2021 US Championship

[Bob Armstrong]

Hi Pargat:

Pargat:

"if the women's pool grows to be the size of the men's pool, AND the women are every bit as good as the men collectively, then the men's pool will continue to have its collection of 2700+ and 2800+ players and the women's pool with have it's own collection of 2700+ and 2800+ players."

Of course you are right.

IF the elite in the women's pool are of the same RATING (As well as strength) as the elite in the men's pool.

The pragmatic issue we face is that in the women's pool, at the moment, the average rating of the elite women is below the average rating of the elite men. You agree to this from all I've read.

Now you magically propose a model where the men's pool (I'd prefer calling it now the open pool, to get rid of your inaccurate sexist charge) average elite rating is the same as the elite rating in the women's pool!

How did you get there? How did the elite of the women's pool get so high, so as to equal the average elite rating of the Open Pool?

Bob A

[Pargat Perrer]

Hello Pargat.
.....
So.....what you're saying is that if you have a set of chessplayers and you create a second set of chessplayers where the elements of the second set are identical to the elements of the first set, then the two sets would be equal? This may be mathematically correct but I don't follow how something so trivial demonstrates anything useful about men's and women's ratings.

It demonstrates that if we agree that men and women are equal at chess abilities, and the total pool size of women chess players equals the total pool size of men players, and the pool size in each case is sufficiently large to fully distribute the Elo ratings once some number of games have been played, then as you said, the elements of the first pool are equal in chess strength to the elements of the second pool, with allowance for statistical error introduced by the Elo rating system.

Which means that in each pool, whose members we are supposing only play amongst themselves,

IF
you have 2 men over 2800 and 15 men between 2700 and 2800,
THEN
you will have 2 women over 2800 and 15 women between 2700 and 2800

Any differences will be statistically within the margin of error and not significant. So you might have 1 woman over 2800 and 16 women between 2700 and 2800.

Since we are supposing that women as a collective group are every bit as equal at chess ability to men, if follows that growing the size of the women's pool to be the same as the size of the men's pool will allow the top women to achieve 2800 / 2700 ratings WITHOUT REQUIRING them to play the men to get there.

And therefore, the effort to get women's ratings up to the same level as men's as a group should be centered on growing the women's pool of players.

I'm not saying there is anything wrong with women playing against men, I've already made that clear in a past post.

I'm saying it isn't REQUIRED that they do so to get their ratings up to men's ratings.

And so the message to women in competitive chess should not be "You must play men to get better!", it should be "Encourage more girls and women to join competitive chess!"

[Pargat Perrer]

..... Pargat is out to lunch on rating systems! He seems to believe that a rating is some sort of absolute measure of chess ability. It isn't. It is just a number arrived at by shuffling around performance points between players you normally play with, yes the "rating pool".

You make it sound like we can dispense with ratings altogether!

Ratings are used by tournament directors to do PAIRINGS! So that the cream rises to the top. That means ratings reflect chess playing strength.

And yes, sometimes a player of lower rating beats a player of higher rating, for one or both of 2 reasons:

1) the rating system in chess is too slow because of the slow aspect of playing chess
2) the lower rated player did some home schooling and improves his or her chess ability in some meaningful way since the last ratings were done


Who is out to lunch?

Women's ratings are lower than men's because there are far fewer of them playing top-level chess

Not quite, but you're almost there. There are far fewer of them playing competitive chess. And that is the point i've been making in this thread.

While quite entertaining, it is a pity to see Pargat's arguments become wilder by the day, as he resorts to nasty little insults. Surely you're better than that...?

Wilder??? I'm still making the same argument I was making all along. The men here are all wearing blinders, nothing i can do about that.

I didn't begin the insults. I only gave back as good as I got.

[Pargat Perrer]

It was clearly a mistake to think ratings could be explained on chesstalk without hostilities.

Yeah, when you call the person you are debating with "young grasshopper" which is almost an ethnic slur, but even without the ethnic element, it is demeaning and making yourself to be the Oracle of Wisdom.

YOU began the "hostilities", dear sir.

Maybe, given your obvious problems with English grammar and getting the correct statements across, you should just stop writing completely.

[Pargat Perrer]

What Bob G and I are trying to get you to understand is that it all depends on the rating configuration of the two pools of players (Let's say equal numbers of men and women in each group).

