Wesley So's chess, by the numbers
October 16, 2026 · 251 games: the last 200 blitz on chess.com, March to October 2026, and 51 classical, 2024 to 2026
So's last 200 blitz games on chess.com and their classical games of the last two years, every move checked by Stockfish, counted the way anyone's games are. Here is what the count found.
So won the 2026 Sinquefield Cup in August, the second year running, by drawing the Armageddon game against Praggnanandhaa after nine classical rounds and two rapid games had left them level. Of the round-2 win over Jeffery Xiong at the 2026 U.S. Championship: "I was just trying to survive the entire game, then I had no idea what was going on," So said (chess.com's round-2 report, by Anthony Levin, October 11, 2026). Two months on, a three-time U.S. champion, they share the lead of the 2026 U.S. Championship after six rounds.
This page counts how they play: their last 200 blitz games on chess.com and their classical games from the Sinquefield Cups of 2025 and 2026, Norway Chess, Romania and two Olympiads, every move checked by the engine, with the same questions a club player's games get asked.
How they play
Every player gets one of four profiles from two measures: how often they force matters with a check or a capture, against what the position asks for, and how rarely one of their pieces is left loose. So's pieces are rarely loose (86 of 100 in blitz, 100 of 100 in classical; the line is at 64), and on forcing they sit a shade under the line at 49 in both, the same spot as Carlsen and Caruana: the Steady Hand. The other blitz scores: closing 88, accuracy 83, recovery 76; in classical, accuracy 95.
1.Nf3 first, ahead of 1.d4 and 1.e4.
41% of 104 blitz games as White opened with Nf3
As White in blitz, So opened with the knight, 1.Nf3, 43 times (41%); 1.d4 32 times (31%), 1.e4 20 (19%), then c4 four times, b3 three and g3 twice. Of the five players counted this week they are the only one whose first move is not a pawn: Caruana and Liang open 1.e4, Carlsen and Mishra are split between e4 and d4. The games most often go into the English Opening (23 of 200, scoring 78%) and the Réti (10, 90%).
Checkmate in one, passed up 17 times.
6 of 23 mates in one on the board were played; the other 17 times they played on
Checkmate in one was on the board 23 times in So's 200 blitz games. They played it 6 times. The other 17 times they played another move, in 7 games, and won all 7. Ten of the 17 came in one game, against Suyarov1 on March 27, 2026, an ending of queen and pawns against a king and pawns in which, from move 54 to move 73, a mate in one was on the board on ten of their turns; they took two pawns, promoted two more, and mated on move 73. The engine's bill for all 17: under 1% of the winning chances they gave away all year.
Once clearly winning, four games got away.
88% of decisive advantages converted in blitz
So reached a decisive advantage in 145 blitz games and won 128 of them (88%); 13 were drawn and 4 lost, the fewest losses from a winning position of the five players counted this week (Carlsen lost 8 of 161, Caruana 7 of 142). Lichess blitz players rated 2000 and up convert 79% of the games they are clearly winning.
Half the pieces taken next move were given up on purpose.
49% of pieces taken on the very next move were trades or sacrifices the engine approved, not slips
Of the 43 times one of So's pieces was taken on the very next move, 21 were a trade or a sacrifice in a live position with the engine's approval (49%); for most players it is about 9%. Carlsen's share is 51%, Liang's 48%, Sindarov's 46%. The piece most often given up: the bishop, 8 times; the queen 5, the knight and the rook 4 each.
20 of 32 losses had no blunder in them.
20 of 32 blitz losses had no blunder in them
Of their 32 blitz losses, 20 have no single move that cost 30 or more points of winning chances: outplayed a little at a time, 62% of the time. Carlsen's share is 15 of 28, Caruana's 21 of 37. Where a So game does turn, the report finds the slips come when their king is out in the open (1.8x their usual rate) and on moves answered with a check (2.4x).
Five answers to 1.e4.
5 different replies to 1.e4 in 55 blitz games as Black, none played more than 27% of the time
Against 1.e4, So answered 1…e5 15 times, 1…e6 14, 1…c5 13, 1…d6 and 1…c6 six times each, and 1…g6 once: five openings in regular use where most players have one or two. Against 1.d4 it is two: 1…Nf6 14 times and 1…d5 13. Carlsen, for comparison, meets 1.e4 with 1…c5 more than half the time (28 of 53); Liang plays 1…e5 in 45 of 48.
Two losses in 51 classical games.
2 of 51 classical games lost, 2024 to 2026; 34 drawn, 15 won
In 51 classical games from the Sinquefield Cups of 2025 and 2026, Romania, Norway Chess and two Olympiads, So lost twice (to Praggnanandhaa at Norway Chess 2026 and to Firouzja at the 2026 Olympiad), drew 34 and won 15; both losses had no blunder in them. The engine checked 218 moments where one of their pieces was left loose and found no real slip among them; all 11 pieces taken on the very next move were trades or sacrifices it approved. Average loss a move against its choice: 0.08 of a pawn. As White 1.e4 in 15 of 24 games; as Black the Berlin in 9 of 27.
The full reports
- So's blitz report
200 blitz games (mostly 5+0 and 3+0) on chess.com as GMWSO, March to October 2026: 134 wins, 34 draws, 32 losses. Download the games (PGN) - So's classical report
51 classical games, 2024 to 2026: the Sinquefield Cup 2025 and 2026 (9 each), the Superbet Classic Romania 2026 (9), Norway Chess 2026 (6), the Olympiads of 2024 (10) and 2026 (9), from Lichess's broadcasts. 15 wins, 34 draws, 2 losses. Download the games (PGN)
Each report opens as a deck of slides, one finding to a slide, with the positions behind every number. A blunder is a move that cost 30 or more points of winning chances; winning chances are the engine's estimate of how likely a player is to win, from 0 to 100.
How this was measured
Plysight analyzes games the same way for everyone: Stockfish evaluates every move, and the findings are counted from those evaluations. The reports compare a player with others at their rating where a fair comparison exists; at So's rating there is none, so those lines are left out. Advice lines, written for a player reading their own report, are left out too. The six house words the reports use are defined on each report's last slide.
Blitz games from chess.com's public archive of gmwso; classical games from Lichess's broadcasts. Plysight is independent of chess.com, Lichess and So. So's FIDE profile. Published October 16, 2026; updated October 16, 2026.
Piece art by Cburnett, CC BY-SA 3.0.