5 ms·

I’ll bet this does not replicate, the effect of stereotype threat is notoriously overvalued. Don’t fall for junk science.

Invictus0 4y ago
Women that dedicate their careers to chess still dramatically underperform men at the professional level. There is even a different rating level and title system for women: women earning the "women grandmaster" title need 200 fewer ELO points than the grandmaster title for men. No woman has ever won the World Chess Championship, and there are far fewer female GMs than male GMs.
- dudeinjapan 4y ago
  How do women and men do versus AI opponents, i.e. without any gender? Wonder if chess.com has the data?
  - Invictus0 4y ago
    Badly? AI is vastly superior to human capability, no human wins
  - icelancer 4y ago
    We all get slaughtered. It's possible Magnus could make one draw with white against a not-current Stockfish on limited depth out of a hundred games, but that would be about the best any of us could do.
    - zamadatix 4y ago
      Chess engines support the concept of skill level. Pretty much any chess program or app supports this out of the box, such as the aforementioned chess.com. The only times they are run unhindered are for analysis, engine testing, or fun.
- epivosism 4y ago
  There's a very interesting point in the thread below which would resolve the question - we could use chess puzzles! If there is a mismatch between performance on puzzles with live results, it would be pretty solid proof that the effect is real. It doesn't have to be full on AI.
- icelancer 4y ago
  Some women (Polgar for example) have argued the separate womens' titles significantly contribute to the gap in skill.
- lelanthran 4y ago
  I've always wondered about this, actually. With almost all sports males have a dramatic physical advantage[1] so it's not really fair to have males compete with females. With chess there is no physical advantage, so why keep separate tournaments for males and females? What is the reasoning for preventing males and females competing in chess? [1] In the 90s, ISTR reading about a tennis match between a #3-rank female player against a #5000-something-ranked male, and the male player simply dominated. I wish I could find that article now, but it was a print mag and I don't remember which one it was.
bee_rider 4y ago
While we shouldn’t fall for junk science, we also shouldn’t fall for cynicism as an excuse to stick to our priors. It might not replicate if you try to do so, but until someone does, we shouldn’t write it off.
- zeroonetwothree 4y ago
  Surely the burden of proof should be on the side making the claim?
  - bee_rider 4y ago
    I’ve in general grown somewhat averse to the “burden of proof points in some direction” line of thinking. They are both making claims. (And I think usually most people are making some sort of claim, unless they are just nitpicking). One is claiming that they remember a study — pretty vague, and easily proven (as far as we can tell at least) by just showing us any study. The other is claiming that that study doesn’t replicate… without having seen it, which seems like a pretty high bar to clear.
    - solveit 4y ago
      Most studies in this reference class (social science) don't replicate, so I'm fine with that being the default assumption.
      - bee_rider 4y ago
        It is hard to speculate without knowing much about the original poster or the types of studies they read. Maybe they prefer “Nature Human Behavior,” who knows?
    - PartiallyTyped 4y ago
      I can offer [1-3] in support of my position. [1] https://onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.440 https://onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.440 [2] https://journals.sagepub.com/doi/10.1177/0956797620924051 https://journals.sagepub.com/doi/10.1177/0956797620924051 [3] https://journals.sagepub.com/doi/10.1177/0956797617736887 https://journals.sagepub.com/doi/10.1177/0956797617736887
ljf 4y ago
https://onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.440 https://onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.440 The study. Note that this is about taking equally matched players, and having them compete via the Internet without knowing the gender of their partner, then again knowing the gender.
PartiallyTyped 4y ago
This reads needlessly aggressive, and the accompanied downvotes seem unjustified. In addition to the other studies published by Wiley, I have two more: https://journals.sagepub.com/doi/10.1177/0956797620924051 https://journals.sagepub.com/doi/10.1177/0956797620924051 https://journals.sagepub.com/doi/10.1177/0956797617736887 https://journals.sagepub.com/doi/10.1177/0956797617736887
PartiallyTyped 4y ago
@dang Why is GP flagged but parent isn’t? Parent adds nothing to the conversation, but apparently GP was worth removing even though it is now supported by 3 different sources.