And what you and Bob G. are not understanding is that if the women's pool grows to be the size of the men's pool, AND the women are every bit as good as the men collectively, then the men's pool will continue to have its collection of 2700+ and 2800+ players and the women's pool with have it's own collection of 2700+ and 2800+ players.

Ask yourself this Bob: WHO ARE THE MEN PLAYING TO GET THEMSELVES A 2700+ OR 2800+ RATING? They are playing AMONGST THEMSELVES!

They are not playing computer engines and stealing points from that rating pool The men's pool reached its current level or ratings all on its own.

And if the women are every bit as good as the men, and have the same pool size, they too will have as many 2700+ and 2800+ ratings, all on their own.

When we agree that women and men are equal at chess ability, there is no "rating configuration" to speak of once the two pools are the same size. Each pool's "rating configuration" is identical, that is part and parcel of the agreement that women are equal to men at chess ability.

And the size of each pool is large enough to make any actual differences statistically insignificant -- IF INDEED women are equal to men at chess ability.

So the real goal should be not to have women play higher-rated men, but to grow their pool to be the same size as the men's pool. Then we can compare the two pools and see whether women are in fact every bit as good as men, and it is my belief that they will be.

....all the top 10 will randomly beat each other, and just trade rating points, and stay at 2500...all the 10 elite. The only way this will not happen is if one of the elite happens to be a Magnus Carlsen, and can beat up on the other 9 elite players. Then that one rating will go much higher than the average rating of the other 9 elite in the end.
........
Strength only increases your rating when you go outside your pool and win points there, because you are in fact UNDER-RATED, because you've been stuck in your pool.

Do you not see the contradiction in your statements? If a Carlsen appears among the women, she will increase her rating much higher by beating up on the other elite players in her pool .... strength only increases your rating when you go outside your pool ..... Gotcha!

No need for personal attacks, or claims of bias, Pargat.......we are just discussing a very important issue, not understood by many chess players, and disagreeing. And it may be that in the end we will just have to agree to disagree, and remain chess friends.

Bob A

I have no personal issue with you. I still say you don't have a bias, you are just not able to see the forest for the trees. Bob G., on the other hand, seems to be in complete disarray. He was cocksure he was going to teach all the "young grasshoppers" the lay of the land, and then whoooops! He found out right off the bat that he didn't have any grasp on it, and so he wasn't going to be teaching us all the "many lessons" we all required because he got schooled. And now he's running away and saying too much aggravation, too much hostilities, when he STARTED that by demeaning language directed towards me.

The problem with agreeing to disagree is that your view is wrongly held by virtually all the men in chess. And the women don't like it. So it creates disharmony between the genders in chess, and that just hurts the overall goal, which is to bring more girls and women into chess.

Now, you say it it time to get rid of the women's-only title system. But given these strains between men and women in chess, and the fact that no men are willing to adjust their views and realize fully that women are the equal of men in chess abilities, I have to think that now is not the time, even if these titles should eventually be removed. If it is done now, I think women will leave competitive chess in droves.

[Kerry Liles]

Well, whether or not the problem is my poor grammar or your comprehension, we clearly are not communicating effectively. That's too bad.
It was nice to play OTB in Waterdown this past weekend, and as usual I got a few questions about ratings. It was a pleasant exchange demonstrating in person communication is so much more effective. It was clearly a mistake to think ratings could be explained on chesstalk without hostilities.

I am happy to address rating questions in person or email, just not on a blog. Life is too short for all this aggravation.

Life is too short for all this aggravation.

+1

I'm out. Thanks for all you do Bob - even suffering fools gladly.

[Peter McKillop]

Hello Pargat.

...... Right now we have Yifan Hou at the top of the FIDE women's rating list, 2658. Let's pretend that Hou has never played a rated game against any men, and that she will only play against other women. In fact, all the women will only play against women. Let's also pretend that Hou's abilities are actually every bit as good as Carlsen's.

That means Hou will start beating all the other top women consistently, and her rating will eventually reach Carlsen's within some margin of error, say 1%. She would become the women's version of Carlsen. Her rating will get to be the same as Carlsens, and this means she will be much more dominant against the women than Carlsen is against the men, because her competition is not as strong. But here rating WILL still reach Carlsen's given enough rated games played.
......

My problems with the above are the word "consistently" and the phrase "given enough rated games played." No one, even Carlsen, consistently plays at their very best. Mistakes will be made for whatever reason, opponents will ocassionally play brilliantly for whatever reason, and Hou *may* never achieve the escape velocity required to leave her sisters on earth.

