PDA

View Full Version : How does the Rating System work?



Alana
30-05-2006, 04:33 PM
Hi everyone. Just wondering, how does the ratings system work? How do they calculate the ratings? I was looking at the results on Ian Rout's site to find out where I came and it said i was going up 32.3 points (or something like that). How is it worked out? How do they know how many points I am going up?

Sorry, I'll stop now, too many questions at the one time :)

Rincewind
30-05-2006, 05:00 PM
Ian would have to verify but I assume that is an estimate based on the Elo system with a particular weighting (k) factor. The ACF rating system is not an Elo system so I would takes Ian's figures are broadly indicative estimates only.

For more information of the working of Glicko check out Prof Glickman's glicko webpage.

http://math.bu.edu/people/mg/glicko/index.html

Desmond
30-05-2006, 09:09 PM
Ian would have to verify but I assume that is an estimate based on the Elo system with a particular weighting (k) factor. The ACF rating system is not an Elo system so I would takes Ian's figures are broadly indicative estimates only.

For more information of the working of Glicko check out Prof Glickman's glicko webpage.

http://math.bu.edu/people/mg/glicko/index.html

Before clicking on the link that had math-blahblah-edu-blahblah I suspected this would not help much. Subsequent to clicking on it my suspicion was confirmed.

Rincewind
30-05-2006, 11:19 PM
Well it has been a few years since I wrote this but perhaps this is of some use...

Glicko Article (http://dreadnought.bjcox.com/downloads/glicko.pdf)

Also there is a spreadsheet to perform calculations

Glicko Spreadsheet (http://dreadnought.bjcox.com/downloads/glicko.xls)

Unfortunately the webiste is currently non-operational in its full form, but the links above should work.

four four two
31-05-2006, 12:30 AM
How does the rating system work?:hmm:

Well there are 3 schools of thought .

A]Badly

B]Mysteriously

C]Effeciently.

Most people seem to agree its a combination of A and B.;)

Garvinator
31-05-2006, 12:37 AM
Sorry RW, but Alana, who is about 11, asked for a simple explaination of how the rating system works. Providing a link to a professor's website to explain how the rating system works in intricate detail probably didnt help much at all.

I think Alana just wanted a simple explaination.

four four two
31-05-2006, 12:41 AM
Sorry RW, but Alana, who is about 11, asked for a simple explaination of how the rating system works. Providing a link to a professor's website to explain how the rating system works in intricate detail probably didnt help much at all.

I think Alana just wanted a simple explaination.

We all want a simple explanation...but you know the likelihood of that hapenning dont you?;)

Rincewind
31-05-2006, 07:36 AM
Sorry RW, but Alana, who is about 11, asked for a simple explaination of how the rating system works. Providing a link to a professor's website to explain how the rating system works in intricate detail probably didnt help much at all.

I think Alana just wanted a simple explaination.

Sorry but I don't assume anything about a person's comprehension based purely on age. Some of the smartest people I know are 11.

BTW The links in my last post were to an article I wrote which was published in Australian Chess Forum a few years back and a spreadsheet that does glicko based calculations that is easy enough that even a Queenslander would be able to use it. ;)

Igor_Goldenberg
31-05-2006, 09:20 AM
http://www.fide.com/official/handbook.asp?level=B0210

Despite being FIDE website, the explanation is quite good. I assume ACF system is simular. If not, perhaps ACF rating officer ought to publish an explanation of ACF rating system simewhere.

Ian Rout
31-05-2006, 09:40 AM
I would think Alana is probably around 14 rather than 11, but I'll let her deal with Garvin.

As suggested by Barry, the estimates are based on the "old" system which is similar to that described in the FIDE Handbook. It works reasonably well for people who have established ratings and mainly play others with established ratings, and play a typical number of games. I wouldn't rely on it to the decimal point if your qualification for the world championship depends on it but in terms of indicating whether you are likely to go up or down and by a big or small number it's not too bad.

Note that each tournament table shows the change due only to that tournament.

jenni
31-05-2006, 09:57 AM
I would think Alana is probably around 14 rather than 11, but I'll let her deal with Garvin.
.
Alana is closer to 15 than 11. At her age she probably wants to be thought older than younger than she is. its only when you get over 35 that being thought 30% younger is a compliment...

Alana
31-05-2006, 04:39 PM
Alana is closer to 15 than 11. At her age she probably wants to be thought older than younger than she is. its only when you get over 35 that being thought 30% younger is a compliment...

Ahh good, at least SOMEONE knows my age around here... :) to be precise I am 5363 days, 16 hours, and 41 minutes old (14 years, 8 months, 31 days, 16 hours and 41 minutes old.

http://junior.ausnetchess.org/games/age.htm

And thanks to everyone who has tried to explain how the system works :)

Alana
31-05-2006, 04:55 PM
and who is ggaryggray anyway??? who thinks i am 11 :@

Alana
31-05-2006, 05:04 PM
The URL below didn't help very well, it is very confusing... :(

http://dreadnought.bjcox.com/downloads/glicko.xls

Oepty
31-05-2006, 05:18 PM
and who is ggaryggray anyway??? who thinks i am 11 :@

ggrayggray is Garvin Gray from Queensland. Just an honest mistake about your age I guess.
Scott

Rincewind
31-05-2006, 05:44 PM
The URL below didn't help very well, it is very confusing... :(

http://dreadnought.bjcox.com/downloads/glicko.xls

:(

Did the pdf help?

four four two
31-05-2006, 05:55 PM
If not, perhaps ACF rating officer ought to publish an explanation of ACF rating system simewhere.

Igor,are you familiar with the expression "When hell freezes over"?:hmm:
Winter is here but dont expect a true explanation this year...;)

antichrist
31-05-2006, 06:09 PM
ggrayggray is Garvin Gray from Queensland. Just an honest mistake about your age I guess.
Scott

Alana, don't be confused by Freddy who signs off as Scott, you know how strange some chessplayers are.

By the way, do you ever miss out on nominating for tournaments because you don't see the notices or forget? We can make special arrangements for you if you do. And who is your favourite poster so far?

Desmond
31-05-2006, 08:36 PM
:(

Did the pdf help?

Yeah I found it much easier to read than the website you posted before. The excel sheet was useful too. Is this what the ACF use?

Rincewind
31-05-2006, 09:35 PM
Yeah I found it much easier to read than the website you posted before. The excel sheet was useful too. Is this what the ACF use?

No. The spreadsheet does a glicko-1 based calculation and so is a more accurate indicator than a Elo calculation but still not what the ACF uses.

Desmond
31-05-2006, 09:45 PM
No. The spreadsheet does a glicko-1 based calculation and so is a more accurate indicator than a Elo calculation but still not what the ACF uses.

So would I be correct to assume that the exact system used by the ACF is not available for the use of the rest of us?

Rincewind
31-05-2006, 09:53 PM
So would I be correct to assume that the exact system used by the ACF is not available for the use of the rest of us?

On the off chance that you are not, you might want to sit down.

The exact system involves several variables which are not available to us. Further to this, the preliminary results of all your opponents in a rating period affects your rating adjustment in a period. So even if you have a perfect spreadsheet, knew all the factors (some of which are not published accurately or at all) you still wouldn;t be able to repeat the rating calculation without all the results of every tournament in a rating period.

four four two
31-05-2006, 09:55 PM
So would I be correct to assume that the exact system used by the ACF is not available for the use of the rest of us?

