Monday, February 08, 2016

The Rudd Test (aka: Why Doesn’t it Rain Indoors?)

Faye’s game wouldn’t pass the Rudd Test

"Why doesn’t it rain indoors?" That’s a pretty good question if you think about it. Especially when it comes from a boy in Mrs Dickinson’s third-year junior school class.

My friend David had noticed something the rest of us hadn’t: it doesn’t rain indoors. Actually we had observed this. It’s just that we’d seen it not rain indoors so often that we didn’t even think about it. The point is not that David had noticed, but that he’d noticed that he’d noticed.

I’m no expert, but I suspect that there’s an evolutionary advantage in being able to forget about stuff that happens all the time. That's all very well, but the downside is that when the everyday passes us by we run the risk of not thinking to question why things are that way. Is it the natural order of things or are they constructed in some way? Brought about by us?

Assuming that we do want to ask why - why, for instance, there are so few women tournament chess players in Britain today - what are we going to do? Well, the first step is that we need to decide that noticing the phenomenon is something that we want to do. Jack Rudd has come up with a way to help us.

More of the Rudd Test later. First, I have to set the scene by talking a bit about women in film.

Mrs Dickinson worked here

You many recall January’s media furore over the absence of nominations (again) for non-White actors in this year’s Academy awards (Similar to the Oscars?). The situation for women in film is clearly rather different. Thanks to Best ActressBest Supporting Actress at least two statues are assured to go to women. Issues around gender and film remain, however, not the least of them being how female characters are portrayed - or not, as the case may be - on screen. 

Women in modern cinema are systematically under-represented in terms of screen time (even in Disney films about princesses), marginalised in roles that are tangental to the main story and very often all but invisible. That’s when female characters are there at all. Of the eight Best Picture nominees for this year’s Oscars, three - Bridge of Spies, The Revenant and The Big Short - have no significant roles for women. Of the remainder The Martian just about does (Jessica Chastain captains the ship that’s on its way to rescue the stranded astronaut but this is very much Matt Damon’s film) and just one the four central characters in Spotlight central is female*. Only Room has a narrative lead by a female character.

Hence the Bechdel Test. Put simply, all a film has to do to pass is have two named female characters appear on screen together having a conversation about something that is not a man. Not exactly setting the bar absurdly high, and yet an astonishing number of films fail to meet even this modest standard.

As instruments of measurement go, Bechdel is certainly a blunt one. Failing the test doesn’t necessarily mean your film is loathsomely misogynistic. Passing won’t necessarily get you the approval of what Nigel Short likes to call the "Tyrannical Feminist Lobby". Nevertheless Bechdel has worth. Principally in helping us notice that we've noticed that it doesn’t rain indoors.

When you consider entire populations of films - all those you’ve seen this month, this year, this life time - the Bechdel Test helps us notice the discrepancy in how male and female characters are portrayed on screen. It reminds us that it’s not The Martian’s gender roles that are the problem. It’s that things are nearly the main character = male, secondary character = female way around.

Jack Rudd
photo by Olivia Netshagen

Last week Jack Rudd proposed we adopt our own version of the Bechdel Test for chess. To pass a tournament must at some stage pair one woman chesser against another. That’s all. It is not, I feel, setting the bar absurdly high.

Jack set some qualifications. The event must not be an all-play-all, not be open only to female participants and not adopt special pairing rules. If I am allowed to fiddle with his idea I might be tempted to add a further point about age - the women involved must be older than, say, 16.

Given these criteria, how many chess tournaments would pass the Rudd Test? Jack suggests that Gibraltar would pass easily. I suspect that most of Sean Hewitt's e2-e4 events - making a welcome comeback at Gatwick in a couple of weeks, btw - probably would too. How many others, do you think? Not many probably. Not enough, certainly.

It doesn’t rain indoors. A tiny proportion of tournament chessers are women. We all know that these things are true. They are so obviously true that we don’t usually pay them any attention. Maybe we should, though. Maybe it’s worth having a think about whether the number of women at chess events is, like rain, a naturally occurring phenomenon or if there’s something else at play. Whether it’s something which we could and should be addressing.

It says something about where we are with respect to gender diversity that every tournament having  even just one game played between two women would be a huge step forward for British chess. The Rudd Test is not the solution to all of chess’s problems. It is, though, something which could helps us think about gender issues. It’s something that would allow us some way to measure progress - or lack thereof - in terms of increasing the numbers of women playing competitively.

The Bechdel Test has become commonplace in discussions about film and the film industry. I hope one day we’ll be able to say the same about the Rudd Test and chess.

* that’s twice as good as The Hateful Eight, admittedly.


Joe Skielnik said...

Q. How many leagues would pass the Rudd Test (apart from 4NCL)?

Jonathan B said...

A good - if depressing - question, Joe.

Matt Fletcher said...

I realised recently that I haven't played a single game against a female player since I came back to chess in 2010 (well over 100 games).

Jonathan B said...

