worth noting that an individual game could be considered tactical or positional while not aligning with the expected score, but I think over the sample size used it should be a pretty good indication.

Posted by novachess-guy

4 comments

  1. Great stuff! It would be interesting to see the names of world champions on x axis.

  2. The colors on the graph are difficult to tell apart, and also they are in the wrong order on the legend (not really a rule but since your regions are sorted why not also the legend)

  3. Could you share how this score, compare for an average game with positional and tactical openings? To have a feeling of what those number mean. Score variation is very small.

  4. I wonder if the dip in the 1960s correlates to the Soviet school of Spassky and Petrosian.

Comments are closed.