Tuesday, March 4, 2008

Photos from Eastern Class / Testosterone

Darmen Sadvakasov

Darmen Sadvakasov
Jorge Sammour-Hasbun


Alex Shabalov and Braden Bournival

Braden Bournival thinks


Braden Bournival winks


Braden Bournival smiles


An idea occurs to Alex Shabalov

Alex is suspicious and frightened


Alex prepares on the smallest computer ever


Alex analyzes

I'm sick of writing stupid captions


whatever

I’m conscious that I may be type-casting myself as one of those people who thinks about gender constantly, but I had some thoughts on the subject this weekend that I’m prepared to share with you. It was towards the end of round four, when I realized I was very likely to lose. And I noticed that didn't bother me all that much, which made me both happy and sad. My normal response to internal conflict is to blame some outside person or factor, so I started thinking something like “Maybe there is no cognitive disparity between genders in chess; maybe men are just more aggressive and everything that comes with that: willingness to endure stress for longer periods of time, willingness to sacrifice happiness for achievement, willingness to invest emotional energy in games."
If that sounds ridiculous to you, let me turn your attention to Ilya Krasik's comment: "I play best against opponents when I hate them." Me, I just can't be bothered to work up any hatred.
And this in turn led me to the idea: maybe a woman preparing for a big tournament or match would increase her chances by taking testosterone supplements (not that this idea is either legal or healthy). But the chain of thought led me to Googling the topic, which uncovered a paper, “Testosterone and Chess Competition.” (Mazur, A. & Booth, A. & Dabbs, Jr., J.M. (1992) Testosterone and chess competition. Social Psychology Quarterly 55:70--77.) It’s not free, but you can download it for $14 at http://www.jstor.org/.

Here are some exerpts:
abstract:
The hormone testosterone (T) has a central role in recent theories about allocation of status
ranks during face-to-face competition. It has been methodologically convenient to test the
hypothesized T mechanism in physically taxing athletic contests, where results have been
supportive, although their generalizability to normal social competition is questionable.
Competition among chess players is a step closer to normal social competition because it
does not require physical struggle, and it is the arena for tests of the T mechanism which are
reported here. We find that winners of chess tournaments show higher T levels than do
losers. Also, in certain circumstances, competitors show rises in T before their games, as if
in preparation for the contests. These results generally support recent theories about the
role of T in the allocation of status ranks.

method:
...We studied 16 male players as they competed along with nonsubject players in one or both of two chess tournaments. Their T was measured from saliva samples taken the day before, just before, and just after each round of each tournament....

...The 11 subjects in the regional meeting include two 16-year-olds who competed in the youth division. Four subjects, including one 16-year-old, won three or four rounds of the four-round tournament and are counted here as winners; the remaining seven subjects won zero to two rounds and are categorized as losers....

...For winners we found a prematch rise in T (from Time 1 to Time 2), as we hypothesized and as is consistent with prior research. A period comparison t-test of T2 minus T1 is nearly significant (p= .08, based on only three winners because T1 is missing for the fourth). Also as hypothesized, the winners' T rises above that of losers, significantly so on the morning of the day after the tournament (Time 7: p = .03, t-test). T1, measured in the morning of the day before the tournament, is surprisingly different for the eventual winners and losers. This difference is not only significant (p = .02, t-test) but remarkably consistent: All seven losers have higher T1 values than any of the winners (we discount one winner for whom T1 is missing). This is not an artifact of the normalization procedure because all of the losers' raw T1 scores (ranging from 11.0 to 12.4 ngldl) also are higher than those of winner; (ranging from 2.0 to 10.7 ngldl). It
is not clear which group, if either, departs from normal morning values. Nonetheless, in relative terms the eventual losers of the tournament had reliably high T the day before competing.

discussion:
T was higher in winners than in losers after victories in both the regional and the city tournaments, confirming Gladue et al.'s (1989) finding that T is affected by nonphysical
as well as physical competitions. In the weeks-long city tournament, the effect occurred
primarily in the final weeks and in games where the players were closely matched, supporting the intuitively appealing conjecture by Salvador et al. (1987) that contestants must take their competition seriously if it is to affect their T levels. Our attempt to replicate the prematch rise
in T, which has been noted in physical competition, produced mixed results. There is
little indication of prematch rise in the city tournament. In the regional tournament,
however, which players regarded more seriously, we found a significant and consistent prematch rise among those players who went on to become winners, but not among players
who would become losers. The presence of the hypothesized prematch rise in the regional winners, and its absence in losers, is especially perplexing. The study design did not permit random assignment of subjects to conditions, and regional winners and losers probably differed in some nonrandom but unknown way. One might guess that the winners were better players than the losers, as is true in the city and most other tournaments, but our regional tournament included winners and losers of comparably high skill ratings. Our analysis of the
matchups in the tournament shows that the strong players who ultimately won had easier
opponents than the strong players who ultimately lost, a difference that these highranked
players should not have been able to anticipate on the day before the tournament.
In the Swiss system for setting matchups, two players of similar rank should not know
which one ultimately will face a tougher slate of opponents. Yet on the day before the
tournament, the ultimate losers had higher T than the ultimate winners, a result that is
difficult to explain in causal terms.
I should now wrap this up by drawing some concluion, but I got a nasty illness at the tournament
... I've been asleep for maybe 18 of the last 24 hours.... not really in a state to think much....
But if this entry isn't bloggy enough for you, try http://main.uschess.org/content/view/8241/443/

16 comments:

Anonymous said...