Yup. Similar experience for me - see Chesserettes from five years ago.

I played two women in Benasque 2011 - and two in Sitges last summer, though.

Jack Rudd said...

I think I've played five competitive games against women in the last year - Sabrina Chevannes and Aigerim Rysbayeva at the South Wales International, and Heather Richards, Sarah Hegarty and Akshaya Kalaiyalahan at the British.

ninkibah said...

I did some maths, but I'm sure some of your readers are better at probability than I am, so please check this.

We have an n-round swiss, with f female players, and m male players.
Assuming that the chance of pairing a woman against a man in each round is:
m/(f+m) for the first woman, (m-1)/(f+m-1) for the second woman, and (m-k)/(f+m-k) for the kth woman, then the chances of a round where all women are paired against only men is:
PROD(0 <= k < f) of (m-k)/(f+m-k)

The probability that in n rounds, women will only be paired against men is that number raised to the power of n, and so the probability that at least one game in a tournament will be between 2 women is:

1 - (PROD(0 <= k < f) of (m-k)/(f+m-k))^n

I wrote a little script to calculate values. It turns out that in 6 round event with 5 women amongst 50 players, you have a 96% chance of at least one female/female match.

Jonathan B said...

Well I’m not the person to check that sort of maths, or any other.

Can anybody else help? Matt Fletcher maybe,

I wonder what proportion of tournaments have 10% female entrants, though.

ninkibah said...

In a 100 player event, you only need 7 women for 95% chance of f-f match, but for a 20 player event, you need 3 women for the same chance.

This is similar to the how many people you need at a party to have a 50% chance of 2 of them sharing a birthday. The non-intuitive answer is 25 people.

Looking at the Gibraltar masters, by my rough count, there were 35 female players from 257 entrants. Even a 1 round event will give a 99% chance of a female-female match up.

Jack Rudd said...

Nikibah - your calculation will probably underestimate the probability of at least one all-female pairing (bear in mind that the pairings are not independent of the pairings in previous rounds).

It's a non-trivial problem to solve, and I'd probably use Monte Carlo methods to estimate the probabilities.

Anonymous said...

As you recently posted, in May 2013 some bloke played four women in a row i.e. all the games he played in that tournament, and those were the only women in the tournament out of 28 (going by the names). What are the chances?


Krantz said...

It's not likely to be a popular discussion topic on a chess blog, but sociologically one might invert the issue, and ask why proportionally males "waste" significantly more effort on chess competition? (And Nigel could inquire why females "waste" more effort on something like shopping or fashion.)
I don't know for sure, but I think nowadays (in the UK) men and women have similar amounts of leisure time (maybe men slightly more), and similar spectra of use (though I'd guess women, say, volunteer to charities a bit more and certainly men are more passionate about sport), whereas 50 years ago I don't think either was necessarily the case.
Recreational chess (compared to say recreational bridge) is something that does not exist too well, and those who partake of it are often considered "bums" for lack of a better term. Many chessers would fall into a "social misfit" category as labelled by society. What percentage of chess players have "socially prominent" jobs, again compared to bridge if you want?
Another social aspect, is that middle-aged guys can be (still) involved with chess, is due to their children, while not as much so (if at all?) with mothers. Chess dads playing side games at scholastic events, can be commonplace in some venue (though their population can be on the "bum" side too).

Anonymous said...

From what I have seen in Germany, in Swiss events (say best player is 2400-2500 range) there is sometimes a woman in the top 10 (or top 5) in ratings, but rarely more than one. Zhukova bemoaned that women don't analyze their games in a post mortem ("like men do" she said) when playing each other, and implied that this sort of camaraderie did not fit with the perception of competitive aspects, among women. Anyone who has been to a Bundesliga event can tell you that even 2700+ players are typically quite eager to analyze for 10 minutes with (much) lower-rated players, while the Frauenbundesliga?

Jonathan B said...

Krantz >>

I very much doubt that your idea of men and women enjoying similar amounts of leisure time is correct.

Jonathan B said...

Anonymous >> I'm not sure I see the relevance of your comment to the subject of this blog.

Jonathan B said...

Another anonymous >> your comment was definitely not relevant to today’s subject. Post it again with a context that explains why it has any meaning for the subject under discussion and maybe it’ll get published.

ninkibah said...

@Jack Rudd: Thanks for the hint. I can take your point that players can't play people they have already played, and put that into the formula:

1 - PROD(0 <= r < n) PROD(0 <= k < f) of (m-k-r)/(f+m-k-r))

I rewrote the script using this formula, and you're spot on. The chances for a female-female match in a 6 round event with 50 players, 5 of whom are women is 97% instead of 96%.

You may well be right in thinking that the swiss pairing rules will also effect the outcome, but I'm not writing a pairing program to run a Monte Carlo simulation!

Jack Rudd said...

Division Four of the 4NCL passed the Rudd Test today. (Although not if you add in Bryant's Addendum.)