You would indeed.

Its top secret.:uhoh:

Desmond
31-05-2006, 10:00 PM
You would indeed.

Its top secret.:uhoh:

My legendary powers of observation detect that this is a sore point.

Oh well, I have used the spreadsheet to calculate my estimated rating changes and I'll be interested to see how much difference there is. I would guess no more than a point or two here or there.

four four two
31-05-2006, 10:33 PM
My legendary powers of observation detect that this is a sore point.



On this bb its a long ongoing saga...not unlike Lord of the Rings.:lol: ;)

We have yet to reach the gates of Mordor...:whistle:

Desmond
31-05-2006, 10:35 PM
On this bb its a long ongoing saga...not unlike Lord of the Rings.:lol: ;)

We have yet to reach the gates of Mordor...:whistle:

Oh well, at least you have that climb to look forward to ;)

Cat
31-05-2006, 11:12 PM
Hi everyone. Just wondering, how does the ratings system work? How do they calculate the ratings? I was looking at the results on Ian Rout's site to find out where I came and it said i was going up 32.3 points (or something like that). How is it worked out? How do they know how many points I am going up?

Sorry, I'll stop now, too many questions at the one time :)

The rating system responds to your playing performance and the number of games you play. If you play infrequently, then your rating will change rapidly depending on your performance. If you play a lot of games over time, your rating will stabilise and move much less, even infinitesimally. Essentially you will be penalised for your endeavours.

So if you & your friend play 10 games and each win (say) 8 against the same opponents, if you have been playing more regularly than your friend, then your friend will gain more points. Makes sense, eh?

Basil
31-05-2006, 11:16 PM
The rating system responds to your playing performance and the number of games you play. If you play infrequently, then your rating will change rapidly depending on your performance. If you play a lot of games over time, your rating will stabilise and move much less, even infinitesimally. Essentially you will be penalised for your endeavours.

So if you & your friend play 10 games and each win (say) 8 against the same opponents, if you have been playing more regularly than your friend, then your friend will gain more points. Makes sense, eh?

David. It's a pleasure to have you back.

Bill Gletsos
31-05-2006, 11:17 PM
David. It's a pleasure to have you back.Unfortunately he is sprouting rubbish as usual.

Garvinator
31-05-2006, 11:20 PM
The rating system responds to your playing performance and the number of games you play. If you play infrequently, then your rating will change rapidly depending on your performance. If you play a lot of games over time, your rating will stabilise and move much less, even infinitesimally. Essentially you will be penalised for your endeavours.

So if you & your friend play 10 games and each win (say) 8 against the same opponents, if you have been playing more regularly than your friend, then your friend will gain more points. Makes sense, eh?
David, this is not fair.

Alana asked for a matter of fact answer and you have given a biased answer (see my bolded emphasis).

The gain of rating points also works the exact opposite way- have a double question mark rating and play a poor tournament for that rating, you will lose alot of rating points. Exactly the same difference.

Garvinator
31-05-2006, 11:28 PM
Unfortunately Alana, you will not get a matter of fact answer on here. As soon as one person replies with an answer, another person will disagree and give their answer and it soon degenerates into another slanging match and you will be left more confused than when you started.

There is a long history of rating debates on here with along of slanging matches.

Basil
31-05-2006, 11:29 PM
I don't want to obfuscate this thread - its still a pleasure to have him back - at this stage, anyway!

Garvinator
31-05-2006, 11:31 PM
Alana,

I believe that the following is accurate, or at least accurate enough to give you a guide to the rating system.

The basics are the same as for any rating system. You gain points for winning games, lose rating points for losing games. Also if you draw against some rated higher, you gain points (but less than a win). Exact opposite for a draw against a lower rated opponent.

When the ratings come out each quarter, they contain one of five symbols. The symbols being !!, !, (blank), ?, ??.

Those symbols refer to how active the player has been in the last few rating periods and is a reflection of how accurate the listed rating is likely to be. !! means that the player is very active and therefore their rating is likely to be a true representation of their playing strength.

If you see a player with a ??, then basically the rating has very little meaning and is just a broad guide.

Hope this is sufficent an answer Alana.

Desmond
01-06-2006, 09:17 AM
Consider this scenario:

I have been away from the chess scene for a few years. I come back and play 2 weekend tournaments. I play poorly and my rating dives, let's say 150 points.

I come back in the next rating period and play 2 more tournaments. I've banged out some rust and I perform at my old rating.

Is the following assumption correct: My new low rating is no longer unreliable because of recent activity, and I don't regain all the points I lost.

Rincewind
01-06-2006, 10:15 AM
Consider this scenario:

I have been away from the chess scene for a few years. I come back and play 2 weekend tournaments. I play poorly and my rating dives, let's say 150 points.

I come back in the next rating period and play 2 more tournaments. I've banged out some rust and I perform at my old rating.

Is the following assumption correct: My new low rating is no longer unreliable because of recent activity, and I don't regain all the points I lost.

After an initial loss, if you then perform at your old rating your published rating will approach your old rating asymtotically. The same is true of any rating system I know of including Elo. The only difference with Elo is you MIGHT not loss as many point initially and you probably MIGHT not get the points back as quickly.

Desmond
01-06-2006, 11:03 AM
After an initial loss, if you then perform at your old rating your published rating will approach your old rating asymtotically. The same is true of any rating system I know of including Elo. The only difference with Elo is you MIGHT not loss as many point initially and you probably MIGHT not get the points back as quickly.

www.google.com
define: asymptotically

Ah, I'm with you now. :)

I might prefer ELO if it won't dive me as far.

Cat
01-06-2006, 09:20 PM
David. It's a pleasure to have you back.

thanks, but i didn't know i'd been away. cheers anyway

Cat
01-06-2006, 09:21 PM
Unfortunately he is sprouting rubbish as usual.

what would you know dum, dum.

Cat
01-06-2006, 09:24 PM
David, this is not fair.

Alana asked for a matter of fact answer and you have given a biased answer (see my bolded emphasis).

The gain of rating points also works the exact opposite way- have a double question mark rating and play a poor tournament for that rating, you will lose alot of rating points. Exactly the same difference.

I think the answer I gave was very fair. All qualititative answers contain an element of bias, how that bias turns out depends on your viewpoint. But what I described is true.

Cat
01-06-2006, 09:30 PM
Consider this scenario:

I have been away from the chess scene for a few years. I come back and play 2 weekend tournaments. I play poorly and my rating dives, let's say 150 points.

I come back in the next rating period and play 2 more tournaments. I've banged out some rust and I perform at my old rating.

Is the following assumption correct: My new low rating is no longer unreliable because of recent activity, and I don't regain all the points I lost.

If I can clarify, you are essentially correct. Basically your rating would be more dynamic in the first period and consequently your rating movement would be greater. Bill will probably make some semantic argument over your descriptions and definitions, but what you have summised is essentially true.

Rincewind
01-06-2006, 09:31 PM
So if you & your friend play 10 games and each win (say) 8 against the same opponents, if you have been playing more regularly than your friend, then your friend will gain more points. Makes sense, eh?