Hello,

I was wondering if you would be kind enough to add my blog to your links.

Getting to 2000 :
http://gettingto2000.blogspot.com

I have added yours to mine

Thanks

Getting to 2000

Anonymous said...

I'm not sure if testosterone really has an effect, but I think I (as an adult male) play better when there are good-looking women in my line of sight or walking around the playing area. How do the researchers account for the fact that a lot of winners at the class level are little boys who, I'd have to assume, have minimal testosterone levels?

HubDiggs said...
This comment has been removed by the author.
HubDiggs said...

The comment from one male player about hate of opponent bringing out the best play is interesting.
I wonder if the Polgar sisters think this way.

There are three Polgar sisters and they are all amazing players, two GMs and one IM. So if this "hate the opponent" idea is the right one you'd think that it would work for them too.

I know that the late Bobby Fischer made statements about wanting to break the opponents will and crushing his opponents psychologically. I wonder if such motivators generally motivate the male chess superstars and if female chess superstars are motivated that way.

Ilya said...

I think what makes one play better is totally individualistic and had little to do with gender or amounts of testosterone. Just to elaborate on the "hate comment"-- it does mean I play better when I hate someone personally but rather something that i develop during the course of the game, it could be based on opponent's mannerisms as well as my own bad disposition.

anjiaoshi said...

The only round I won at the last Boylston Chess Club $10 Open was against a guy who'd been bugging everyone with his constant etiquette violations. I may have been a little more motivated to beat him than I'd been against my other opponents.

Ray Cheng said...

I have only thought about this for a couple of minutes (must go roast a chicken shortly), but I think one possible explanation is the association between testosterone and aggressiveness. Consider two players with identical ratings but low versus high T. The low-T player will play more solidly and cautiously, and perform closer to rating level most of the time. The high-T player will be more aggressive, take more risks, and thus have more spectacular successes - and thus more tournament wins - along with horrific losses (and on average still perform at the same rating level). Would that account for the observations?

Elizabeth Vicary said...

Ray, I think it might explain some things, but for me the confusing part is why the "losers" had much higher testosterone levels the day before games. The study graph is showing and average of more than twice the level of testosterone in losers. Ideas?

Ray Cheng said...

Yes, that is confusing - I skipped over that bit earlier. Having now looked at the Wikipedia entry for testosterone, I still cannot come up with a theory. Even the connection with stress goes the other way: chronic or acute stress (such as that preceding a tournament) tends to suppress the hormone, not raise it. Or is it that the aggressive players that are stewing the night before?

BTW, my chicken was terrific. Garlic, rosemary and lemon peel.

Globular said...

Was the Eastern Class the tournament of Bad Haircuts?

Ironically, Shabba has the best haircut of the bunch (I exclude Jorge, and his Latin good looks of course).

Tom Panelas said...

Now this:

"Russian women to use men as chess figures on Women's Day"

"We had difficulties in finding men to volunteer. It was hard to convince young men to be under [women's] control for the games,"
a spokesman for the organizing committee said.


I'm not sure what light, if any, it sheds on the current debate, but it struck me as vaguely apropos.

Chris said...

Here is my photo gallery from the Eastern Class Championships. You will notice that Elizabeth managed to sneak into one or two of them too! http://www.chesstournamentservices.com/photos/thumbnails.php?album=search&search=ecc2008

Anonymous said...

Globular, best haircut award goes to the almost bald man?? WTF

Globular said...

My point exactly.

Anonymous said...

Yeah Sadvakasvov or Borat or whatever his name is as well as Braden Bournival have 2 of the ugliest cuts I've seen in a long while.

Check2Check said...

“Maybe there is no cognitive disparity between genders in chess; maybe men are just more aggressive and everything that comes with that: willingness to endure stress for longer periods of time, willingness to sacrifice happiness for achievement, willingness to invest emotional energy in games." -I think you've hit on a very important and valid point.

Women it seems are more balanced in life and tend not to obsess on any one particular thing. Since chess skill for the most part correlates to the amount of time invested it would seem logical that women who don't invest an equal amount of time as their male counterparts will not and should not achieve that level of play. There are always exceptions where natural talent is involved of course but I can't think of one woman player in the history of the game who had the rare "gift". i.e. in the likes of a Morphy or Capablanca.

Take Judit Polgar for instance. She is the only female to break into the world top ten FIDE. Did she have any natural talent for the game? Perhaps. Did she work hard and put in many many hours? Absolutely.