Heresy on the chessboard

Chess960 (also known as Fischer Random or Freestyle Chess) is often presented as a bracing corrective to the alleged decadence of modern chess. It randomises the back-rank pieces into one of 960 possible starting positions and its advocates insist that by abolishing opening memorisation it restores honesty to the game, compelling players to think for themselves from the very first move. There is some truth in this claim. Yet it is equally true that the question of fairness, which is usually waved aside with a rhetorical flourish, is far from settled. Recent studies suggest that certain starting positions may confer a measurable advantage on White, thereby undermining the notion that randomness is the same thing as justice.

Is Freestyle chess a tougher game, since it removes opening theory? In standard chess, elite players memorize vast amounts of opening theory, often playing the first dozen moves automatically. Chess960 makes pre-game memorization impossible and, advocates claim, emphasizing a purer test of skill, creativity, and understanding of general chess principles. Players, including former world champion Magnus Carlsen, are forced to calculate and adapt from the very beginning of the game, rather than relying on prepared lines.

By promoting unfamiliar positions, Chess960 leads to more blunders and less frequent draws. This is seen as a positive for professional chess, where “draw death” at the top levels is a concern.

All this sounds admirable, and indeed it appeals strongly to an age that mistrusts tradition and worships novelty. But it is one thing to make a game harder, and another to make it fairer. While the intent of Chess960 was to level the playing field by removing opening theory, recent analysis suggests the fairness is not absolute.

A study by physicist Marc Barthelemy, highlighted in The Times in a recent report on January 5, used the powerful Stockfish engine. He found that the traditional White advantage (White moves first) remains in almost all 960 positions, but the size of that advantage varies significantly. In the following, I have drawn on an insight to his paper by Paul Arnold.

Some starting positions amplify White’s edge, while a few might slightly favour Black. The study indicated a wide range of positional balance across the 960 possibilities.

The analysis suggested that the most balanced position is #198 (QNBRKBNR), while position #80 (BBNNRKRQ) offers the largest advantage for White. The standard classical chess starting position is actually more asymmetrical and less balanced than the average Chess960 starting position.

Supporters of the variant respond that these imbalances are determined by chance before the game begins, rather than by months of selective opening preparation. This, they say, makes Chess960 a superior measure of raw chess ability. The International Chess Federation (FIDE) has accepted this argument in practice, officially recognizing the variant and hosting World Championships for it.

Yet one should beware of mistaking official recognition for intellectual soundness. I find it sickening to see the heretical proponents of so-called Freestyle Chess mafficking about the 100% (9/9) score by Magnus Carlsen in the recently concluded Grenke Open. Fanatical adherents of the new heresy are ululating with gay abandon and celebrating the fact that Magnus has outperformed, in a ratings sense, the superhuman rating performances of the chess Immortals, such as Bobby Fischer and Alexander Alekhine.

Such misplaced jubilation overlooks the elementary fact that Freestyle ratings cannot legitimately be compared with established chess ratings. Freestyle chess consists almost entirely of tactics, since the random disposition of the pieces at the start of the game renders such strategic concepts as pawn structure virtually irrelevant. To pretend otherwise is to engage in a form of intellectual doublethink. One might as well compare rating performances in chess with exploits in Tiddlywinks, Marbles, Shatranj (the ancient ancestor of chess), Shogi (Japanese chess), Xiangqi (Chinese chess), or Weichi, aka Go, the venerable oriental board game, where territory counts rather than checkmate.

A further opinion was shared with me by my colleague Adam Black. He thinks that the rationale the new freestyle game proceeds from, is fallacious. “(The shuffle of pieces) does not remove theoretical learning from playing the game, in fact, it makes the mountain of learning (eventually) 960 times greater, because a database of learning will spring up, and those that operate more from memory than intellect, will enjoy an even greater advantage than is currently the case. The counter-proposal would be to look at the new strategic approaches that sophisticated and novel analytic processes already bring to the traditional game. AlphaZero immediately comes to mind. How different might the outcomes of Barthelemy’s research have been had he used AlphaZero, rather than Stockfish to analyse the 960 openings’ status?”

The error is not in playing Freestyle Chess, which may be an entertaining and even instructive diversion, but in claiming for it a moral and historical superiority it does not possess. To confuse difficulty with depth, and novelty with truth, is a mistake that chess, of all games, should be careful not to make.

