Main Forum | Post Reply
Top Players
Play gin-rummy tournaments online
It's now:   Oct 13, 3:10pm EST
Top Players
Posted by Menkman74 (VIP) 28 Apr 2004 10:10am     


I noticed that one of the sections in 'Top Players' is Free Games played and the way it is ranked is by rating. Well, we all know that a players rating is no indication of skill or talent and is virtually meaningless in here . I would like to propose a switch from ranking it by rating and switching it to 'winning percentage' (with maybe a minimum games played such as 100 otherwise 1-0 and 2-0 players will all be up top). Not most games won, but actually winning percentage. I think the top players in here with the best record/winning % deserve more recognition than players who just play each other with high ratings to keep them high.

Bottom line, a player should be judged on their long term success and their overall record compared to a 2100 rating whose significance in here has little or no meaning. Thoughts?

Re: Ratings vs. Percentage of win
Posted by webmaster 28 Apr 2004 1:49pm     


Although ratings may be at times invalid, they were designed (by professor Elo many years ago) to serve as an indicator of performance across all games.
As opposed to the seemingly logical & straightforward win/loss ratio, ratings depend on who you play against.
For example, a strong player A who plays mostly against strong players and a weak player B who plays mostly against weak players will in all likelyhood have the same win/loss ratio. Their ratings, however, will reflect their strength in most cases.

Simply Not True
Posted by calc_guy 28 Apr 2004 9:29pm     


Although ratings may be at times invalid, they were designed (by professor Elo many years ago) to serve as an indicator of performance across all games.

No. The title of Professor Elo's book was 'The Rating of Chessplayers, Past & Present.' The formula has been adapted by others to various games, such as Go, Backgammon, and Football, by adjusting the constant factor representing the variance of a game -- ie. the 'luck' facor. (See, for example, http://www.cs.arizona.edu/-gary/backgammon/elo.html , or http://gobase.org/studying/articles/elo/ .) But when I asked how it had been adapted for gin rummy I was told, on this forum, that the FIDE system (the chess system) is being used here.

I'm not trying to be a trouble maker, but I think it's an important point. It's silly to say the Elo ratings given here are 'meaningless,' because the ratings are clearly correlated with the win/loss record. It's just that the correlation coefficient is not nearly as high as it could be and should be.

They are an indication of skill, just not a very good indication.

calc_guy

I agree
Posted by (VIP) 28 Apr 2004 10:17pm     


I would like to see the top players here listed by win-loss % b/c that is a long term indicator of who the top players actually are. A 1600 rated player who never minds who they play is most of the time a MUCH better player than a 2100 rated player who just plays other 2100 rated players. Sometimes double clicking on a high rated players record, I've actually seen a below .500 record! 'Top Players' should not be affiliated with 'Highest Rating' players. They are two totally different indicators and only a player whyo has the BEST record in here will get the nod from me as being a top player. Very good post Menk and hopefully Game Colony will take this positive feedback and implement it.

Julie

Sorry, have to disagree ...
Posted by calc_guy 29 Apr 2004 3:12pm     


... with your statement 'A 1600 rated player who never minds who they play is most of the time a MUCH better player than a 2100 rated player who just plays other 2100 rated players.'

When I first started here, mere weeks ago, I always invited everyone and amassed a 2 to 1 win/loss record, with over 700 games played. Then I started playing for ratings and invited the highest rated players first. My win/loss percentage went way down to something like 4 to 3.

There may be a few 1600 rated players that are better than a few 2100 players, but your statement is much too strong. In fact, you could turn it completely around and make a true statement: 'A 2100 rated player who just plays other 2100 rated players is most of the time a MUCH better player than A 1600 rated player who never minds who they play.'

calc_guy

LOL calc_guy
Posted by (VIP) 29 Apr 2004 4:28pm     


Loosen up a little! Why are you dissecting my post and narrowing it down to YOU and what youve done here in the last 700 games? What kind of sample is that?? Ok, so mayyybe by stating that most 1600 players are better than 2100 players is a little off... but this isnt an SAT test where If A then B and if B then C etc etc. And if some is true then ALL is true and on and on and on. LOL I'm not looking to be exact here and ok so maybe I was exagerating a littttttle. The point trying to be made is that Game Colony considers 'Top Players' as ones that have the highest rating and that is simply NOT TRUE (unless in the extreme rare occurance where the best record has the highest rating as well). A top player with the best record/winning % holds more water than someone who happens to have a high rating at the time.

By the way, my post and your answer to my post was never about you. So no matter whom you've played and what records you've amassed, the fact of the matter still remains that I've seen just as many 1600 rated players with the same record as 2100 rated players and vice versa. Bottom line, rating in here is so ridiculous that it unfortunately gives players playing for rating a false sense of importance and possibly skill. You happen to be in the minority as you seem to have both, but with all things being equal, your record holds SO much more weight than what your current 'rating' is today.