Invictus0 4y ago

Women that dedicate their careers to chess still dramatically underperform men at the professional level. There is even a different rating level and title system for women: women earning the "women grandmaster" title need 200 fewer ELO points than the grandmaster title for men. No woman has ever won the World Chess Championship, and there are far fewer female GMs than male GMs.

dudeinjapan 4y ago
How do women and men do versus AI opponents, i.e. without any gender? Wonder if chess.com has the data?
- Invictus0 4y ago
  Badly? AI is vastly superior to human capability, no human wins
- icelancer 4y ago
  We all get slaughtered. It's possible Magnus could make one draw with white against a not-current Stockfish on limited depth out of a hundred games, but that would be about the best any of us could do.
  - zamadatix 4y ago
    Chess engines support the concept of skill level. Pretty much any chess program or app supports this out of the box, such as the aforementioned chess.com. The only times they are run unhindered are for analysis, engine testing, or fun.
epivosism 4y ago
There's a very interesting point in the thread below which would resolve the question - we could use chess puzzles! If there is a mismatch between performance on puzzles with live results, it would be pretty solid proof that the effect is real. It doesn't have to be full on AI.
icelancer 4y ago
Some women (Polgar for example) have argued the separate womens' titles significantly contribute to the gap in skill.
lelanthran 4y ago
I've always wondered about this, actually. With almost all sports males have a dramatic physical advantage[1] so it's not really fair to have males compete with females. With chess there is no physical advantage, so why keep separate tournaments for males and females? What is the reasoning for preventing males and females competing in chess? [1] In the 90s, ISTR reading about a tennis match between a #3-rank female player against a #5000-something-ranked male, and the male player simply dominated. I wish I could find that article now, but it was a print mag and I don't remember which one it was.

bee_rider 4y ago

While we shouldn’t fall for junk science, we also shouldn’t fall for cynicism as an excuse to stick to our priors. It might not replicate if you try to do so, but until someone does, we shouldn’t write it off.

zeroonetwothree 4y ago
Surely the burden of proof should be on the side making the claim?
- bee_rider 4y ago
  I’ve in general grown somewhat averse to the “burden of proof points in some direction” line of thinking. They are both making claims. (And I think usually most people are making some sort of claim, unless they are just nitpicking). One is claiming that they remember a study — pretty vague, and easily proven (as far as we can tell at least) by just showing us any study. The other is claiming that that study doesn’t replicate… without having seen it, which seems like a pretty high bar to clear.
  - solveit 4y ago
    Most studies in this reference class (social science) don't replicate, so I'm fine with that being the default assumption.
    - bee_rider 4y ago
      It is hard to speculate without knowing much about the original poster or the types of studies they read. Maybe they prefer “Nature Human Behavior,” who knows?
  - PartiallyTyped 4y ago
    I can offer [1-3] in support of my position. [1] https://onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.440 https://onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.440 [2] https://journals.sagepub.com/doi/10.1177/0956797620924051 https://journals.sagepub.com/doi/10.1177/0956797620924051 [3] https://journals.sagepub.com/doi/10.1177/0956797617736887 https://journals.sagepub.com/doi/10.1177/0956797617736887

ljf 4y ago

https://onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.440 https://onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.440 The study. Note that this is about taking equally matched players, and having them compete via the Internet without knowing the gender of their partner, then again knowing the gender.

PartiallyTyped 4y ago

This reads needlessly aggressive, and the accompanied downvotes seem unjustified. In addition to the other studies published by Wiley, I have two more: https://journals.sagepub.com/doi/10.1177/0956797620924051 https://journals.sagepub.com/doi/10.1177/0956797620924051 https://journals.sagepub.com/doi/10.1177/0956797617736887 https://journals.sagepub.com/doi/10.1177/0956797617736887

PartiallyTyped 4y ago

@dang Why is GP flagged but parent isn’t? Parent adds nothing to the conversation, but apparently GP was worth removing even though it is now supported by 3 different sources.