Actually that does make sense.

Cat
01-06-2006, 09:35 PM
Actually that does make sense.


Now wots going on here Barry - you say I make sense and Bill says I'm spouting rubbish. Are you trying to pick a fight with Bill deliberately? You know how sensitive he is about these things, lets tread carefully, eh??

There's absolutely no need to be making such contentious statements right now.

ElevatorEscapee
01-06-2006, 09:59 PM
Now wots going on here Barry - you say I make sense and Bill says I'm spouting rubbish...

Actually Bill said you were 'sprouting' rubbish... (careful that one doesn't get marked down as another 'inaccuracy'! :lol:)

Bill Gletsos
01-06-2006, 10:00 PM
what would you know dum, dum.When it comes to ratings you have previously demonstrated you have not got a clue.

Rincewind
01-06-2006, 10:22 PM
Now wots going on here Barry - you say I make sense and Bill says I'm spouting rubbish. Are you trying to pick a fight with Bill deliberately? You know how sensitive he is about these things, lets tread carefully, eh??

There's absolutely no need to be making such contentious statements right now.

Your question of makes sense was obviously trying to be ironic. However, unwittingly you described a scenario which does make sense. So actually I was disagreeing with you. But of course you knew that, you're just gone back into trolling mode.

Cat
02-06-2006, 07:55 PM
Actually Bill said you were 'sprouting' rubbish... (careful that one doesn't get marked down as another 'inaccuracy'! :lol:)

Point taken

Cat
02-06-2006, 08:05 PM
Your question of makes sense was obviously trying to be ironic. However, unwittingly you described a scenario which does make sense.

Only if you consider all children to be drones incapable of adaptive responses, which of course you & Bill do. With Bill, I can understand it, as we all know he is a drone and he earnestly believes humanity is one big beehive. But you Barry have occasionally lent yourself to human interaction, even to the point, dare I say it - where you could even be described as tepid on occasions ( as opposed to warm). I believe you are capable of deeper insights, but it does require you open your eyes! Just a bit.

ElevatorEscapee
02-06-2006, 08:30 PM
Cat, have you considered the possibility that Barry's response about your comment being ironic may in and of itself be ironic? ;) :P

Cat
02-06-2006, 11:19 PM
Cat, have you considered the possibility that Barry's response about your comment being ironic may in and of itself be ironic? ;) :P

Barry's posts are repleat with double entendres. I never know what he's thinking. Whereas as I stand by what I say on both sides of the fence!

Rincewind
03-06-2006, 12:05 AM
Only if you consider all children to be drones incapable of adaptive responses, which of course you & Bill do.

David, the one who is underestimating the intelligence of others is you. But that is probably just due to an over-inflated self-evaluation. Rest assured that your squirming does not appear to be anything other than just that.

Cat
03-06-2006, 12:20 AM
David, the one who is underestimating the intelligence of others is you.

If that appears to be the case, maybe the evidence justifies my evaluation? Impress me.

Alana
03-06-2006, 01:35 PM
Hi everyone!

Sorry I haven't responded for a while, I have been sick these last couple of days and so haven't had much time on the internet... I have been in bed or on the lounge mainly.

I haven't got around to reading the posts properly becuse i have a lot of homework... (evil teachers :( ) but i will soon. Thanks to everyone who has tried to explain it... :( :( :( :confused:

qpawn
15-06-2006, 08:34 AM
The rating system works on a comparative system of player A playing Player B. Both players have ratings. Statistically it is able to be predicted that the higher rated player will win more often than 50 percent of the time against the lower rated one. The bigger the difference in ratings the greater this winning percentage becomes for the better player. When the game concludes the higher rated player will get some rating points for winning or drawing ; but the lower rated player will get somewhat more points than that if he or she wins or draws. After all, it would be a greater achievement for the underdog.

That is my attempt to be dispassionate. Now I will present my opinions of the rating system.

The ACF has just adopted a modified form of the ELO system called the Glicko system. In Glicko a player's rating is given a measure of "stabilty". A player who plays frequently will find it easier to keep their rating stable than someone who plays seldom. Now, this idea of Glicko's is IMO abject nonsense. In 2006 we have certain very nice tools for playing in your private Idaho. One such tool is the internet. Another useful gizmo is the chess computer. Perhaps someone spent their break from tournaments playing over the net??? How on earth can it be deemed for certain that someone's abilties at chess have got worse?? Glicko's reasoning is so laughable that I am astonished that the ACF has adopted a Glicko system. Why should someone's rating go down because he or she has spent the last month on ICC ?? Is this the first ever systemn in the world that actually PUNISHES someone for improving their abilities?? I am speechless. I really am. That a maths professor could conjure up such garbage is mind-blowing. It is the sort of flawed reasoning that I would hope gets dispelled at the undergraduate stage.

Rincewind
15-06-2006, 09:20 AM
The rating system works on a comparative system of player A playing Player B. Both players have ratings. Statistically it is able to be predicted that the higher rated player will win more often than 50 percent of the time against the lower rated one. The bigger the difference in ratings the greater this winning percentage becomes for the better player. When the game concludes the higher rated player will get some rating points for winning or drawing ; but the lower rated player will get somewhat more points than that if he or she wins or draws. After all, it would be a greater achievement for the underdog.

This is an error in this paragraph. The high rated player does not gain any rating points from drawing, he loses them. The next paragraph has number of statements I would call inaccurate...


The ACF has just adopted a modified form of the ELO system called the Glicko system.

Not sure what you mean by "just". The ACF has been using either Glicko 1 or 2 for many years now. Also Glicko is quite a bit different to Elo and to call it a modified form of Elo is not doing it justice. The only similarity is that Glicko is scaled to give numbers in the same range (0-3000) as Elo.


In Glicko a player's rating is given a measure of "stabilty".

In Glicko 1 there is no measure of stability, only reliablity. Unreliable ratings might change more rapidly (if a result is far and away from the expected) but that is all. Glicko 2 has a measure of volatility but I doubt this is what you are referring to.


A player who plays frequently will find it easier to keep their rating stable than someone who plays seldom.

Everyone's rating will move towards their demonstrated strength over time. Ratings with lower reliability will move more quickly as a result of recent performance, because the system has greater confidence in the recent performances being indicative of current chess strength.


Now, this idea of Glicko's is IMO abject nonsense. In 2006 we have certain very nice tools for playing in your private Idaho. One such tool is the internet. Another useful gizmo is the chess computer. Perhaps someone spent their break from tournaments playing over the net??? How on earth can it be deemed for certain that someone's abilties at chess have got worse??

The system does not deem that anyone has gotten worse unless they lose rated chess games. What the system does is says, if we haven't seen a game from this player for a number of years, I'm not going take their rating as reliable as someone who turns in 50 rated games a year.

So If there is someone with a 1500 rating who hasn't played a rated game for 5 years comes back and plays like a 2200 player over a number of tournaments then their rating will rapidly rise to 2200. This is an example of how Glicko more accurately takes into account player's ability to improve via 'net and computer play.


