11 ms·

This makes little sense no? With 200 ELO diff Hans should beat Magnus ~1/4 games.

by avip 4y ago

icelancer 4y ago
Only by K factor; those calculations don't hold at the highest levels, the distribution is skewed. Magnus also rarely loses with the white pieces in classical; his last loss was in 2018 at Biel vs. a much better player than Hans.
- avip 4y ago
  You're right, it's ~1/25. So still should happen.
  - sharedfrog 4y ago
    For sure. On its own, it's no more than a raised eyebrow. The next few days will show if there's any solid evidence. Btw. the 24% chance of "winning" against a 200 Elo higher rated opponent refers to "winning a point" - it includes draws as well.
    - texasbigdata 4y ago
      Rookie question but don’t draws occur more than 20-25% of the time at this level? Or is that just a unique feature of chess versus say StarCraft. Restated, are sports specific ELO scores real? Alternatively phrased, does an X ELO score in one competitive endeavor represent the same relative strength as the same ELO in another subject?
      - sharedfrog 4y ago
        Draws are the most common outcome at the "supergrandmaster" level, yes, to the extent that calls for changes to tournament formats or the game itself are often called for. Chess960 is the most popular chess variant, in which the starting positions of the pieces are shuffled in some way, resulting in 960 possible starting positions. There's also been experiments done with no-castling chess or tweaks to how the pawns move, or the ability to capture your own pieces. I'm not sure about your Elo question but I think a 200 points difference should meant the same thing regardless of sport.
      - lesuorac 4y ago
        > Restated, are sports specific ELO scores real? Alternatively phrased, does an X ELO score in one competitive endeavor represent the same relative strength as the same ELO in another subject? Yes and No. Yes but in the long term. ELO is a self-fulfilling prophecy so if your ELO states you should win say 30% of the games vs a different ELO but you win 50% your ELO rises until you have an ELO that states you should win 50% of the games vs that ELO (at which point you lose the same amount of points in a loss as your wins so it stops rising. Previously you gained more points for a win than a loss so you gradually increased in rating). No because not every sport starts ELO at the same baseline and not every sport league (i.e. Chess website) starts the base ELO at the same (i.e. Lichess starts everybody with 1500, Chess.com starts everybody with 1000). It's also a relative metrics amoungst active players (who indirectly/directly play each other) so if a GM 500 years ago reached an ELO of 2100 that doesn't mean they're the same strength of a current player with an ELO of 2100.
        sharedfrog 4y ago
        > No because not every sport starts ELO at the same baseline [...] (i.e. Lichess starts everybody with 1500, Chess.com starts everybody with 1000). It's not just where they start - they actually use different rating systems, and neither of them is the actual Elo rating system. Chess.com uses Glicko and Lichess uses Glicko-2. This is also a big part of what makes comparing players from different player pools a bit meaningless. What I think OP is asking is "If you used the same rating system in two different sports, would a 200 point difference mean the same thing?", for which I think the answer is "yes".
        lesuorac 4y ago
        > What I think OP is asking is "If you used the same rating system in two different sports, would a 200 point difference mean the same thing?", for which I think the answer is "yes". Well, I agree with your points but I'm not sure that is what OP is trying to do.
- systemvoltage 4y ago
  Not correct, the current rating doesn't reflect growing player's strength. It takes time for them to accrue rating. Alireza used to be beat players 200 points higher as he was climbing through the ladders. Hans just topped 2700. There is also the reverse side of the coin that top players peak at a rating and then decline as they age. Not saying Magnus is, but it is not a possibility that can be ruled out.
  - peter422 4y ago
    Magnus has played tons and tons of chess games this year and has maintained his rating. There is no doubt his rating is very accurate unless you think his skill started to deteriorate in the last month.
    - systemvoltage 4y ago
      What if Hans is actually rated higher than his current rating, seems quite plausible, no? ELO has pros and cons; it is not some law of nature: https://en.chessbase.com/post/what-s-wrong-with-the-elo-system https://en.chessbase.com/post/what-s-wrong-with-the-elo-syst...
      - peter422 4y ago
        I think your point is fine for Hans, I’m just ruling out the possibility that Magnus is overrated. The reason a lot of older players are overrated is because they stop playing as often and their rating doesn’t catch up to their decline. Magnus plays a ton so that isn’t a concern. For elite younger players, no matter how much they play, they are getting better so quickly that the rating can’t catch up. Younger players also get better much much faster than older players decline. It might take a decade for a GM past their prime to drop 100 ELO, whereas a young player can easily add 100 ELO in a year on their way to super GM status.
  - squidbeak 4y ago
    Alireza was able to speak accurately and in depth in his post game interviews along the way. Niemann doesn't seem able to do this.
tzs 4y ago
That was the result of the four games they played at the FTX Crypto Cup in Miami about 3 weeks ago. Hans, playing back, crushed Magnus in their first game. Then afterwards when an interviewer tried to interview him about the game he just said "The chess speaks for itself" and walked away. A lot of commentators interpreted that as a bit of trash talk about Magnus, who convincingly won the remaining 3 games.
sdwr 4y ago
https://www.youtube.com/watch?v=DCeJrItfQqw&t=377s https://www.youtube.com/watch?v=DCeJrItfQqw&t=377s I don't see a cheater, I see someone who's been grinding for a while and finally popped off. Dunno if he can keep it up though, might be one of those miracle performances that he never recovers from (ex. Linsanity). He comes across as arrogant and a little affected (the hand flourishes, stylistic pauses), but I like him a lot better than Hikaru. Hoping he continues to see success and smooths out his interview performance over time. If anything, I'm suspicious it's was a throw from Magnus, or at least subconsciously choosing when to relinquish the throne. This is the guy you want to see in the spotlight. He's interesting, confident, and puts his heart into the game.
tzs 4y ago
Elo gives expected outcome in points, not wins. If the formula says ~1/4, it means you should get ~1/4 of the points, not ~1/4 of the wins. If Hans were to score 1 point off Magnus in a 4 game match it would be far more likely to be be by drawing twice than by winning once.

