Updated Ratings!

[waynegober]

I am completely opposed to round by round Elo.

Eric, can you supply a link showing USCF is scoring round by round in tournaments rather than using average pre-tournament strength of all opponents?

I have been unable to find that anywhere.



Here is another useful link. http://www.glicko.net/research/acjpaper.pdf

[waynegober]

If round by round scoring is most accurate and it is all the same to those running the tournament, I would like to play the highest rated player available each round(in round robin tournaments). The Old Okie won't explain why, but shouldn't make any difference

[Alex_Moiseyev]

I understand your concerns but the ELO system that I am using is the same system yahoo, kurnik, pogo, and other websites with two player games use.

Can you extend this list by any "offline" profesional Federation ?

Having this system in "online" games makes sense, but not sure if it works efficient in static, real events. I can predict paradoxes - due to less number of games and events.

Wayne, Strange system - round-by-round scoring, definely is not most accurate in my eyes.

But Eric is good anyway

[Eric Strange]

Well if you guys don't like it, then by all means, tell Joe to stop doing ratings. I am sure he wont mind Other than assisting Joe with advice and some database work + owning the system coded for this. I really have nothing to do with them

[waynegober]

One more question.....

Are the rounds scored scrictly on win or draw of round?

Surely they are scored on a game basis per round except at a National where Masters is winner take all.

At a tournament such as Alabama Open.
If Pollard wins both games he is getting scored differently than a win and a draw, right?

Thanks Wayne

[Alex_Moiseyev]

you guys

I am talking for myself. If ACF likes it - so be it, I don't object. If ACF doesn't like it - also so be it !

I don't care much unless they start award titles based on these numbers

I am still curious - how it could work on long run. This is fun

But Eric is good anyway.

BTW. Competitive checkers are not the same as phantom, online checkers.

Regards,

Alex

[waynegober]

Well if you guys don't like it, then by all means, tell Joe to stop doing ratings

The system is fine with me...
If I can get answer of k values and scoring in 2 game rounds,
I think I can ajust to where and how I need to play for maximum advantage should it ever become neccessary to take ratings seriously lol.

[Chexhero]

If just a few people disaprove with the ratings I won't stop. But if most people don't like these ratings and think they are crap, then I will not continue with them. But here is what I do know. Before me, nobody stepped up to do any ratings at all. So if I stop, what is going to happen with the ratings? I honestly don't mind the round by round scoring. There is a consistent formula used round by round and I don't really see the problem with that. But, then again, I just put in the numbers, I dont know enough about the formula differences and all to know what is really the best. What I am going to do is finish the ratings up to the present as promised. Then perhaps conduct a poll so I can see what most people really want.

[waynegober]

Joe,

Not trying to get you to stop and maybe it will be just fine.

Your so close you should finish and I thank you and Eric for you work.

I absolutley know it is flawed with round robin though, if you think about it you will agree.

As far as the coding ect.

Kinda like the fellow raising the hood when buying a car. Sees there isn't a motor and asks salesman..."How does this thing run" and Salesman says, "If you don't like the way it runs, don't buy it"

The nice ending to the story is, "I don't need a car"

[Alex_Moiseyev]

I think any system is always better then no system. I support what Joe and Eric are doing and appreciate their efforts.

If I criticize something, it's my personal opinion only and it has nothing to do with simple and obvious fact, that we need to bring ACF ratings back and start running it.

Eric is good anyway.

Alex

[waynegober]

If I criticize something, it's my personal opinion only and it has nothing to do with simple and obvious fact, that we need to bring ACF ratings back and start running it.

Yes, we agree. A car without a motor is better than no car at all.

Last post on this for me

[Chexhero]

Wayne, I think every rating system probably has its flaws. One thing to keep in mind is that most tournaments are not round robin in the ACF. It is actually pretty rare. If a consistent formula is used each round, same for each player, then no, I don't see the big problem. You could be right, but there is no proof on anything. But if you look at the ratings themselves, in my opinion, they becoming more accurate. Alex and King are clearly the top ranked players. Everyone on top of them have not had their rating caculated enough in the system to even be considered close to accurate. That is a bit of a problem, but will go away through time. Your rating is fine where it is. Your low 2000 rating is in the high major low master field, and that is about where you are right? Anyway, I understand any system is better than no system, but if people don't like or believe in the system, then I would not feel right continuing with it. It is more than just accepting it for what it is. People need to at least trust it to a degree, get what I am saying?

[Eric Strange]

Well I tried to make it calculate game by game as close as possible, so a win and a draw is calculated as 1.5 and two draws are calculated as 1 and two wins are 2.0 and a loss and a win are still 1.0, because we don't have the breakdown in order of game by game we calculate is as how many points you got for the round.

for K factor I am using what in research was said to be the most accurate.

Players below 2100 -> K factor of 32 used
Players between 2100 and 2400 -> K factor of 24 used
Players above 2400 -> K factor of 16 used

I do not see how round robin is flawed, you play everyone and get scored vs everyone based on how well you did vs your opponents, it is the way anything should be.

Peace

[Alex_Moiseyev]

A car without a motor is better than no car at all.

Wayne, you didn't undertand ... motor exists, but this is not motor for car, this is airplane motor !

[waynegober]

I do not see how round robin is flawed, you play everyone and get scored vs everyone based on how well you did vs your opponents, it is the way anything should be

I had previously said, I would not post anymore on this thread.
However, I appreciate the civil manner in which you replied to the game scoring Eric. I am very glad you broke it down that way which is right IMO.

Let me attempt to explain my thought on round robin once again, which also applies to Swiss if you consider it.

Lets take 4 players all coming in equally ranked. They are the only 4 players in tournament for my explanation. Three of these players are actually increased in strength through study ect. since their last rating but one player has not progressed. All of the 3 advancing players have advance exactly the same. Now the first player plays our weak player and beats him 2 games gaining several points in rating. Next player plays him and wins 2 but now because his rating has lowered considerably, the second player gets fewer rating points. 3rd Equally strong player plays and again wins both games, but now weak player is way down and player 3 gets hardly any points. The 3 stronger players draw all games and tie for top honors in tournament. However, their respective ratings are now not equal as they should be, but biased according to the order they played the weaker player. He did not gain or lose strength by round but entered the tournament weaker. My argument is all three stronger players should be judged and rated according to his strength when entering the tourn.

Hope this makes sense to you,

Best Regards and intentions
Wayne