Glicko's reasoning is so laughable that I am astonished that the ACF has adopted a Glicko system. Why should someone's rating go down because he or she has spent the last month on ICC ?? Is this the first ever systemn in the world that actually PUNISHES someone for improving their abilities?? I am speechless. I really am. That a maths professor could conjure up such garbage is mind-blowing. It is the sort of flawed reasoning that I would hope gets dispelled at the undergraduate stage.

These last few sentences are just nonsense. As stated above, people's ratings do not go down from lack of play. Ever. The only thing that causes a rating to move (blanket uplifts excepted) is a result. A good result you go up, a bad result you go down. All rating systems are pretty much the same in this regard.

four four two
15-06-2006, 11:40 AM
While its true that not playing doesnt decrease your rating in of itself,lack of activity will affect how many points you do lose or potentially gain.

For example ,take 2 players who are 1800. One plays tournaments regularly,the other has taken an 18 month break due to work/family.
Lets say they then play in the same tournament and have the same performance level...in this case 1700.
The player who plays regularly will lose less rating points then the player who has taken a break...even though they are playing at the same strength.
Conversely if both players had performed at 1900 strength the player who has been inactive will gain more points then the player who plays regularly.

Rincewind
15-06-2006, 11:52 AM
For example ,take 2 players who are 1800. One plays tournaments regularly,the other has taken an 18 month break due to work/family.
Lets say they then play in the same tournament and have the same performance level...in this case 1700.
The player who plays regularly will lose less rating points then the player who has taken a break...even though they are playing at the same strength.
Conversely if both players had performed at 1900 strength the player who has been inactive will gain more points then the player who plays regularly.

And this is quite correct. The regular 1800 player is known to the system to be reliably 1800 and so this one tournament is less significant than a player who has eben inactive and whose strength may have increased or decreased in that period.

qpawn
15-06-2006, 11:59 AM
In my last few sentences I never denied that results would send someone's results up or down. My point was that the system uses a modifier that assumes someone's behaviour away from the chessboard, namely that someone's level of abilty has been adversely skewed.

I admit that my other point about the Glick system being a modified version of the ELO system could have been articulated better. Both systems use a statistical system based upon a normal Gaussian probability curve; that was the similarity I had in mind.

Of course it doesn't matter what my opinion is of the rating system in the final analysis; I am just a lowly rated nobody according to the Kevin Bonhams et al of this world. I expect that after the Vic Open I will go below 1000 Elo/Glick [??] whatever that means. I always took 1000 as being someone having the ability to play only a legal chess game, 1500 as being average, 2000 as expert, 2300 as IM, 2500 plus as GM. The ratings system has done absolutely nothing for me. In the Vic Open I finished 10 places better than the seeding I began with. But my only reward is to become some sub-1000 know-nothing scumbag who wouldn't be able to move a bishop or a knight. That is extremely depressing. I do know something. Barry, in that first game I played Lasker's defence to the QGD. I only went downhill with a positional inaccuracy [18...Qf6] that ended with a tempting blunder on move 28 in a tough position. Frankly, I don't know whether I will be hanging around in OTB chess to see my rating do down the chute to 3 digits like that.

So there is my bias against the ratings system if there is one. It has done bloody nothing for me apart from telling me that I should give up OTB chess and take up marbles.

I have a lot of respect for players below 1000. In the last game one of them taught me something new in move 4 of my trusted French defence. I have shown it to a few other people and it is a very interesting novelty. Of course according to the ratings I should be saying "French defence? Queen gambit? what on earth are they?"

Rincewind
15-06-2006, 12:33 PM
In my last few sentences I never denied that results would send someone's results up or down. My point was that the system uses a modifier that assumes someone's behaviour away from the chessboard, namely that someone's level of abilty has been adversely skewed.

The system doesn't skew them at all. The most you can say is thatthe system views their ratnig as less reliable. But that could mean their true strength is higher or lower than their rating. You are taking the position that the system assumes they are just lower rated which is simply not the case.


I admit that my other point about the Glick system being a modified version of the ELO system could have been articulated better. Both systems use a statistical system based upon a normal Gaussian probability curve; that was the similarity I had in mind.

This is true. However, they systems are different enough in their mechanics for a description of Glicko as benig a modified form of Elo to be more misleading than enlightening. The ACF in days of yore eomplyed many doctorings of the Elo system for that to be called a modified Elo system. In fact the FIDE system with rating based k-factor system is a modified Elo system too.


Of course it doesn't matter what my opinion is of the rating system in the final analysis; I am just a lowly rated nobody according to the Kevin Bonhams et al of this world. I expect that after the Vic Open I will go below 1000 Elo/Glick [??] whatever that means. I always took 1000 as being someone having the ability to play only a legal chess game, 1500 as being average, 2000 as expert, 2300 as IM, 2500 plus as GM. The ratings system has done absolutely nothing for me. In the Vic Open I finished 10 places better than the seeding I began with. But my only reward is to become some sub-1000 know-nothing scumbag who wouldn't be able to move a bishop or a knight. That is extremely depressing. I do know something. Barry, in that first game I played Lasker's defence to the QGD. I only went downhill with a positional inaccuracy [18...Qf6] that ended with a tempting blunder on move 28 in a tough position. Frankly, I don't know whether I will be hanging around in OTB chess to see my rating do down the chute to 3 digits like that.

I think 1000 strength players know much more than just legal moves. There are many reason why people may lose, draw or win a game of chess. All of these factors go together to make up a players strength and therefore rating. Some 1000 players have much better opening knowledge than me (a 1700 player) but expertise in one facet of the game is not enough in itself to increase rating strength.


So there is my bias against the ratings system if there is one. It has done bloody nothing for me apart from telling me that I should give up OTB chess and take up marbles.

Wel you could play chess for free on FICS, however they use a Glicko ratings system too. :)

I would just suggest that you play and enjoy chess and not worry too much about ratings. Unless you are a junior or elite level player vying for selection, it doesn';t really have that big an impact on your chess playing experience. It can be a good indication of your progress, but if you feel you have a better idea of your own chess enlightenment then go with that.


I have a lot of respect for players below 1000. In the last game one of them taught me something new in move 4 of my trusted French defence. I have shown it to a few other people and it is a very interesting novelty. Of course according to the ratings I should be saying "French defence? Queen gambit? what on earth are they?"

Yep, like I said not all 1000 players are the same in all facets of the game. hat the rating system says is that generally lose more games to 1500 players than they win. This could be poor endgame play, inaccurate tactics or a whole bunch of other reasons and permutations of the same. The point is I think forming too much of an opinion of a player from their rating is a mistake. That is a recipe for losing to low rated players.

In our game for example the position was fairly equal for quite some time and it was only after you missed a tactic that I took the advantage. There is every possibility your opennig knowledge is superior to mine, particularly in openings like the French which I have noly studied a small amount and not played recently at all. So just because we have a rating difference is no reason for me to assume that something you tell me about the French in a bad idea. It would probably more meaningful than anything I could say on the French.

I guess the message is
Give every man thine ear, but few thy voice:
Take each man's censure, but reserve thy judgement.
...
This above all,---To thine ownself be true.

Ian Rout
15-06-2006, 01:41 PM
I always took 1000 as being someone having the ability to play only a legal chess game, 1500 as being average, 2000 as expert, 2300 as IM, 2500 plus as GM.
Not sure how long "always" is, but about 12-15 years ago this was largely true. However the system has re-based somewhat over time so that a beginner will be well below 1000, and will have exhibited far more competence than in the past before achieving that level. If you are correlating a 1000 rating now with that meant in the early 1990s then you will understandably become despondent.