Game Colony, I would like to log on to the top player listing one day and not find someone with a record of 890-860 on top with a 2200 rating. I would like to see who truly are the top 20 players in here by virtue of their superior record, deserved over many games played here on your site.

Jule

Thanks
Posted by Menkman74 (VIP) 29 Apr 2004 6:23pm     


Julie, Thanks for your support. I echo your comments to the tee.

I disagree some more
Posted by calc_guy 29 Apr 2004 8:57pm     


Julie, I am PERFECTLY loose, and I am NOT narrowing it down to what's happened to me. I disagree with just about everything you've said in your last post.

Eg. 700 games is a prefectly good sample. It's not uncommon in statistics to make inferences based on samples as small as 20 or 30 (widgets, baseball games, whatever). More is better, of course. I was just presenting my case as an example, of what I believe to be a general truth -- that higher rated players are, in general, tougher opponents than just anyone off the streets.

I also don't agree that win/loss percentage is the sine qua non of rating methods. I guarantee you, if you invite higher rated players for a week or two, your win/loss percentage will go down, if you invite lower rated players, it'll go up.

If win/loss percentage supplanted the rating system, you'd have many people doing exactly that, just as they now 'play' the rating system, or 'play' the ladder system, taking advantage of the flaws in those systems.

The solution, as I've suggested before, is to fix the rating system, which is a simple process, and has already been done for football and several other games.

I'm dissecting because I think it's an important point, and one that needs addressing.

calc_guy

ok now
Posted by Rampage11 (VIP) 29 Apr 2004 8:59pm     


ok if u guys really wanna get into whos the best and all that junk, let me say this i can name 4 people in this site that chew me up and spit me out( no matter what rank or rating is) U NEVER NO IT ALL IN GIN read all the books all the %s u want. i fri my brain to beat these players but on the % they win ,to me I GET LUCKY and win 1 here and there, THERE skill is just better than mine, what can u do ? nuttin sit back and take ur beating and mayb learn a thing or 2. now for the next thing in my p brain, i also think people who play in tournys ,also r at a higher level as well, any 1 can play 1 game and win but in a tourney u play more than 1 and differnt peeps at that . i no people that r scared to death to enter a tourny room..... but r xellent players. o well love the dbate ... but just remeber this...... no matter how good u think u r , theres always sum1 better !!!! i myself am a 50 50 player i like to think im average or mayb a hair above,, o well peace out , and who is theboatman............. RAMP

y not try
Posted by Rampage11 (VIP) 29 Apr 2004 9:07pm     


top ten rating mungers and top 10 ladder spot holders in a tourny or may 3 tournys at differnt times , just to c the %s of wins and losses. makes me wonder, but so do pop rocks ...... any how just a thought , and those r few and far btween. RAMP

to calc guy
Posted by (VIP) 29 Apr 2004 10:16pm     


JUST WANT YOU TO KNOW I ISSUED A CHALLENGE TO CALC GUY AND WAS DENIED, I GUESS MY RANKING, MY WIN LOSS% AND MY ABILITY NEEDS SOME WORK. SO IM GOING TO TRY TO WORK VERY HARD TO GET TO LEVEL WHERE I CAN BE A FORMIDABLE OPPONENT. AND I KNOW CALC MAY READ THIS AND GIVE ME A CHARITY MATCH CAUSE HE FEELS SORRY FOR ME. I WILL OF COURSE DECLINE UNTIL I GET TO THAT LEVEL WHERE IM WORTHY OF HIS PLAYING AND TALENT. THANKS AGAIN CALC FOR PUSHING ME TOWARD THAT LEVEL. EVERYONE KEEP KNOCKIN AND GINNING AND HAVE FUN.
THINKORSWIM

What a Fabulous Discussion
Posted by Ed_Blue (VIP) 29 Apr 2004 11:52pm     


This is a tough one and I tend to side with Yankee/Menk and Think. I've been quite fortunate to have played all of you and my w/l percentage is admireable in these matches.

Let me state, all of you are wonderful gin players and its a pleasure meeting you on the field of battle.

I've been invited by the Calc_guy twice in the last week. Yankee_Baseball, Menk and Think. I don't thonk any of you consider me a pushover and I can state I am 50/50 with the Calc_Guy. He is a formidable player but he tends to select the 'R' function when he plays lesser tha 2000 rated players and that is not what I consider sportsman-like. Us mere mortals with mid 1600 ratings have proven ourselves over thousands of matches in 1:1 and in Tourneys.