This makes little sense no? With 200 ELO diff Hans should beat Magnus ~1/4 games.

by avip 4y ago

This makes little sense no? With 200 ELO diff Hans should beat Magnus ~1/4 games.

Only by K factor; those calculations don't hold at the highest levels, the distribution is skewed. Magnus also rarely loses with the white pieces in classical; his last loss was in 2018 at Biel vs. a much better player than Hans.

avip 4y ago
You're right, it's ~1/25. So still should happen.
- sharedfrog 4y ago
  For sure. On its own, it's no more than a raised eyebrow. The next few days will show if there's any solid evidence. Btw. the 24% chance of "winning" against a 200 Elo higher rated opponent refers to "winning a point" - it includes draws as well.
  - texasbigdata 4y ago
    Rookie question but don’t draws occur more than 20-25% of the time at this level? Or is that just a unique feature of chess versus say StarCraft. Restated, are sports specific ELO scores real? Alternatively phrased, does an X ELO score in one competitive endeavor represent the same relative strength as the same ELO in another subject?
    - sharedfrog 4y ago
      Draws are the most common outcome at the "supergrandmaster" level, yes, to the extent that calls for changes to tournament formats or the game itself are often called for. Chess960 is the most popular chess variant, in which the starting positions of the pieces are shuffled in some way, resulting in 960 possible starting positions. There's also been experiments done with no-castling chess or tweaks to how the pawns move, or the ability to capture your own pieces. I'm not sure about your Elo question but I think a 200 points difference should meant the same thing regardless of sport.
    - lesuorac 4y ago
      > Restated, are sports specific ELO scores real? Alternatively phrased, does an X ELO score in one competitive endeavor represent the same relative strength as the same ELO in another subject? Yes and No. Yes but in the long term. ELO is a self-fulfilling prophecy so if your ELO states you should win say 30% of the games vs a different ELO but you win 50% your ELO rises until you have an ELO that states you should win 50% of the games vs that ELO (at which point you lose the same amount of points in a loss as your wins so it stops rising. Previously you gained more points for a win than a loss so you gradually increased in rating). No because not every sport starts ELO at the same baseline and not every sport league (i.e. Chess website) starts the base ELO at the same (i.e. Lichess starts everybody with 1500, Chess.com starts everybody with 1000). It's also a relative metrics amoungst active players (who indirectly/directly play each other) so if a GM 500 years ago reached an ELO of 2100 that doesn't mean they're the same strength of a current player with an ELO of 2100.
      - sharedfrog 4y ago
        > No because not every sport starts ELO at the same baseline [...] (i.e. Lichess starts everybody with 1500, Chess.com starts everybody with 1000). It's not just where they start - they actually use different rating systems, and neither of them is the actual Elo rating system. Chess.com uses Glicko and Lichess uses Glicko-2. This is also a big part of what makes comparing players from different player pools a bit meaningless. What I think OP is asking is "If you used the same rating system in two different sports, would a 200 point difference mean the same thing?", for which I think the answer is "yes".
        lesuorac 4y ago
