Chess rating system

You can't just start people high... if everyone started 500 points higher, then eventually everyone's ratings would end up 500 points higher, because you'd pick up more rating points from the new people whose ratings haven't been lowered enough yet.  If you start people out at 1700, then the ratings system will adjust itself until the average new person IS a 1700.  And the people who used to be 1700 will now be 2200, etc.

Using official ratings from USCF or other bodies as a starting rating would probably be fine from a ratings standpoint, but how would you prove that you are who you say you are?  Premium membership using a credit card with your name on it?  Obviously they verify the titled players already, but it seems like it would be a pain for the staff to have to do it for everyone.

And as far as the chess engine goes, if someone's using an engine then they're going to be out of your ratings range pretty fast.  Assuming they don't just get caught and tossed from the site.

I agree that discipline is essential. You make a mockery of discipline however, when a very disciplined player, through being disciplined, attains a level of education and play.Then you mock his intellegence and discipline by treating him like less than he is. A good norm for both parties would be to treat them as they are instead of your own biasedly leveled playing field.

I have to admit you make a good point about the ratings starting high. I guess thats my own emotional biases kicking in and over compensating. I agree also that it would be a headache for the staff. They might as well be subsidiaries of the major sanctioning bodies.

However, that line of reasoning doesn't quell my hunger for fairness and justice. If you are sated of your desire for these things simply because, it is said that life isn't always fair, then we might as well through out the rule book altogether. Oh wait a minute, that would include the structure(rules) of the game. Therefore, I am defending the integrity, structure and discipline of the game I love. I am not doing this simply because I am upset that I think I deserve a higher rating and don't want to do twice the work,or oh wait, maybe three times while I add in the loses to cheating.

Believe me discrediting the cheater is only the first step. It doesn't fix the damage done to the 1700-2000 players doing it honestly. You might have overlooked how many of these a predator must first devour, before a legit player with a GM/IM type elo will even gime them a game. Then how many more does it take before they are found to be fraud who won't play a live tournament.

This is why in general I have submitted a few examples of why I think some form of reform would be welcome. I would love it if someone would come up with a better system for online play and ratings in particular. Yeah yeah you don't have to reply. I already know.

  I agree 100% !!!

I LIKE PIE!

 hi! u didnt tell about  chess!! appriciate ur pie favouritism!!!!

I try to stay away from this thread but I can't :)   Much earlier in the thread I was involved in a debate about the merits of allowing people to supply their rating from another site or organization as the starting rating (I believe it should be permitteed and encouraged).  The main rebuttal to my position was that each rating pool is independent, so why try to inter-relate them ("it doesn't matter").  My reply was that if it doesn't matter, then let me start with a rating I supply.  In the comments quoted above, there was a mention of needing to verify the rating that is submitted.  I personally don't think that would be necessary and agree that it is impractical.  Sure there is a risk that someone would enter a false rating, but I think most people who would take the time to enter a rating would enter a "real" one.  This has to be better than arbitrarily assigning people a 1200 rating to start. 

You can be good at different types of chess games, and you have different ratings in each chess type. It makes sense that you are better at longer games than shorter games, more than the other way around. You can be good at short games by being a pawn ahead, and then ice the person you're playing. You might play short games only to do that, and have a good rating in that type of chess. In long games though, you can't really ice your opponent too well, therefore you'll lose a lot, and not have a very high rating in that type of game. But, the ratings from each type of chess you play does not combine to just form one rating, they stay in separate in whole different ratings that have nothing in common.

I don't follow the rationale but your point is correct.  First, in absolute terms, people will be better when they have more time to think.  Second, different people have different relative strength depending on the time control.  I've done fairly well here in turn-based chess but I get killed in bullet games (and I'm not much better at blitz).  I have theories why this is, but that is off-topic.

By the way, how the subject of the relative value of the pieces ever worked its way into this thread is beyond me. :)

To the first point, that is why even within Chess.com there are several different ratings depending on the type of play (Live, Blitz, Online/turn-based, 960, Tactics Trainer, etc.).

 ....agree!!

I feel that a lot of people have the wrong idea about ratings and it's purpose.1200 is not a bad starting point and you have to start somewhere. If you are pretty good and play someone with a say 1700 and win, then your rating will improove very quickly. It's an encouragement to do well. The more games you play, the more accurate your ratings reflect your abilities. If you start with say 1700 and keep loosing to similar rated players, then your elo will drop and level out until you can hold your own against like rated players, eg. you may play against 1400's and win some then loose some and win some again. After you've played a few hundred games, your ratings should reflect fairly accurately as to how good you are. As soon as you start a different chess format, you start from the beginning again, until you've reached your normal level. As far as cheats are concerned, ie. people who use chess engines and the like, I think chess.com has a very good way of determining this. From what I understand it has to do with percentage moves. eg. if a player of a certain rating plays consistently at their  level and all of the sudden keeps beating players rated much higher, then there is probable cause to investigate his or her games. Most players make moves with a variable percentage in strength, even world champions. A lower ranked player may make 20%, 30% and maybe the odd 40% move. A higher ranked player may make 30%,50% and maybe some 60% moves. A chess engine makes perhaps consistent 60% or 80% moves, but they do it all the time. Even the best players in the world aren't that good. I understand that chess.com kicked over a thousand people out last year for cheating in various ways. They can never play on this site again!                                                                                            I get a lot of enjoyment out of my chess, win or loose. Sure, I don't like loosing to a lower rated player, but you learn to loose OK, pick yourself up, learn from your mistakes and go on to the next game. The next game you could beat a higher rated player, it not only makes up for the loss, but also gives you a sense of achievement.   So enjoy what you do and have fun.  :D

One other common occurrence on Chess.com is players losing due to timeout.  In many cases the culprits are strong players who over-commit and are unable to keep up with the volume of games.  This has to have some effect on the integrity of the ratings, but a smarter statistician than I will have to explain what that effect is.

I haven't kept up with all the comments, but I think I saw somewhere that time management is also part of this game and it holds quite true. eg. if you play OTB's in a tourney you are given very specific time controls which equate to timeout's online. So you may have a winning position against a lower rated player, but may very well loose on time, for which your ratings then drop accordingly. This applies in particular to 'lightning' and 'super lightning' chess. Of course if you take on more games than you can handle on line, then you may timeout on some, due to other commitments in your life. The other effect that a larger volume has, is that the quality of your game may drop somewhat and I know that from my own experience. So take on, what you think you can handle and adjust the volume of your games to your time availability.

wow, idk what to say...

im just posting to get points lol

This is advice I am taking myself.  As I finish my games, I am putting more time into the ones I am still playing, not taking on any more games.  Once all my games are finished, I wont take on more than 3 at a time, which is comfortable for me.  I can play them with all my strength and much quicker too.

i like the rating system. 

the more you play the better.

 it's ok.!!!!!!

Iplayed a 1798 guy at 960 and in a totally losing position (being about 2 pieces down) he was banned from the site for cheating and I got about 155 rating pts.(althoughg I lost them later)))