Arrogant-One
15-06-2006, 02:55 PM
How does the rating system work?:hmm:

Well there are 3 schools of thought .

A]Badly

B]Mysteriously

C]Effeciently.

Most people seem to agree its a combination of A and B.;)

Apart from (c) I agree with you. the previous system was vastly superior. For instance at that Qld open 2006 tournament - I was plagued by having to play lots of kiddies (one of whom was playing at least at the 1700 level). that kid beat Tony Weller, and caused me heaps of problems and I had to claw my way to a draw.

Anyway, the point is with the current rating system if you have to play lots of kiddies, and you don't crush all of them because they're so cute and adorable, your rating takes a 150 point dive.

Thats the gist of it 442.

Rincewind
15-06-2006, 03:06 PM
Anyway, the point is with the current rating system if you have to play lots of kiddies, and you don't crush all of them because they're so cute and adorable, your rating takes a 150 point dive.

My recollection of rating system in Australia is that the current rating system is better at rating juniors at closer to their actual playing strength than previous rating systems. This is because a high initial RD factor allows juniors who have not played many senior rated events to move to their actual rating level much more quickly than was possible under the mechanics of Elo. Therefore perhaps you would like to include some justification of your statement that the previous system was "vastly superior".

Regarding your reluctance to "crush kiddies" this could be one of two things. A personality flaw affecting your competitiveness that needs to be addressed. After all, these kids are playing competitive senior events and no quarter should be asked for nor given. Or else it just provides you with a convenient excuse when you lose to a low rated junior. However, neither of these rating system issues.

Arrogant-One
15-06-2006, 03:16 PM
My recollection of rating system in Australia is that the current rating system is better at rating juniors at closer to their actual playing strength than previous rating systems. This is because a high initial RD factor allows juniors who have not played many senior rated events to move to their actual rating level much more quickly than was possible under the mechanics of Elo. Therefore perhaps you would like to include some justification of your statement that the previous system was "vastly superior".

Regarding your reluctance to "crush kiddies" this could be one of two things. A personality flaw affecting your competitiveness that needs to be addressed. After all, these kids are playing competitive senior events and no quarter should be asked for nor given. Or else it just provides you with a convenient excuse when you lose to a low rated junior. However, neither of these rating system issues.

Did I ever tell you that I thought, seriously, about quitting chess after I lost to Ly Maulthan (molton) at the Gold Coast Open when he was just 1400.

No one had told me he was a dynamo. How bloody unfair was that?

Anyway, I then withdrew from the tournament believing that Santa Claus no longer existed and that life had come to an end because a little 1400 had crushed me after 40 odd moves.

Then Graeme Gardiner reported me to the CAQ and put me on an abuser list!

Needless to say I have never withdrawn from a tournament since, but I have made the point about the bad things that can happen to you when you play kiddies.

PHAT
15-06-2006, 03:35 PM
Not sure how long "always" is, but about 12-15 years ago this was largely true. However the system has re-based somewhat over time so that a beginner will be well below 1000, and will have exhibited far more competence than in the past before achieving that level. If you are correlating a 1000 rating now with that meant in the early 1990s then you will understandably become despondent.

This sounds remarkably like it would manifest as a decade of deflation for stable players.

Hmmm, is this not a consequence of calculating and using relative ratings rather than benchmark ratings?

And, is this not also a consequence of rating improving players, using an algorithm that assumes symetry of possible outcomes?

And , is there not also a consequence of assuming that the individuals in the population have a net zero change in actual ability?

And, is that not what Glicko does?

:owned:

Bill Gletsos
15-06-2006, 03:39 PM
Did I ever tell you that I thought, seriously, about quitting chess after I lost to Ly Maulthan (molton) at the Gold Coast Open when he was just 1400.

No one had told me he was a dynamo. How bloody unfair was that?

Anyway, I then withdrew from the tournament believing that Santa Claus no longer existed and that life had come to an end because a little 1400 had crushed me after 40 odd moves.The records show that you actually withdrew from the event not after you lost to Moulthun in round 3 but after you lost to Jacob Edwards in Round 4.

Also Ly's rating history under Glicko2 is as follows:

Players Historical Rating Performance Normal
Period Rating Perf Score Games
Jun 2006 2213 2459 5.0 7
Mar 2006 2171 2198 18.5 31
Dec 2005 2147 7.0 7
Sep 2005 2071 2501 11.5 14
Jun 2005 1979 2154 4.5 6
Mar 2005 1963 2079 14.5 22
Dec 2004 1881 1804 10.0 13
Sep 2004 1894 1980 14.5 22
Jun 2004 1856 1891 19.5 32
Mar 2004 1811 1907 17.5 28
Dec 2003 1647 1770 19.5 34
Sep 2003 1461 1741 7.5 13
Jun 2003 1282 1494 9.0 13
Mar 2003 1171 1143 8.0 16
Dec 2002 1165 1169 5.5 12

where as under ACF Elo it would have been:

Players Historical Rating Performance Normal
Period Rating Perf Score Games
Jun 2006 2117 2454 5.0 7
Mar 2006 2084 2184 18.5 31
Dec 2005 2013 7.0 7
Sep 2005 1967 2489 11.5 14
Jun 2005 1900 2171 4.5 6
Mar 2005 1884 2042 14.5 22
Dec 2004 1812 1750 10.0 13
Sep 2004 1820 1938 14.5 22
Jun 2004 1778 1855 19.5 32
Mar 2004 1680 1864 17.5 28
Dec 2003 1478 1741 19.5 34
Sep 2003 1316 1739 7.5 13
Jun 2003 1220 1433 9.0 13
Mar 2003 1124 1051 8.0 16
Dec 2002 1145 1152 5.5 12

As can be seen he is significantly lowered rated under Elo compared to Glicko2 in virtually all rating periods.

Bill Gletsos
15-06-2006, 03:41 PM
This sounds remarkably like it would manifest as a decade of deflation for stable players.

Hmmm, is this not a consequence of calculating and using relative ratings rather than benchmark ratings?

And, is this not also a consequence of rating improving players, using an algorithm that assumes symetry of possible outcomes?

And , is there not also a consequence of assuming that the individuals in the population have a net zero change in actual ability?Actually it is Elo that is zero sum, not Glicko.

And, is that not what Glicko does?

:owned:I see you are as clueless as ever.

PHAT
15-06-2006, 03:48 PM
Actually it is Elo that is zero sum, not Glicko.
I see you are as clueless as ever.
Yes, Elo is zero sum at every level.

Glicko pretends to be asymetric by increasing RD for volitile players. However at its core, it assumes directional symetry of that volitiliy. That is why there is (as you celebrate) no net inflation/deflation.

:owned:

Garvinator
15-06-2006, 03:56 PM
Matt and Bill, stfu in relation to replying to each other.

Rincewind
15-06-2006, 03:58 PM
Did I ever tell you that I thought, seriously, about quitting chess after I lost to Ly Maulthan (molton) at the Gold Coast Open when he was just 1400.