......
Now let's pretend some new woman comes along who is every bit as good as Caruana in the men's group. So this new woman starts beating all the top women regularly, except she loses to Hou at the same rate that Caruana loses to Carlsen. This new women's rating will eventually reach Caruana's rating.

If there were an identical woman for every man in the mens' pool, then each identical woman would have the same rating as her identical counterpart in the men's group, again accounting for margin of error. This would happen once the women's group reaches the same size as the men's group. And all without women playing against men. ......

So.....what you're saying is that if you have a set of chessplayers and you create a second set of chessplayers where the elements of the second set are identical to the elements of the first set, then the two sets would be equal? This may be mathematically correct but I don't follow how something so trivial demonstrates anything useful about men's and women's ratings.

[Fred Harvey]

You seem to be saying that chess players cannot improve by playing other players in their so-called "rating pool", whatever that is (never heard that term before).

It seems to me that this astounding admission from an earlier post is the clue as to why Pargat is out to lunch on rating systems! He seems to believe that a rating is some sort of absolute measure of chess ability. It isn't. It is just a number arrived at by shuffling around performance points between players you normally play with, yes the "rating pool". If one were to examine the formulae used, you would see how difficult it is to rise above the top ratings in you pool. Women's ratings are lower than men's because there are far fewer of them playing top-level chess - nothing to do with ability, possibly because of different wiring and priorities. While quite entertaining, it is a pity to see Pargat's arguments become wilder by the day, as he resorts to nasty little insults. Surely you're better than that...?

[Bob Gillanders]

I was silly enough to say that my correction to Bob G.'s grammar was nitpicky, but really, it wasn't nitpicky at all. The grammar was atrocious! He really made it out to be that you modify the opponent's rating and leave your own rating alone!

Keeping in mind that many of Canada's residents are not having English as their first language, this guy should NEVER write anything for the CFC that is for general public consumption!

Well, whether or not the problem is my poor grammar or your comprehension, we clearly are not communicating effectively. That's too bad.
It was nice to play OTB in Waterdown this past weekend, and as usual I got a few questions about ratings. It was a pleasant exchange demonstrating in person communication is so much more effective. It was clearly a mistake to think ratings could be explained on chesstalk without hostilities.

I am happy to address rating questions in person or email, just not on a blog. Life is too short for all this aggravation.

[Brad Thomson]

Therefore, to blame women for a system that was, in effect, forced upon them by men, seems sexist...

If there is no sep. but par. system for women then men still might not benefit because some/most/all of the women's prize money probably comes from sponsors who *want* to support the women's section; i.e. no additional money for the open section.

My inent is not to specifically blame women for the sexism, I am stating that the sytem is sexist. Perhaps true, but not justifcation for the continuation of sexism.

[Brad Thomson]

Brad, if a woman plays in an open section, and finishes 2nd, does she win BOTH the 2nd-place money AND the top woman money (assuming a man finished overall first)?

What has this to do with the price of tea in China?

[Bob Armstrong]

Hi Pargat:

You -

"- 2 separate groups of Elo rated players,
- each group the same size, at least several hundred members and possibly into the thousands,
- each group member only plays rated games against the members of his or her own group,
- each group is equally proficient at chess, and works just as hard at improving their chess

then the 2 groups, after many years of rated games played within their group, will have a rating Bell curve matching to the other group's Bell curve almost exactly, with allowance for whatever margin of error is built into the Elo rating system."

What Bob G and I are trying to get you to understand is that it all depends on the rating configuration of the two pools of players (Let's say equal numbers of men and women in each group).

Pool A (20 players) - if there are an elite 10 players and are all in the 2400 range, they will indeed pick up points from the lower rated sacrificial lambs. Lets say each of the 10 is able to pick up 100 points from the bottom 10 when they beat them. They will raise the elite pool average rating to 2500. But then that becomes the ceiling to the pool, because all the top 10 will randomly beat each other, and just trade rating points, and stay at 2500...all the 10 elite. The only way this will not happen is if one of the elite happens to be a Magnus Carlsen, and can beat up on the other 9 elite players. Then that one rating will go much higher than the average rating of the other 9 elite in the end.

