What if both players started with the same number of pieces as normal, but the novice could replace a piece that was captured after a certain number of turns. So if the novice loses their rook, after X turns they can put that rook back in its starting position. The more novice the player, the fewer turns they have to wait to replace the piece. A complete novice may have to wait only a few turns, while a more intermediate player would have to skip 5 or 10 turns. This would allow the novice to remain strong throughout the game, while the grandmaster would have plan for the replacement pieces coming into the game at a certain spot X moves in the future.
I guess my question is “how so”? IOW what makes a great chess player “great”? Is it an ability to see 20 moves ahead? Is it having encyclopedic knowledge of various openings and gambits and whatnot?
In the case of my father, he frankly admitted he could only see 5-6 moves ahead (I struggled to see three moves). What made Dad “very good” was that he knew every opening and how to play them. It didn’t require seeing that many moves ahead, he just played the book. What kept him from being an “excellent” player (again, by his own admission) was that when his opponent went against the book, Dad might stumble for a few moves trying to come up with the right counteroffense. And when you have two players who are pretty evenly matched, when you give one of them two or three moves in a row, that can be almost insurmountable.
How about he grandmaster requires two pieces to apply check. (The novice isn’t in check until the grandmaster has at least two pieces attacking the king.)
And we’re now being humbled to realize that no human really understands chess.
I’ve seen a variant where one player moves twice per turn, but has absolutely nothing but the king and the six centermost pawns, while the other side plays as normal. It’s a ludicrously huge advantage for the double-moving player, enough so that I’m pretty sure that I could win even against perfect play. For one thing, any pawn promotion is instant victory, because a double-moving queen almost anywhere on the board can threaten almost every square. For another, a double-moving king is a real powerhouse of a piece.
I’m not certain, but I think even a lone double-moving king, without any support at all from any pawns or other pieces, might already be enough to force a win (though that would take a bit more skill). Though the other side might be able to pull it off, if they can mobilize their queen and both rooks quickly enough (without losing any of them).
Doesn’t the other side have no pawns? That’s what I recall from years ago.
I recall that sometime in secondary school, probably middle school, I was in a chess “class” taught by a high schooler who would occasionally play in tournaments. I wasn’t a complete novice, but I had very little hope of beating him even I was playing white and he had no queen. I had been playing chess for quite a while and had done some studying, and thus knew most of the openings and all of the theory, but I didn’t have very good perception of how things would develop beyond a couple of moves. Thus it wasn’t too hard for him to get good forks or x-ray attacks or other sorts of free pieces constantly, because I never saw what he was doing to develop those attacks. That will get the queen advantage back real quick even if you’re playing correctly from a wider theory view. Thus, I suspect that if you make the gulf in skill even greater, you’d need some really massive material advantage. I’d probably consider starting at both rooks and the queen, but I’d suspect we’d also have to remove one or both knights to get anywhere close.
I play on chess.com where I’m ~1300 on 10 minute controls. Occasionally when an opponent resigns I’ll play the remainder against the computer. In the end game I’ll probably need a 5 point advantage and in the middle game probably a full queen or more. I expect from the start I would need 20 points. I’m better and a complete amateur and stockfish is better than Magnus Carlson, but maybe the delta is similar?
I think this would be difficult for the grandmaster to win - it’s very rare in chess for the king to be in check by 2 pieces simultaneously, as it involves the piece directly giving check moving out of the way of another piece behind that then also gives (discovered) check (there might be even rarer possibilities involving an en passant capture). King and queen vs king would no longer be a win and I don’t think even king, queen and rook could force a win in those conditions.
Having said all that - I now recall seeing several Ben Finegold videos in which he defeats a much lower-rated opponent by taking all their material, then promoting several of his pawns (once he managed to get 9 queens, and on several occasions he has promoted enough pawns to restore his entire original army). So on that basis, yes it could be done. Still quite a challenge for the GM to figure out how to force checkmate with a double check, though. I’d be interested to see it done.
This is interesting. It feels to me that at the endgame a 3+ advantage should be enough if I have time on the clock, but it depends on how complicated the position is. I would definitely draw a lot of games that I should win. I’m 1800 Rapid on Lichess (which typically has higher ratings than chess.com I think, so probably pretty similar to you as a player).
But 20+ feels like too much at the start. I just played against Stockfish 14 lvl 8 on Lichess up a queen from the start and won twice very easily (it puts that as +12). Maybe I didn’t have the engine cranked up all the way (not sure how to make it any higher, but I noticed that the post-game analysis marked a few “inaccuracies” in the engine which probably shouldn’t happen if it was playing perfectly) but just trading down aggressively end to a won end game isn’t that difficult. Maybe I will play a few up only a rook later, I feel like that would be much much closer, and perhaps even favoring the engine.