No one had told me he was a dynamo. How bloody unfair was that?

It's a dog eat dog world. I try to do a little research on my opponents when I can, and a few friends who have been around the traps are a valuable resource as to possible lines, style of play, that sort of thing. So you got ambushed by one of the most talented players this country has produced. It might be discouraging but not worth giving up chess for, is it?

My point is under the old Elo system Moulthan would have probably been rated at around 1100 at the same point in his senior tournament playing career. I can't prove that but it is the sort of thing which Glicko is designed to handle better than Elo normally would.

Edit: Bill replied while I was away. You can see from his post the numbers behind my basic argument. Noit exactly as I state above but I leave them unedited as the general argument was hand-waving and in the right ball-park anyway.

Bill Gletsos
15-06-2006, 04:06 PM
Matt and Bill, stfu in relation to replying to each other.Your contribution to the thread is noted.
However this is a ratings thread and Matt as usual has no clue.

Garvinator
15-06-2006, 04:12 PM
Your contribution to the thread is noted.
However this is a ratings thread and Matt as usual has no clue.
Just basically that most ppl realise that Matt has no clue and so replying to him is pointless. Both replying to each other will result in no other posts from other posters.

Unless the point is to see Matt get another month on the sidelines :whistle:

Bill Gletsos
15-06-2006, 04:18 PM
Just basically that most ppl realise that Matt has no clue and so replying to him is pointless.Yes most people do realise, but not everyone. They also need to be aware of his cluelessness. ;)

Both replying to each other will result in no other posts from other posters.You posted. :hand:

Unless the point is to see Matt get another month on the sidelines :whistle:How he behaves is his responsibility.

qpawn
15-06-2006, 06:30 PM
I don't see how any rating system. ELO or Glicko, can do justice to someone's skillset. Suppose I am playing a good player who is near 2000 rated. I have some miserable rating such as 1200. But I have spent the last month poring over endgames and working my guts out at every endgame problem I can find. Let's say that we both chop a lot of wood and we get to the endgame. Is it still a contest of a 2000 rated player versus a 1200 rated player? I think not :D

I admit that my correlation of 1000 as denoting a minimal knowledge of the pieces' function goes back to about the late 1980s when I read a book that mentioned it. I have to catch up with modern developments in the rating scene. I am sooooo behind the times. Where is my gramophone? When is the next Jimi Hendrix concert? :lol:

PHAT
15-06-2006, 08:13 PM
Matt and Bill, stfu in relation to replying to each other.
u stfu ;)

ElevatorEscapee
15-06-2006, 08:24 PM
Is Saint Fu on the rating list? (I assume that Saint Fu was cannonized by the Catholic Church!)... Or will St Fu's rating forever remain a mystery? :D

PHAT
15-06-2006, 08:28 PM
Just basically that most ppl realise that Matt has no clue and so replying to him is pointless.

You and I have had no problems in the past. It would be a mistake on your part to assume that Gletsos has anymore of a clue than me. He applies code that overly simplifies reality. I understand the behaviour of complex systems. Suffice to say, his approach is smart like a silicon chip, my appoach is wise like a neural network.

Arrogant-One
16-06-2006, 01:00 PM
The records show that you actually withdrew from the event not after you lost to Moulthun in round 3 but after you lost to Jacob Edwards in Round 4.


Thats like saying its not being severely wounded by schrapnel from an exploding grenade that killed you, rather it was the bullet to the head you received two seconds later. A small detail.

Arrogant-One
16-06-2006, 01:02 PM
It's a dog eat dog world. I try to do a little research on my opponents when I can, and a few friends who have been around the traps are a valuable resource as to possible lines, style of play, that sort of thing. So you got ambushed by one of the most talented players this country has produced. It might be discouraging but not worth giving up chess for, is it?

My point is under the old Elo system Moulthan would have probably been rated at around 1100 at the same point in his senior tournament playing career. I can't prove that but it is the sort of thing which Glicko is designed to handle better than Elo normally would.

Edit: Bill replied while I was away. You can see from his post the numbers behind my basic argument. Noit exactly as I state above but I leave them unedited as the general argument was hand-waving and in the right ball-park anyway.

At the time I didn't know he was Auatralia's next Grandmaster. Thats what made it so hard to digest.

Bill Gletsos
16-06-2006, 02:06 PM
Thats like saying its not being severely wounded by schrapnel from an exploding grenade that killed you, rather it was the bullet to the head you received two seconds later. A small detail.Not at all. Your post was saying you withdrew after losing to Ly. That wasnt the case. You withdrew after losing to Edwards and being on 1.5 out of 4.

qpawn
16-06-2006, 05:41 PM
I still contend that the Glicko system is based upon flawed reasoning. Rincewind strongly disagrees with that. And I assume that Bill Gletsos agrees with Rincewind.

Leaving all that aside for the moment, what is the scale of the rating system now? Since 1000 = raw beginner, 1600 = average club player seems to be outdated.

What rating, for example, would a total beginner have who knew the moves only? What is deemed to be an "average" club player?

Without these answers my rating is meaningless. It is like saying that today's temperature is "45678" without having any idea what the freezing point is.

Garvinator
16-06-2006, 06:07 PM
I still contend that the Glicko system is based upon flawed reasoning. Rincewind strongly disagrees with that. And I assume that Bill Gletsos agrees with Rincewind.

Leaving all that aside for the moment, what is the scale of the rating system now? Since 1000 = raw beginner, 1600 = average club player seems to be outdated.

What rating, for example, would a total beginner have who knew the moves only? What is deemed to be an "average" club player?

Without these answers my rating is meaningless. It is like saying that today's temperature is "45678" without having any idea what the freezing point is.

There is no scale per se. Every player receives a rating after they have played nine acf rating games. Therefore, your rating matches your results in your first couple of rated tournaments.

From what I understand of the current system, I am a supporter and the complaints that most players have cannot be addressed by any rating system and especially not by the acf rating officers.

Cat
16-06-2006, 07:07 PM
There is no scale per se. Every player receives a rating after they have played nine acf rating games. Therefore, your rating matches your results in your first couple of rated tournaments.

From what I understand of the current system, I am a supporter and the complaints that most players have cannot be addressed by any rating system and especially not by the acf rating officers.

That's just nonsense Garvin. We can fly a man to the moon, explode a nuclear bomb and grow babies in a test tube, yet you seriously believe a better rating system cannot be found ?? Really!!

You're listening to individuals that have shown little leadership or understanding for the affairs of Australian Chess, yet you seem to believe implicitly that their deep frozen thoughts hold merit. There is nothing that can't be done or imagined. Lack of imagination is killing chess in this country and the rating system is the embodiment of the rot.

A burnt out Zombie could offer something more inspirational. The ACF is locked in a 30 year group-think with their heads up their asses. Better you keep quiet Garvin than peddle their foolishness.

Bill Gletsos
16-06-2006, 07:15 PM
I still contend that the Glicko system is based upon flawed reasoning. Rincewind strongly disagrees with that. And I assume that Bill Gletsos agrees with Rincewind.You can contend it as much as you like. You havent backed it up in any way whatsoever.

Leaving all that aside for the moment, what is the scale of the rating system now? Since 1000 = raw beginner, 1600 = average club player seems to be outdated.