        > What I think OP is asking is "If you used the same rating system in two different sports, would a 200 point difference mean the same thing?", for which I think the answer is "yes". Well, I agree with your points but I'm not sure that is what OP is trying to do.
systemvoltage 4y ago
Not correct, the current rating doesn't reflect growing player's strength. It takes time for them to accrue rating. Alireza used to be beat players 200 points higher as he was climbing through the ladders. Hans just topped 2700. There is also the reverse side of the coin that top players peak at a rating and then decline as they age. Not saying Magnus is, but it is not a possibility that can be ruled out.
- peter422 4y ago
  Magnus has played tons and tons of chess games this year and has maintained his rating. There is no doubt his rating is very accurate unless you think his skill started to deteriorate in the last month.
  - systemvoltage 4y ago
    What if Hans is actually rated higher than his current rating, seems quite plausible, no? ELO has pros and cons; it is not some law of nature: https://en.chessbase.com/post/what-s-wrong-with-the-elo-system https://en.chessbase.com/post/what-s-wrong-with-the-elo-syst...
    - peter422 4y ago
      I think your point is fine for Hans, I’m just ruling out the possibility that Magnus is overrated. The reason a lot of older players are overrated is because they stop playing as often and their rating doesn’t catch up to their decline. Magnus plays a ton so that isn’t a concern. For elite younger players, no matter how much they play, they are getting better so quickly that the rating can’t catch up. Younger players also get better much much faster than older players decline. It might take a decade for a GM past their prime to drop 100 ELO, whereas a young player can easily add 100 ELO in a year on their way to super GM status.
- squidbeak 4y ago
  Alireza was able to speak accurately and in depth in his post game interviews along the way. Niemann doesn't seem able to do this.

tzs 4y ago

That was the result of the four games they played at the FTX Crypto Cup in Miami about 3 weeks ago. Hans, playing back, crushed Magnus in their first game. Then afterwards when an interviewer tried to interview him about the game he just said "The chess speaks for itself" and walked away. A lot of commentators interpreted that as a bit of trash talk about Magnus, who convincingly won the remaining 3 games.

sdwr 4y ago

https://www.youtube.com/watch?v=DCeJrItfQqw&t=377s https://www.youtube.com/watch?v=DCeJrItfQqw&t=377s I don't see a cheater, I see someone who's been grinding for a while and finally popped off. Dunno if he can keep it up though, might be one of those miracle performances that he never recovers from (ex. Linsanity). He comes across as arrogant and a little affected (the hand flourishes, stylistic pauses), but I like him a lot better than Hikaru. Hoping he continues to see success and smooths out his interview performance over time. If anything, I'm suspicious it's was a throw from Magnus, or at least subconsciously choosing when to relinquish the throne. This is the guy you want to see in the spotlight. He's interesting, confident, and puts his heart into the game.

tzs 4y ago

Elo gives expected outcome in points, not wins. If the formula says ~1/4, it means you should get ~1/4 of the points, not ~1/4 of the wins. If Hans were to score 1 point off Magnus in a 4 game match it would be far more likely to be be by drawing twice than by winning once.