There is no way to accurately measure our talent with a player logging less than 800 matches. Calc, you deal in statistics but the probabilty of the 'luck factor' creeping in to these matches is quite high. Just the other day, I was in combat with Menk, or maybe Think. I was well ahead and by some miracle he pulled a very quick gin and nailed me with 81 points on the 2nd or 3rd card. He was a amazed as I was but he won fair and square.

What I am trying to say is the talent of a player cannot be judged until he or she has played in a comparable number of games as their competitors. I ask GC to change the rules and not permit the 'R' function so that all matches are included in both opponents statistics.

There are a great many players, at lower levels, that 'seem' to be better than the very limited higher rated players that restrict their matches to only high rated players.

Hope this clears the water a bit.

Hey, thinkorswim
Posted by calc_guy 30 Apr 2004 12:33am     


You may well be a better player than I am. I didn't play at all for at least seven years before I discovered this site. I'm still getting used to computer cards. Hell, I throw away melds every other day. Couple days ago I threw away GIN!!

I used to be PDG, but I have a looooong way to go before I'm ready for money play again. It's more like TWENTY years since I played for serious money.

Even with all that, I still think I'm above average. I just thought it would be fun to play for ratings for a while, so I could say 'I did that.' I'll probably get tired of that after a few weeks and start playing for ladder, or win/loss, or who-knows-what.

Again, it would be foolish for me to play you -- a good player -- in a rated game while your rating is 400 points below mine. You would have AT LEAST a 50/50 chance to win, but would be getting odds of 30 or 40 to 1 on the ratings. Again, this is a flaw in the system.

Since the rating system isn't perfect, we definitely should be encouraging people to play with the -R option until it's fixed. I'd be happy to play you in an unrated game any time, Mr. think.

Ed, demanding that the -R option be removed would accomplish one thing: there would be a lot fewer games played between high and low rated players. Eg. The games you and I have had and enjoyed would never have happened.

I can tell you I do play rated games almost every day, and I promise you this: when I lose my rating goes DOWN, and when I win it goes UP -- without fail.

So far, my rating is still on a random walk with an upward drift.

calc_guy

ok, so disagree... it's fine, so be it.
Posted by (VIP) 30 Apr 2004 12:05am     


Calc, you can disagree all you want. Whatever it is, it is. i just tend agree with Menkman74 over you and there is no disputing that. His basic comment that judging the best players in here b/c they 'happen' to have a high rating is right on the money. I just went into another site (can I say (edited) in here w/out it being edited? (nope) well we'll see lol) and clicked on 15 of the higest rated players all around 3000. 9 out of 15 of them were BELOW .500! ya know what, regardless of that, you can give me all the statistics and inferences and 'sine qua non' and quotes on Pythagorus and ELO and whoever else had nothing better to do on a Saturday night all you want... but I'm just sorry to say that I will tip my Yankee hat to a 65% win player with a 1600 rating WAY before I tip it to a 55 to even 60% player with a 2200 rating. Change the rating system, I agree. It's broken and needs to be directly proportional to the deserved skill level of each person. Until that happens, again, Game Colony, please find a way to list the top player column as best winning percentage based on wins/losses, minimum of what I suggest to be 100 games.

Thanks calc_guy
Posted by webmaster 30 Apr 2004 9:17am     


Great suggestion on ratings!

You have convinced us that there needs to be rating variance throughout the games.

We'll add it to our job jar.

Thx,
webmaster



Oh Calc_Guy???
Posted by Ed_Blue (VIP) 30 Apr 2004 1:35pm     


First, hi my good friend and thanks for the reply. I'm sorry to say that you finally did it, you proved my point with your following statement:

Ed, demanding that the -R option be removed
would accomplish one thing: there would be
a lot fewer games played between high and
low rated players. Eg. The games you and I
have had and enjoyed would never have
happened

In the first place, I would never 'demand' anything from the great masters, GC, of this site. I just suggested what I consider to be an inequity based on my experience.

Secondly, I sincerely feel bad you have this opinion of the rest of us on GC. We play for fun and for money but the commonality is we enjoy the site and the game.

To isolate yourself by avoiding play with lower rated players is a defeatest attitude and I don't believe, in my humble opion, its the spirit of most other players on GC. Okay, I may be wrong but I do play with 1400 rated players so they have a chance to climb up the ladder or increase their ratings.

Of all the postings, I must say that your last posting tells a very sad story. If you need the recognition of a high rating I feel bad for you. You explained why you're not a money player but I also didn't play for 40 years, I'm 60 now, but I must have gotten something wrong because I did start playing for money and I think it adds a bit more edge to the matches.