What rating, for example, would a total beginner have who knew the moves only? What is deemed to be an "average" club player?

Without these answers my rating is meaningless. It is like saying that today's temperature is "45678" without having any idea what the freezing point is.Currently 65.82% of active players have a rating higher than 1200.
You are currently in a group of 7.71% of active players rated between 1100-1199.
22.75% are rated under 1000.

Bill Gletsos
16-06-2006, 07:16 PM
That's just nonsense Garvin. We can fly a man to the moon, explode a nuclear bomb and grow babies in a test tube, yet you seriously believe a better rating system cannot be found ?? Really!!

You're listening to individuals that have shown little leadership or understanding for the affairs of Australian Chess, yet you seem to believe implicitly that their deep frozen thoughts hold merit. There is nothing that can't be done or imagined. Lack of imagination is killing chess in this country and the rating system is the embodiment of the rot.You havent a clue as usual.
In fact you contributed absolutely nothing to the ratings debate except babbling on about your maturation factor.

A burnt out Zombie could offer something more inspirational. The ACF is locked in a 30 year group-think with their heads up their asses. Better you keep quiet Garvin than peddle their foolishness.The only person peddaling rubbish is you.

Desmond
16-06-2006, 07:19 PM
I think that a rating system sistered by a series of honourary titles would be worth considering. For example, once you reach 1500, you get a "silver" title or something similair. Something that could stay with you forever even if your rating slipped.

Ratings are viewed (correctly or otherwise) as a personal achievement. Watching your rating slide is much more painful for many than never participating again. I think that is very sad, and one of the root causes of poor tournament numbers in this country.

If the rating system is indeed the best it can be (which seems to be the position of the official who should know), why not introduce a second element to appease the masses and let chess move forward in the country?

Bill Gletsos
16-06-2006, 07:24 PM
I think that a rating system sistered by a series of honourary titles would be worth considering. For example, once you reach 1500, you get a "silver" title or something similair. Something that could stay with you forever even if your rating slipped.

Ratings are viewed (correctly or otherwise) as a personal achievement. Watching your rating slide is much more painful for many than never participating again. I think that is very sad, and one of the root causes of poor tournament numbers in this country.

If the rating system is indeed the best it can be (which seems to be the position of the official who should know), why not introduce a second element to appease the masses and let chess move forward in the country?There hasnt been any clamouring by the masses for any sort of "titles" for various categories of players.

Desmond
16-06-2006, 07:27 PM
There hasnt been any clamouring by the masses for any sort of "titles" for various categories of players.

Hi Bill
I don't dispute this. What I am offering is an idea. One that could take the place of what the masses HAVE been clamouring for (ie removal of Glicko). I am not married to the idea, but perhaps some discussion of various options would be a positive step.

Bill Gletsos
16-06-2006, 07:34 PM
Hi Bill
I don't dispute this. What I am offering is an idea. One that could take the place of what the masses HAVE been clamouring for (ie removal of Glicko). I dont agree with that statement either. There are some who oppose the Glicko but there are equally many who support it. The vast majority of course couldnt care less one way or the other.

Glicko is demonstrably far better than Elo in predicitve accuracy and certainly in handling improving juniors.

Desmond
16-06-2006, 07:36 PM
Ok Bill, have it your way ... I'll keep my ideas to myself and you can get on with the job secure in the knowledge that you have not considered all options.

Kevin Bonham
16-06-2006, 07:38 PM
Of course it doesn't matter what my opinion is of the rating system in the final analysis; I am just a lowly rated nobody according to the Kevin Bonhams et al of this world.

Your rating isn't relevant to the merit of your views on ratings. I've known 1000-rated players who understood ratings systems well and 2200+ players who didn't have a clue.

These days 1000 is a lower-level club player to me. A 1000-rated player usually has some very basic idea about openings and will not typically lose games through elementary blunders of a piece or more, but will only score the occasional point against average serious club-strength tournament players.


I have a lot of respect for players below 1000. In the last game one of them taught me something new in move 4 of my trusted French defence.

I have a lot of respect for the ability of players of ratings around or below 1000 to sometimes come up with something good. I've lost to 1300s (and not just juniors) and been given good tough games by players with 3-figure ratings. However it's a statistical fact that most times I play a 1000-rated player they will be two pawns or more down by move 15, and that even if they're not it remains overwhelmingly likely I will win the game later. I still might not; no-one should take anyone completely for granted.

As for your example of a 2000 player against a 1200 who has done lots of work on the endgame, even if a 1200 can play an endgame at 2000 strength, the 1200 normally won't reach a drawn endgame against a 2000 anyway. The rating is an average across all a player's skills. Many players know 1400s who play 1700-strength openings and 1100-strength endgames, or the other way around. Of course no rating system can predict what a player has done to improve their game since their last game, but in 99/100 cases the answer is "nothing effective".

Another thing to stress is that what the rating system tests is performance under tournament conditions. Some players have a better understanding of chess than their performance under tournament conditions ever indicates. Correspondence chess may be one way to test that that is accompanied by an accurate rating system. Some books offer rating systems against a self-testing scale (stuff of the "Test Your Chess IQ" nature) but these are always suspect and not many people use any specific one.


I have shown it to a few other people and it is a very interesting novelty.

I'd be interested to know about it. Last year a sub-1000 played 3.c3? against my French. It's unsound but I'd never seen it before and after 3...dxe4 4.Nd2 I made an error from playing too fast 4...c5? and white recovered and played a good game.

You probably get the impression I have negative views of low-rated players based on one of our exchanges a while back. But all I was doing there was criticising you for making dogmatic statements about particular openings that a player of your rating is extremely unlikely to be a good enough player to make (no matter what qualifiers or excuses we care to add). And even then I was mainly making that point in such sarcastic fashion because you had got me offside with an unjustified personal attack in another thread.


Without these answers my rating is meaningless. It is like saying that today's temperature is "45678" without having any idea what the freezing point is.

The primary purpose of rating systems is comparisons between players. Another purpose is comparisons between yourself at one point in your career and another - although when you are just starting there is a lot of bouncing about.

Bill Gletsos
16-06-2006, 07:41 PM
Ok Bill, have it your way ... I'll keep my ideas to myself and you can get on with the job secure in the knowledge that you have not considered all options.I'm not ignoring your comments, I'm just splitting your comments into two parts. One part being Glicko the other being local titles.

With regards your titles suggestion, I'm simply pointing out there isnt any push for it.
Note even in your Australian Masters poll that is evenly split 50-50.

qpawn
16-06-2006, 08:01 PM
For the first time in any thread Bill has actually made my rating MEAN something. So my 1200 approx is a certain percentage of the chess populace. We are getting somewhere.

What is the average rating of all rated players in Australia? I even wouldn't mind a mathematical rating distribution if one could be provided, with the mean, mode etc, rather like an IQ bell curve.

Believe it or not, some players like myself are interested in that degree of ratings data.

Desmond
16-06-2006, 08:01 PM
I'm not ignoring your comments, I'm just splitting your comments into two parts. One part being Glicko the other being local titles.
yes, that is what you did. What you did not do was engage in any discussion on the idea. The purpose of my post was to generate discussion.