Pool B (20 players) - suppose the 10 elite players are all in the 2600 range. Same problem.......their cap will be set at their increase from the bottom 10; lets say 100 points again. So the ceiling for this group is 2700, and will never rise above this no matter how long they play in the pool, nor how long they study, and increase their actual playing STRENGTH. Strength only increases your rating when you go outside your pool and win points there, because you are in fact UNDER-RATED, because you've been stuck in your pool.

No need for personal attacks, or claims of bias, Pargat.......we are just discussing a very important issue, not understood by many chess players, and disagreeing. And it may be that in the end we will just have to agree to disagree, and remain chess friends.

Bob A

[Pargat Perrer]

Bob doesn't even get his grammar right, to the extent he makes it seem that for provisional rated games in CFC, you should modify your opponent's rating and not your own, this guy says I don't comprehend! LOL

English is not my first language and I understand and comprehend better than him.

I was silly enough to say that my correction to Bob G.'s grammar was nitpicky, but really, it wasn't nitpicky at all. The grammar was atrocious! He really made it out to be that you modify the opponent's rating and leave your own rating alone!

Keeping in mind that many of Canada's residents are not having English as their first language, this guy should NEVER write anything for the CFC that is for general public consumption!

[Pargat Perrer]

Hi Bob A.

You are amazing. I don't know anyone as patient as yourself.

I understand what you are saying, nothing sexist about it. The problem is not what you are saying, but what Pargat is hearing.
I have had the same problem with him. It can be exhausting trying to correct him comprehension.

And yes, I can verify statement #3, my friend.

Ah, so indeed it is the "Old Boys Club" here. Two old men telling women that they MUST play men to improve at chess!

So we have Bob G., the person who said he was going to teach all of us everything about the way the CFC does Elo ratings, and then when his initial experiment failed, he disappeared without answering any questions, and in fact had to ask for "math experts" to help him out.... and he is BACKING Bob Armstrong on Elo ratings knowledge!

Bob. G, who disparagingly calls me "young grasshopper", then gets upset when I give back as good as I got, now says I am not comprehending what he and Bob A. are saying! Why? Because I disagree!

I disagree, so I am not comprehending what they are saying! LOL

Bob doesn't even get his grammar right, to the extent he makes it seem that for provisional rated games in CFC, you should modify your opponent's rating and not your own, this guy says I don't comprehend! LOL

English is not my first language and I understand and comprehend better than him.

Yes, it's the old "don't attack the message, attack the messenger" trick by Bob G. I actually thought he had more character than that.

Neither of these dunderheads understand this: if there are

- 2 separate groups of Elo rated players,
- each group the same size, at least several hundred members and possibly into the thousands,
- each group member only plays rated games against the members of his or her own group,
- each group is equally proficient at chess, and works just as hard at improving their chess

then the 2 groups, after many years of rated games played within their group, will have a rating Bell curve matching to the other group's Bell curve almost exactly, with allowance for whatever margin of error is built into the Elo rating system.

That means if one group member has a 2850 Elo rated member, the other group will have a corresponding 2850 member, plus or minus the margin of error.

So much for these self-professed experts. I can see why women make such fun of men.

[Bob Gillanders]

Hi Pargat:

1. A woman of Carlsen's talent may well be buried in the separate, parallel, women-only title system. But to date, no woman has streaked ahead of her peers in the women-only system. Hou Yifan got to top ranking for women because for a while she refused to play in the women-only system, and began playing only in "Open" tournaments (Not "men-only"). Then her rating began to rise.

2. It is my error to have said that women must play "men" to improve their ratings. What I should have said is that the women-only title system should now be abolished (It has served its "incubator" purpose), and then women will play in the "Open" tournaments (Where, at the moment, and likely not for long if what I want happens, the top players are all men). When women enter the top echelons of ratings, then other women will be playing them as well as top men and will thus raise their ratings.

3. You are going to find out from Bob G, that I do very well grasp the fundamentals of the international ELO system.

4. I appreciate that despite my argument "appearing" sexist to you, you have taken note that I am far from sexist. My whole argument is for the purpose of benefiting women's chess internationally (Despite the fact that high-rated women, being in a conflict of interest position, will not support my project, because a good portion of their income comes out of the FIDE women's title system, and they do not want the system to disappear).

Bob A

Hi Bob A.

You are amazing. I don't know anyone as patient as yourself.

I understand what you are saying, nothing sexist about it. The problem is not what you are saying, but what Pargat is hearing.
I have had the same problem with him. It can be exhausting trying to correct him comprehension.

And yes, I can verify statement #3, my friend.