It’s also probably worth noting that the engine plays it straight, allowing trades and not really going for the kind of tricks that a real GM would probably do. But I dunno, @glee, do you think you would have any chance down a full queen against a “club level” player?
Here’s one of the games:
- e4 Nc6 2. d4 e6 3. Bb5 a6 4. Bxc6 bxc6 5. Nc3 a5 6. Nf3 Nf6 7. Bg5 Ba6 8. Bxf6 gxf6 9. a3 h5 10. Qd2 Bh6 11. Qd1 O-O-O 12. b4 axb4 13. axb4 Kb7 14. Rxa6 Kxa6 15. Qd3+ Kb7 16. O-O Bf8 17. Ra1 Ra8 18. Rxa8 Kxa8 19. Qa6+ Kb8 20. Qa5 Bd6 21. e5 fxe5 22. dxe5 Be7 23. b5 Bc5 24. bxc6 Bb6 25. Qb5 dxc6 26. Qxc6 h4 27. h3 Rd8 28. Qb5 Kb7 29. Na4 Rd5 30. Qxd5+ exd5 31. Nxb6 Kxb6 32. Nd4 c5 33. Nf5 c4 34. Nxh4 Kc5 35. Nf3 Kb4 36. g4 d4 37. Nxd4 Kc5 38. Ne2 Kb4 39. g5 Kc5 40. g6 fxg6 41. e6 Kd6 42. Nf4 g5 43. Ne2 Kxe6 44. c3 Ke5 45. Nd4 Ke4 46. Kg2 Kd3 47. f4 gxf4 48. Kf3 Kxc3 49. Ne2+ Kd2 50. h4 c3 51. Nxc3 Kxc3 52. Kxf4 Kd4 53. Kf5 Kc5 54. h5 Kb4 55. h6 Kc3 56. h7 Kc4 57. h8=Q Kb3 58. Qd4 Ka2 59. Ke4 Ka3 60. Qb6 Ka4 61. Kd4 Ka3 62. Kc3 Ka2 63. Qb2# 1-0
Just saw this on preview, but I’m way lower level player than that and found Queen odds very easy against Stockfish level 8. The jump from Rook to Queen odds is from +8.8 to +12.8 according to Stockfish, just for reference.
Yeah, I was surprised as well that it took him that long to finally win, but it seemed he had positions he should have one before he got to queen odds. Here’s the video, if interested:
He only got a draw with 8 points of material off. The queen game looked pretty easy, though. I would think that once you’re in the 1500s range, queen odds should be enough to beat almost anyone.
Oh, he’s playing blitz chess.
I definitely could lose even with Queen odds playing blitz. I’d just run out of time. Or lose to some simple tactics like he did in a few of the Rook + Pawn games. He would never get his Queen skewered like he did in the Rook + 2 Pawns games in a slower time format.
And in the Rook + Knight game he up a whole Rook but blundered badly to lose a piece and then lost on time from a won position. I don’t really count that.
His Queen game looked a lot like mine did - trade everything and win. I even game up the queen for a Rook in one of my games just to get a winning pawn ending.
Oh, good God, how did I forget to mention that it was bullet chess? Yeah, when I was recalling the game, I forgot that it was one minute time limits. A very important point. So that 1000 Elo advantage is probably a lot more in less time constrained formats. Probably more like the 2000 mentioned before. (And, yes, those couple blunders you noted are ones that I alluded to where I thought he could have won with less odds.)
I do wonder what it would be like against full-strength (3500 ELO-ish) Stockfish.
When I play with my nine year old grandson he uses the treacherous King advantage. He says in the real world if a King were threatened he would simply change places with any of his own men so the King can’t be hemmed in by his own pieces.
A lot of these are not real chess. A way to play would be:
Play the game as normal. The only rule is that the grandmaster cannot checkmate the opponent.
After X number of moves they switch places.
I’m wondering how many moves it would take before there is a big enough hole that the grandmaster could not get out of.
That’d lead to both players trying to get themself into the absolute worst position possible for the first X moves. The grandmaster is probably better at that, too, and certainly better at recovering from the bad position their opponent came up with. I’m not sure how this would make things any easier for the opponent.
Yeah, you’re right. I just went and did this with queen odds and it was trivial on chess.com with full stockfish. I’m curious if my recollection was worse previously or if those positions were just harder for some reason.