Major Proposed Change: Ratings Cap Above 2400 for Unverified Players

Sounds good.  I hope it works.  And I take it the 2400 cut-off is due to the FIDE classification as master (2400)?

Interesting.

Forgive me if I am wrong (my computer knowledge is limited), but wouldn't IP bans work just as well?

Hmm, possibly the biggest problem could be a whole bunch of cheaters caped out at 2400 :)

HOPFULLY that means they're easier to find... if it gets ignored to where there is a big spike at 2400 on the distribution curve I'm going to puke :)

Otherwise I think it's great.  Titled players get membership for free right?  So the only people this hurts are near-master (or master) level players who never got the title and can barely break 2400... this has got to be an incredibly small number.

Meanwhile the leaching group of cheaters takes a big hit.  I really like it, hope to see this implemented asap :)

I like the idea of this change, but I'm not the pros outweight the negative effects. There are a handful of honest players in both correspondance, and especially bullet who are very strong players but untitled and on basic accounts.

2400 is a pretty good cap, but those that are ~2500-2700 are going to cause a bit of rating deflation because they can consistently play above that level, even against titled competition.

Though a gold membership is very inexpensive and comes with good benefits, I don't think this is the best way to go about things, but overall we'll have to see what impact it has on the system after it's applied.

EDIT: deepgreene makes some pretty valid points. I suppose if someone cares enough about their rating to be discouraged by the cap, then it shouldn't be a very hard decision to go for membership and make use of the extra benefits as well. If they just want to smash people and don't care about cheating, then I can think of a few sites off the top of my head where they might be better off. However, I do know one individual in particular this will affect and the tradeoff between relative anonymity and rating cap might be an interesting dilemma.

I agree that this is a bad idea. You could just ban the cheaters' computers.is over 2400-strength, but doesn't have the money to achieve a master title?

I thought there might be some sentiment this way... that players can pay to make their cheating more effective.  That's not too accurate, this group of cheaters will move on to other free sites where you can cheat for free rather than pay. 

I think under the circumstances it's justified.  10 staff can't police 50,000 players (or however many active users they have).  I don't feel like it's "the ends justifying the means" but more like the good is vastly outweighing the bad.

The bad is players like you who are (or will be) close to the 2400 line and are untitled.

The good is as deepgreen points out, a populous and leach group of cheaters gets hit hard.

The point is chess.com's resources are limited so I think this is an honest and good move given the circumstances.  In all honesty it's quite apparent that chess.com can't control cheating.  There are a few 2800+ players who post on the forums (that [named deleted by staff] guy I think his name is, and [name deleted by staff]) who are obviously engine users and don't get banned for [what seems like years]

BTW your 1850 ICC bullet rating makes me feel like I'm not so terrible :)  What I mean is my USCF rating doesn't necessarily have to match my ICC ratings.

For what it's worth, the motive for posting first in the Cheating Forum was to elicit feedback first from a group who obviously thinks a lot about this problem (as the solution is still something of a work in progress).

Anyway, like some of the restrictions around the Chatrooms in Live, this sort of initiative is clearly something we see as more necessary than desirable. And to the extent that it discourages high-level cheaters & improves our credibility with truly excellent titled chess players and coaches, it's good for everyone.

Measure the benefit of seriously discouraging aggressive repeat-offenders against the offense taken by this group:

• Honest, awesome chess players (>2400), who...
• Have no chess title, and who...
• Don't think a Gold Membership at Chess.com is worth the money, and who...
• Don't mind playing a significant # of their games vs cheaters

...and things clarify pretty quick.

As long as I can find challenging opponents, the ratings don't matter to me.

I kind of get the whole idea behind this. Like DeepGreene said, it's definitely gonna bother the large amount of cheaters more than the rare 2400s out there. Anyone 2400 in blitz or standard is either a badass or a cheater, so that should be easy to detect. Although, the only place this would really occur commonly would be bullet I imagine. The problem is that anyone 1800+ in normal time controls could easily hit this in bullet with the inflation here not to mention the people who don't resign and the fast mouse movers. I wonder what motivation there would be to play a game once you reach 2400 though. Just a small problem. 

Something to consider:  When players reach the 2400 cap, run your cheat-detection algorithm and remove the cap if they appear to be clean. The population of 2400+ players might be small enough that this is feasible.

fezzik, you say this is announcing we can't control cheating. to some extent, yes. we can't control cheating if every time we catch and close an account someone just opens another one and continues cheating. that's pretty obvious though.

but if we can cause those cheaters to incur a cost, then we'll be able to control them.

the suggestion of ip-banning them doesn't work. any other suggestions?

I'm 1820 USCF and don't think class A range players have much of a chance to break 2400 in any category here.  I think it will mostly be strong experts like yourself who are facing this delima.

"A bunch of players will be butting up against the ceiling"

"A bunch" ?  You yourself said you're not quite there yet, so who will make up this bunch?  I'm curious how many untitled near master to master level players do you think are on this site?

I don't think this will be perfect, but I do think the good really outweighs the bad.  I suppose IP bans aren't a go to because public computers or families with multiple users?  I'm also curious if you have another suggestion to help the cheating problem?  This one seems pretty good to me.

I think the 1800-1999 range of players (rated by a federation, not online) would find it very hard to break 2400 here even on bullet.  It's true some bullet ratings are inflated many hundreds of points, but not as much as 500-600 yet :)

What do you do then, with the players at 2400? Look through their games and see if they are cheating?

I thought that chess.com does not have any cheaters. That is the official position, no cheating is mentioned anywhere. It does not exist! When I wrote once that, in my opinion, ICC has fewer cheaters, my post was deleted. There are no cheaters on chess.com, all the cheaters play on ICC.

So, why bother?

Good point. I doubt that every cheater out there is really going to care that their rating is frozen at 2400. Some will, but probably not a majority.

Anyway, I have another idea, but it depends on how expensive running the algorithm is. My idea is running it (automatically) on new members' games (Until 25?) and knowing that player's established rating. If that person has a mysterious rating hike (more than X in Y amount of time) then run the algorithm on some of their more recent games.

we can take measures to deflate the bullet rating pool (and any others) when they become inflated, so this will continue to only affect a small portion of players.

  It's no problem to me because I don't play LIVE or Online that much, and I am many years away from 2400, but as said it doesn't matter to Premiums anyway. So why am I posting this?  I don't know.     

  -Mark

I'm a high Class B USCF player and I've been up to the mid-2300 range here in CC.  I don't know if I'll ever make it over 2400, but this change bothers me immensely.

My only other idea at the moment would be to have something based on membership time - i.e, you must have been active for more than 3 months to break that rating cap.  That doesn't quite seem feasible (though there may be a way to tweak it), but I don't want to deal with a rating cap.  CC often becomes my main chess focus when I'm too busy to travel to tournaments, and I'd like to keep that option open.

shoop that's actually an interesting suggestion as well.

i should point out that even if your rating hit the cap, you could still keep playing chess here. so you should still be ok. why does this bother you immensely?