With regards your titles suggestion, I'm simply pointing out there isnt any push for it.
Note even in your Australian Masters poll that is evenly split 50-50.
The poll is a bit different; many people on the against side gave reasons that amounted to there being no point due to existing FIDE titles. I disagree, but let's leave that aside for the moment. There is no FIDE equivalent for the types of mini-titles I'm suffesting here.

The local title concept was introduced by another poster (Axiom, from memory) and I liked the idea. I'm not sure why there has to be a preponderous push for something to make it worth discussing.

I'd like to mention that I feel I have shown you courtesy. I'm not ragging you or your rating system. There seems to be battle lines drawn very clearly between your side and those who would like to see Glicko torn apart. Much of this debate has been going for so long that a new-comer like myself cannot distinguish between who started what, much less understand the discussions when the inevitable response is "you haven't got a clue". IMO, you are the ratings officer and I take your statements over anyone else's unless I have reason to do otherwise.

Maybe the idea is bad; maybe not. Maybe from discussing it, a new idea will emerge that will be shown to be a leap forward for Australian chess. Maybe we will never know.

Desmond
16-06-2006, 08:03 PM
For the first time in any thread Bill has actually made my rating MEAN something. So my 1200 approx is a certain percentage of the chess populace. We are getting somewhere.

What is the average rating of all rated players in Australia? I even wouldn't mind a mathematical rating distribution if one could be provided, with the mean, mode etc, rather like an IQ bell curve.

Believe it or not, some players like myself are interested in that degree of ratings data.

I too would be interested in such a graph.

Cat
16-06-2006, 08:04 PM
You havent a clue as usual.
In fact you contributed absolutely nothing to the ratings debate except babbling on about your maturation factor.
The only person peddaling rubbish is you.

Why don't you just but out and let the debate run its natural course? What are you so frightened about, that a few people might put some good ideas together and come up with something substantial perhaps?

I'll tell you what, why don't I & others throw their contributions in, have an orderly discussion and you just keep quiet for a while and try not to sabotage the discussion. Then in the freshness of the clean air, let's just see what constructive ideas emerge.

Bill Gletsos
16-06-2006, 08:06 PM
For the first time in any thread Bill has actually made my rating MEAN something. So my 1200 approx is a certain percentage of the chess populace. We are getting somewhere.Yes but you could have gleaned that information for yourself from the ACF ratings page or even from http://www.chesschat.org/showpost.php?p=104327&postcount=13

What is the average rating of all rated players in Australia? I even wouldn't mind a mathematical rating distribution if one could be provided, with the mean, mode etc, rather like an IQ bell curve.You can see the breakdown in the two sources I mentioned above.

Believe it or not, some players like myself are interested in that degree of ratings data.The information is available on the ACF ratings page.

Bill Gletsos
16-06-2006, 08:09 PM
Why don't you just but out and let the debate run its natural course? What are you so frightened about, that a few people might put some good ideas together and come up with something substantial perhaps?You certainly didnt and others dont need to be diverted by your dribble.

I'll tell you what, why don't I & others throw their contributions in, have an orderly discussion and you just keep quiet for a while and try not to sabotage the discussion.You have previously demonstrated you are only capable of misrepresentative dribble.

Then in the freshness of the clean air, let's just see what constructive ideas emerge.If an idea is stupid, it deserves to be identified as such as soon as possible to avoid wasting peoples time on it.

Bill Gletsos
16-06-2006, 08:17 PM
yes, that is what you did. What you did not do was engage in any discussion on the idea. The purpose of my post was to generate discussion.With regards titles you can discuss it as much as you like. I was just pointing out there hasnt been a push for it.

The poll is a bit different; many people on the against side gave reasons that amounted to there being no point due to existing FIDE titles. I disagree, but let's leave that aside for the moment. There is no FIDE equivalent for the types of mini-titles I'm suffesting here.

The local title concept was introduced by another poster (Axiom, from memory) and I liked the idea. I'm not sure why there has to be a preponderous push for something to make it worth discussing.I wasnt commenting on whether it was worth discussing or not. I commented that there hasnt been a push for its implementation.

I'd like to mention that I feel I have shown you courtesy. I'm not ragging you or your rating system. There seems to be battle lines drawn very clearly between your side and those who would like to see Glicko torn apart. Much of this debate has been going for so long that a new-comer like myself cannot distinguish between who started what, much less understand the discussions when the inevitable response is "you haven't got a clue". IMO, you are the ratings officer and I take your statements over anyone else's unless I have reason to do otherwise.And I havent responded to you with "you havent got a clue". I reserve that for those that demonstrate they havent. :)

Maybe the idea is bad; maybe not. Maybe from discussing it, a new idea will emerge that will be shown to be a leap forward for Australian chess. Maybe we will never know.By all means disccus it, but I think you would be better off starting a seperate thread devoted soley to it.
That way it wont get lost in cats dribble.

PHAT
16-06-2006, 08:45 PM
Thats like saying its not being severely wounded by schrapnel from an exploding grenade that killed you, rather it was the bullet to the head you received two seconds later. A small detail.

I understand what you mean. The details are unimportant to what your message was.

(BTW, try to ignore those who would present you as the fool because they cannot understand you.)

PHAT
16-06-2006, 08:50 PM
... complaints that most players have cannot be addressed by any rating system

That is a big call.


... and especially not by the acf rating officers.

You are risking BB death with that statement. :lol:

Kevin Bonham
16-06-2006, 08:55 PM
(BTW, try to ignore those who would present you as the fool because they cannot understand you.)

Stop encouraging others to employ your trick (though you're not the only one) in which the poster blames the reader for the poster's inability to write clearly.

PHAT
16-06-2006, 08:59 PM
... why not introduce a second element to appease the masses and let chess move forward in the country?

:clap: I totally agree with Boris. Chess must compete with MTV, ball sports and the fast living. We need as many bells and wistles as can be promoted.

Look at the martial arts and their belts - what a great way to keep people in the fold.

PHAT
16-06-2006, 09:02 PM
Ok Bill, have it your way ... I'll keep my ideas to myself and you can get on with the job secure in the knowledge that you have not considered all options.

No, don't keep them to yourself. Chess needs people like you. Ignore those who say "it cannot be done."

Cat
16-06-2006, 09:37 PM
You certainly didnt and others dont need to be diverted by your dribble.


Well let them make their own decisions Billy-boy. Nobody's passing any judgements here, there's no opinion given yet. All I'm saying is give people some space to say what's on their mind without you wading in with caustic comments. Hold your tongue for a while, so that people can be free to air their opinions without fear of ridicule or condemnation from you, Bill.

Desmond
16-06-2006, 09:39 PM
No, don't keep them to yourself. Chess needs people like you. Ignore those who say "it cannot be done."

Thanks ... see my new thread for continuation of the discussion

Bill Gletsos
16-06-2006, 10:23 PM
Well let them make their own decisions Billy-boy. Nobody's passing any judgements here, there's no opinion given yet. All I'm saying is give people some space to say what's on their mind without you wading in with caustic comments.I made no caustic comment to Boris. I simply pointed out to him there hadnt been any push for local titles.

Hold your tongue for a while, so that people can be free to air their opinions without fear of ridicule or condemnation from you, Bill.All long as you and others are going to spread misinformation then I am going to respond.