Regarding your refussal to play the great Menkman. Shame on you. I've found him to be a delightful person and a great gin player. When you alluder to the fact his rating and/or w/l percentage is too high for you to tangle with, you've just told us all:

YOU'RE MUCH DIFFERENT THAN US

I guess all of us can make our own desisions as to whether we want to play against you or not. We have greatr matches and I can state that if the rating was turned on, I probably would have shot up to the mid 1700's to the low 1800's and you would have come down to the more earthly ratings of the rest of us.

This is not intended to be a negative Reply, it's just a statement of fact. For this reason, I am saddened. I did consider you in that group I call 'friends'.



I FORGOT ONE POINT
Posted by Ed_Blue (VIP) 30 Apr 2004 5:53pm     


The issue of the 'R' function is a toll that's obviously being misused or maybe not. I'd like to hear from the Webmaster regarding the 'intended' use of this function. It's obviously being used as a shield against faor play and a true competitve spirit.

From webmaster
Ed, I am in total disagreement with your statement here...
I have seen [unrated] 'R' option used by players who preferred to play for relaxation & preferred not to be pressured by ratings issue. How can you fault those who play for relaxation??


I Have the Answer !!!
Posted by Ed_Blue (VIP) 30 Apr 2004 6:01pm     


After all of this jargon, no-one's come up with a viable solution to the many problems inherent in the rating/ranking system. Until now, that is.

How about this idea:

Since it's inherently unfair to judge all players alike and the facts are apparent that the more matches you play the better chance of a lowerrating. Create an 'apprentice' category in which any player with less than a certain number of matches will be included. This category will only include those that played, say, less than 1,000 matches.

Look what this does, it eliminates the flash in the pan players that need early recognition as better players. Matches all loyal GC players to themselves limited to players if the same intention and qualifications.

It's only a matter of time before those 'high-on-the-hog, boy I like that, players get bored playing the same people over and over. Maybe then they'll join us guys and gals who've earned the right to be grouped as loyal and ongoing players in the GC family.

What a great, and very simple, solution to this obvious problem. What do you think?

This is addressed to The Webmaster and also the players that have played many, many matches.



Provisionaly rated players
Posted by webmaster 30 Apr 2004 9:30pm     


Those who played fewer than 20 games are currently rated provisionally.

These players are helped [via separate but fairly standard ELO formula implementation] to ramp up their rating faster.

This means that if they lose, their rating may either go down a little bit or not at all.

At times we see a reluctance of higher-rated players to play with provisionally-rated newbies. Just because newbies' ratings ramp up at a higher speed ...


Rating
Posted by (VIP) 30 Apr 2004 10:05pm     


As I stated in a diff gin forum, I don't play gin here. But IMO you have a choice who or who not to play, if you are 'rating' yourself against someone whom you think 'cheats' by playing lower ranked players or changes his/her name to get a higher rating, then you are just degrading yourself by putting him/her in your class. Lighten up play whom you want to play and don't worry about what the player 5 rows up/back is doing. Just my opinion

Tx

Ed, Ed, Ed
Posted by calc_guy 30 Apr 2004 11:53pm     


I would truly LOVE to believe your statement: 'YOU'RE MUCH DIFFERENT THAN US,' because more than anything, I value individuality and hate uniformity. I would really love to be different. Unfortunately, I can't buy your reasoning, because I see EVERYONE with a rating over 2000 behaving exactly as I do.

I have absolutely NOT -- never, nyet, nada -- EVER refused to play with Menkman, thinkorswim, or anyone. I have only refused to play a rated game at horrible, ridiculous odds. I am glad to play anyone an unrated game at any time, since I wish to educate people about the -R option, and encourage its use.

Why do you feel the need to criticise my approach? Right now, I'm going for ratings -- just because I've decided that would be fun for a while. I assume you play for money for similar reasons.

Suppose someone asked you to give 20 or 30 to 1 odds when playing for money and you refused. Then suppose they said 'shame on you' and called you 'isolated' and 'different.' What? You don't think that's a fair comparison?

What's really sad, Ed, is you seem to feel I must believe everything you believe, and play the same as you play, in order to earn your friendship. BUT ... if that's the way you feel.

calc_guy



Main Forum | Post Reply
Bookmark and Share    ...and Earn Free Tickets!
Play gin-rummy tournaments online

At GameColony.com you can play games of skill only -- play for free or play for $prizes!. According to the statutes of most states in the United States, gambling is defined as: "risking something of value upon the outcome of a contest of chance". (Also see No Gambling!).   The skill (as opposed to chance) is predominant in games of skill. Playing games of skill for $prizes, therefore, has nothing to do with gambling as it is not a contest of chance -- the more skillful player will win far more often. The chance element of a 'gamble' is either insignificant or missing. When players compete in tournaments or games of skill for $prizes -- it is "competitive entertainment" rather then "gambling". The more skilled winner will always win more matches, tournaments and $prizes.
Affiliate Program

Copyright © 2024

Site map