Data analysis: Difference between Male/Female ratings

watcha, if you write any more you'll have basically written half of the article im writing XD, I can confirm that your graphs looks VERY similiar to mine.

I didn't include female participation in my article though. Very cool graph in post #25. I'll have to see if I can generate this somehow.

What program do you use to generate all of the material you are outputting? I use R, but am open to using other programs.

Originally I used simply Open Office to generate charts, but it become inconvenient to manually copy paste the data into the spreadsheet.

Finally I decided to write a charting program from scratch, which understands the format in which the data is stored and can generate charts on the fly.

The program is written in Scala using the JavaFX GUI platform, so it takes Java 8+ and Scala installed on your computer to run it.

Here is the source code:

https://github.com/scalasbtprojects/scalasbtprojects/tree/master/fideplayers/1.0

Drawing a chart is basically a function call to a general charting function, which can take several parameters. Scala allows the use of callbacks, so I can pass a function as a parameter to the charting function. For examle if I want to filter age, I pass the 'age_ok' function as a parameter for checking X values.

Here is the call for the participation chart:

It seemed to depend a lot on age. Like, the 35 to 45 and the 45 to 55 females had a huge hump, but not as much in other age groups.

"Whatever. its not a huge difference and any individual female could possibly the best player ever."

True, but it's probably not as likely as an individual man being the best player ever. In other words she would be a more far fetched "exception" than for an individual man to be the best player ever. But yeah I do think that in these discussions, people forget that everyone is an individual. Statistics probably "suggest" I should be a 1600 because somewhere around there is probably the average uscf rating. But look at that I'm actually much higher rated than that. Looks like I know myself better than the data. It's good to know what the stats talk about and what they don't talk about.

I never thought it was about one gender being inherently superior to the other. There's a lot of peer-reviewed research involving hormonal influence on risk tolerance. Testosterone encourages not just muscle growth, but also risk-taking behavior. (To a lesser extent, so does estrogen. Cortisol and progesterone, on the other hand, are positively correlated with risk aversion, and both of these hormones are higher in women -- and of course, progesterone is virtually non-existent in men).

In our evolutionary past, this makes sense. In a communal context, it is good for group survival if the gender that bears and raises children doesn't charge headlong into buffalo herds. On the other hand, members of the other gender are individually more expendable, and greater risk tolerance can mean the difference between a successful hunt and starving over the winter -- even if one or two individuals are killed in the process. We don't face such risks any more, but culture evolves at a much greater rate than biology, so this sexual dimorphism -- both physical and behavioral -- is likely to be with us for a long time.

These same differences affect our behavior on the chessboard, where greater risk tolerance often translates into the difference between a full point and a half point. (Of course, it also can mean the difference between a half point and zero.) This alone would account for a higher standard deviation among males in almost all pursuits, and statistics do seem to bear out this prediction.

It would be interesting to see whether there is a correlation between female chess performance and progesterone levels. I suspect there would be a fairly strong connection.

Deleted.

"These same differences affect our behavior on the chessboard, where greater risk tolerance often translates into the difference between a full point and a half point. (Of course, it also can mean the difference between a half point and zero.) This alone would account for a higher standard deviation among males in almost all pursuits, and statistics do seem to bear out this prediction."

Although it does seem like chess players in general can often be pretty risk averse, at least in their chess. GMs playing for a quick draw, or drawing a much better position because their opponent is higher rated. Although it's just based on some general observations, so it just seems that way to me, maybe there are a lot of risk takers in chess as well.

Wow, cool, I'll have to download it and try it out. Thanks for sharing!

Elubas: I'm not talking about foolhardy risk here, but things like moving into complications when the outcome is less than clear. Think Kasparov or Perosian or Mickey Adams. Female chess players tend to be more likely to seek simplification, when the result of a particular decision is a bit more concrete.

That said, males are more likely to make speculative sacrifices that are less than sound, too. Tal or Grischuk would be of this latter category. (Well, Grischuk tends to sacrifice his clock more than material, but it's in the same category of risk.)

Source?

@0110001101101000: Admittedly anecdotal. Fischer, when asked by Tal what he thought of former Soviet women's champion Larissa Volpert's style of play, described it as "too cautious." Magnus Carlsen echoed those sentiments, saying "Perhaps the ladies must dare to be tougher on the board. I think they are just too cautious, maybe."

Perhaps I am overgeneralizing, but I strongly suspect that an analysis of a large number of games would support these statements, just as analysis of investment strategies has done.

@FiveofSwords: See my previous answer on this point, where I differentiate risk from foolhardiness. Risk worked just fine for Tal, who was world champion. A different sort of risk seems to not hurt Alexander Grischuk, who will let 3/4 of his clock go away on a single move, counting on seeing every variation in a complex position (and not uncommonly missing something, so the risk doesn't always pay.)

Wilford-n :
It seems to me that women are rather more aggressive than men, and tend to take more risk.
Admitedly it could be a bias because the women games I tend to look at are in the 2500-2600 rating range, and it may only be that GM (male and female) tend to play more aggressive/tactical openings than 2700+ players.

@Angelo, 0110001101101000: See this excerpt from a chess-specific study by Gerdes and Gransmark of Stockholm University, posted on this website here. A relevant quote:

"We have looked at different outcomes aimed at measuring strategic behavior. We have found that women choose more cautious strategies than men, which we interpret as women being more risk-averse on average."

Useless generalisation.

I was always critical on my wife for playing way to tactically/ agressively even if there is no need for such measures. But this is really her style, she is used playing like that, and she enjoys it- in the last tournament we both played (a few days ago), she won the tournament by tactically outplaying her opponents out of "slightly" suspect positions, while I was just able to pick half a point less.

@ Wilford-n : it seems it's just me then.

@FiveofSwords: I do not discount the facts your present about neurons (and, it turns out, the same can be said of neural connectedness), but to a great extent, we are certainly subject to hormonal response. Try "overwriting" your natural response in a true fight-or-flight situation. Cerebral response is short-circuited when we are flooded with hormones like adrenaline, cortisol, testosterone, progesterone, prolactin... the list goes on. In many situations, even our much-vaunted "higher processes" are subject to a cascade effect that originates in the white matter of our mid-brain. You can no more change it than you can change your patellar reflex.

No, but competition does result in a release of hormones that have a measureable effect. It's not a question of whether these hormones influence behavior, but how much they influence it.

This is so strange. I don't feel or act any different when playing a female opponent.

I guess for some men (and women apparently) it makes a big difference.

@0110001101101000: I'm assuming your response is to the finer points of the paper cited, and not my specific quote... where the data shows that both men and women tended to use more aggressive strategies against female opponents. The fact that you "don't feel or [think you] act any different" is kind of the point. Hormonal influence and instinctual behavior act at a level below consciousness. The study participants probably weren't aware they were acting differently, either... after all, the games analyzed were drawn from real-life playing situations. They weren't played in a lab.

I suppose it's much too cliche to say "I am an exception!"

heh, oh well.

It should not be possible to measure this anyway... if I'm not aware of calculating or planning any differently compared to previous games, and when I analyze the games I don't see a difference, then how will someone who studies my games see a difference?

"Your knight takes pawn was risky"

No, it was a miscalculation, I simply missed the reply.

"But you were trying to win because they were female"

No, knight takes pawn intended to force a draw, as I said, I miscalculated.