What is the required Elo to beat a Grandmaster with queen odds?

I heard someone saying today that he would beat Garry Kasparov if Garry played without a queen.

Can a Grandmaster give queen odds and win against a 2000 ranked player? Rook odds?

• I think you will find this link useful: en.wikipedia.org/wiki/Chess_handicap Aug 25, 2013 at 21:35
• Methinks the material is a bit underestimated here. (Even a 1900 should bring home a knight against Alpha Zero, unless blundering horribly.) Feb 13, 2023 at 20:20
• @HaukeReddmann AlphaZero would easily beat a 1900 down a knight Jul 25, 2023 at 18:12
• @AspectOfTheNoob: [shrug] I know of no actual experiment to settle the score. A GM won 5:1 against Komodo: chess-international.com/?p=21900 Jul 25, 2023 at 19:48

It's hard to say for sure, but even a grandmaster wouldn't have a very good chance without a queen unless the opposing player made big blunders.

GM Larry Kaufman estimates that Kasparov would "have chances" with giving knight odds to a player rated 2115, and be slightly favored in giving pawn and move odds to a GM rated around 2500.

A knight is about 3 pawns, and a queen is 8.8-9 pawns. So if at levels at 2500 vs 2850, a pawn is roughly a 300 point advantage (according to Kaufman, a pawn would presumably be worth maybe 200 at the 2000-2200 level vs Kasparov), and a knight is around a 700 point advantage at 2115 vs 2850, then a queen, being worth almost 3 times as much as a knight, would be worth much more.

[T]he Elo equivalent of a given handicap degrades as you go down the scale. A knight seems to be worth around a thousand points when the "weak" player is around IM level, but it drops as you go down. For example, I'm about 2400 and I've played tons of knight odds games with students, and I would put the break-even point (for untimed but reasonably quick games) with me at around 1800, so maybe a 600 value at this level. An 1800 can probably give knight odds to a 1400, a 1400 to an 1100, an 1100 to a 900, etc. This is pretty obviously the way it must work, because the weaker the players are, the more likely the weaker one is to blunder a piece or more. When you get down to the level of the average 8 year old player, knight odds is just a slight edge, maybe 50 points or so.

http://en.wikipedia.org/wiki/Chess_handicap

• I've been told that a 1500 player would be favorite (but not a "lock) against a grandmaster who gave queen odds. The grandmaster would be a favorite at rook odds. Sep 26, 2016 at 16:50
• I've found queen odds is the threshold where you really can't play "normal" chess. Even down a rook, you can maneuver, build up an attack, etc. Down a full queen, all trades become so favorable for your opponent you mostly just have to mill around and hope they blunder.
– Nate
Mar 17, 2017 at 20:52

I have a hard time believing that there's a player on the planet who could beat me with Queen odds. The strategy would be simple: trade off pieces pieces when possible. I'm a USCF B player.

Here are some 'Queen Odds' games. Play by the weak side is sometimes very dicey. Paul Morphy loses his game. None of the games the GMs won went past 15 moves. After that the Queen asserts too much power to overcome.

Here's a rant, lol.

I have played against computer engines with material odds for myself.

At queen odds I win easily.

Minor piece odds gave me a lot of good practice. I was losing them to start off but managed to improve my play and start winning them.

I play a relatively closed game in such circumstances. and ensure my kingside cannot come under a major attack.

Exchange odds would be an interesting one to try out.

• Not a great example since the computer is just going for optimal play against a "perfect" opponent, while a human player against a weaker opponent with odds would try to play tricky moves that keep the game complicated, whether those moves are objectively the best or not. Feb 14, 2023 at 23:38

I'm a senior master (2400+), and I've played thousands of odds games with players up to expert level (2000+). In the 90s I was able to give a local 2100 the odds of piece-for-pawn (White removes his queen's knight, black removes his f-pawn), and I would win around 80%. In those days I would give 1800-2000 players anywhere from a knight to a rook, with a positive win-loss ratio. The internet has changed things, though, and I think modern class players are much more savvy and accurate in blitz, so I can't say I can repeat those numbers today. But I would say that Queen odds is more of a chess joke, where there really isn't much to do unless the opponent is a real tyro. The clock can work both ways, because many weaker players won't necessarily come up with a better move given more time, whereas I can usually forestall doom if there's plenty of time to think. Beer helps. On both sides.

I was able to beat Chess.com 3200-rated bot starting it with a rook and a knight down. And I'm rated 850. I guess, all you need to ensure is to trade pieces as much as possible ensuring you don't blunder anything.