Levon Aronian vs. Magnus Carlsen

Freestyle Grand Slam Finals (unorthodox), Grootbos, 2025, round 3

d4 f5 2. f3 e5 3. dxe5 Nbc6 4. Ndc3 Ne6 5. e4

Here there is trade-off between deterring the e6-knight’s mobility, and releasing the c1-bishop. White wasn’t happy about …Nc5 putting pressure on e4, so he pre-empts this. But he should prefer, 5. f4 b6 6. Bc4 Bb7 7. Nd2 g5 8. Nf3 gxf4 9. exf4 Nb4 10. Bb3 Bc5 11. Qf1, when it is White that has a marginal edge in the position.

5… fxe4 6. Nxe4

This is consistent with the plan behind his previous move, but the plan is sub-optimal and gifts Black a very healthy advantage. This could have been constrained more effectively with,  6. Be3 Bb4 7. a3 Ba5 8. b4 Bb6 9. Nd2 Qf7 10. Ndxe4 O-O 11. f3 Nxe5 12. O-O-O.

6… b6

This rather dogmatic continuation aiming to castle long, returns all Black’s gains, leaving the position a level one. Instead, after 6… Ned4 7. Kd1 d5 8. exd6 Bg4+ 9. f3 Nxf3 10. gxf3 Bxf3+, Black retains a very comfortable advantage.

7. Nbc3 Bb7 8. f4 O-O-O

Premature. Black should prepare this move with 8… Nc5 9. Nxc5 Bxc5 10. Be3 Bb4 11. Qf2, and only now, 11… O-O-O.

9. Be3 d5

A case of stopping digging when finding oneself in a hole. If White can get his king to safety, he is slightly favourite in mutual attacks on opposing wings. After, 9… Qf7 10. Qf2 Bb4 11. Be2 Rhf8 12. O-O Ne7, Black can enjoy a pull with either 12… Ng3 or …Bc4.

10. exd6 Bxd6 11. Qf2 Qf7??

A blunder. Black should try, 11… Bb4 12. Bd3 Ned4 13. a3 Be7 14. O-O-O Nf5, which at least leaves Black no worse. However, the move played presents White with a considerable advantage, and despite a very few minor deviations, He should land this catch. And demurring  further errors, Aronian brings home the bacon960, with splendid technique.

12. Bc4 Qf5 13. O-O-O Kb8

Probably 13… Nc5 or …Na5 are better.

14. Rd5 Qf8 15. Rd6 Rxd6 16. Nxd6 Qxd6 17. Rd1 Qe7 18. Re1

While there are a few more ups and downs as the game progresses, Black has a sizeable enough majority, and a technique sufficient to bring home the Carlsen in the remainder.

And the rest is a matter of applying that technique. White is now dominant.

18… Na5 19. Bd3 Qd7 20. f5 Nd8 21. Bf4 Ndc6 22. g4 h6 23. a3 Nb3 24. Kb1 Nc5 25. Bc4 Rd8 26. b4 Qd4 27. Qxd4 Rxd4 28. Re8+ Bc8 29. Bxc7 Kxc7 30. Nb5 Kd7 31. Nxd4 Kxe8 32. Nxc6 Ne4 33. Kb2 a6 34. Be2 Kd7 35. Ne5+ Kd6 36. Nc4 Kc7 37. h4 h5 38. g5 Bxf5 39. Bxh5 Be6 40. Be2 Ng3 41. Bd3 Nh5 42. Ne3 a5 43. Kc3 axb4 44. axb4 Nf4 45. Kd4

When White’s position is both dominating and supreme.

Black resigns 1-0

Ray’s 206th book, “Chess in the Year of the King”, written in collaboration with Adam Black, and his 207th, “Napoleon and Goethe: The Touchstone of Genius” (which discusses their relationship with chess) can be ordered from both Amazon and Blackwell’s.

His 208th, the world record for chess books, written jointly with the late chess playing artist, Barry Martin,  Chess through the Looking Glass ,  is now also available from Amazon. 

A Message from TheArticle

We are the only publication that’s committed to covering every angle. We have an important contribution to make, one that’s needed now more than ever, and we need your help to continue publishing throughout these hard economic times. So please, make a donation.