Major Proposed Change: Ratings Cap Above 2400 for Unverified Players

any plan that involves checking for cheating again and again is not really helpful. so just saying hey, this ip might be a cheater doesn't change the fact that we wasted resources analyzing and banning that member once before, and we'll know have to do it all over again.

ghost players: you said they would not lose points if they lost to the ghost player. but you also said people wouldn't know who the ghost players are. but if you finish a game and lose without losing points, you'll know it's a ghost player. also, no way to force people to play against the ghost player. also, the ppl who lose to the ghost players will submit cheat detection reports, and be upset about those games!

finally, the main point is that people out there want to play against people. introducing more computer engines into the pool of players won't help that.

Could you introduce a mandatory special fee for untitled players who break the 2400 barrier?  That seems like it would solve the problem.

that was the original idea, but just requiring they purchase gold membership works just as well with less development effort, and then also gives those members extra benefits.

$5 / month? 

I disagree. They can change their 'names' after each game. Also, if you loose one game, so what? People don't submit cheat detection after loosing one game surely? That is ridiculous. Otherwise you would have MILLIONS of reports each day....do you? Surely not! Exactly. I win. ha

So you try to catch cheaters by effectively creating "sanctified cheaters"?? 

Clever, clever. I can not imagine anybody being the slightest bit mad at losing to a program, as long as it is created by the "system." 

:D sorry if this sounds harsh...

Would this change be applied to bullet ratings too? I think that the rating inflation on there is such that many 2000+s with fast mouse fingers can easilly break 2400. I should know from experience :P Plus is it even possible to check moves against an engine in the space of a 1min game?

I personally would welcome the change at blitz and above though, the chances of breaking 2400 on that without being FM level or above is slim.

Players could give a proof to chess.com that they have a consistent Fide rating for their 2400+ here.At that level you don't play chess as a hobby it's a lot more than that.I have no problem that they play Really well..just a proof.

IMDeviate-> "Thing is chess.com's turn based ratings are roughly 400-600 points higher than they should be" 

Do you have data to back this up?  I've never ever played in any USCF or FIDE tournament, so I have no idea what my "real" rating is, I thought the general consensus was around 250-300 pts higher? Not that it matters that much in the end, but my 1600 here looks really bad if you're telling me it's only worth 1000(barely class E). 

OK, back to the discussion on hand: This shouldn't affect me since I'm not realistically planning to reach the online equivalent of Expert (Which I thought was 2300 but Deviate here seems to think it's ~2400-2600) until I finish grad school, but surely there are better ways to go about doing this than making people effectively pay for ratings?  

An equivalency test, interviews, etc seems impractical, but how about, as somebody mentioned, that somebody with six months or more on this site could "provisionally" get their rating cap lifted? The way I see it, if you're not a titled player, surely it'll take you more than half a year to get to a 2400 anyway, no?

no draconlord. because of provisional ratings you can hit that ceiling in a day or two.

also, as i responded earlier to this suggestion, i have thought of a way for cheaters to get around the 6 months on the site issue.

it's a false concern that this will affect many people who arent cheaters. there are going to be very few members who are strong enough to make 2400, but don't have otb titles, and are not interested enough in chess that they want to be premium members on here. various rating pools sometimes get inflated, which would make it possible for eg. a 1600 to hit 2400, but as that happens we can notice and take steps to deflate it, so that again, a 1600 has a rating of 14-1800.

You won't loose points; just a game. Man up, you gain knowledge from every loss.

I suggested a solution here, I see that it was discarded easily. What the staff proposed is wrong: I don't want the same idiot to make an account, cheat until getting to 2400, then make another account, get that to 2400 and so on. Not to say that, if the staff proposal will be implemented, some people will think "on chess.com you can cheat provided that you pay". It's ugly and wrong.

But anyway, I'm not too interested in what chess.com does. I will never buy an account here until they will have a reasonably effective cheating detection policy.

here's the deal:

there is ZERO cost or barrier to somebody creating a new account, loading up their favorite cheating program, and cheating their way on up. it's free and easy to do. the only downside is that eventually they get caught (http://www.chess.com/forum/view/community/chesscom---list-of-caught-cheaters ) and they lose all of their work. but who cares - they can just do it again. 

there IS cheating on Chess.com (and all chess/gaming sites). but it's not as big and scary as some chicken littles would have you believe. but the problem is that it does impact legitimate players at the top, and we want to respect them by creating the cleanest possible playing environment. 

the reason that we are proposing the cap is that now instead of just catching cheaters AFTER they cheat, we can now dissuade them BEFORE they start (or from doing it over and over). 

first, it creates a frustrating barrier for cheaters. now they can't just quickly move up and then get their giggles by taking down legitimate strong players and titled players. they won't have access to those members. secondly, if they decide to pay, they are then running a financial risk and disclosing personal and uniquely identifiable information that is more than just an easily changable IP address.

as for the artificial lump at 2400, well, that gives us a nice and easy place to begin looking for problems. any potential ratings impact this may have on the players as a whole is minor when compared to the other impacting factors (people timing out, cheating in general, etc).

as for Chess.com ratings being inflated, we actually ran statistics on this not too long ago and found that our ratings were slightly high, so we made a change, watched it work, and then re-ran the study to find that on the whole we are NOT higher. of course this isn't true for all members, but averages across all players. 

finally, please remember that this only impacts an extremely small number of players. only ~1,300 people have ever been over 2400 in any one rating. 

the way i see it (and correct me if i'm wrong), there are only a few different kinds of players who can hit 2400+.  

titled players: these are players who play in OTB events and earn recognized titles. they have clearly demonstrated that they are capable of 2400+ ratings. 

cheaters: people who get their kicks "winning" artificially. it's pathetic, but it appeals to a sad, small slice of humanity. 

very strong non-titled players: while it is difficult to become very highly rated without playing in OTB tournaments, it does happen (especially as internet chess grows and OTB chess shrinks)

now there are two kinds of those strong non-titled players:

a. those who are premium members because they like chess.com and all of the features. (you'll notice that many of them fall into this category already: http://www.chess.com/echess/players.html )

b. those who aren't premium members yet, but who feel the benefits of having a more vetted top-tier of players is worth the cost (about $2.50 a month for a yearly gold membership).

c. those who are strong players, but who don't want to pay any amount of money for any reason. 

this last group, c, is tough. to me it falls under the "you can't please everyone" clause. do we, in order to please this group, NOT use this tool to combat cheating? or do we just have a few potentially frustrated non-paying strong players so that we can better serve the titled players and paying premium members?

thoughts?

+1

All of this fuss over the cost of just one gallon of premium gasoline? You gotta be kidding me. If 5 dollars will make ALL of the ratings at chess.com 100% accurate, I honestly can't see why the highly skilled and serious Expert level chess players would be even in the least bit upset.

Expert players - Your ratings will increase over time, since you won't be losing to cheaters/computers ever again. Also, you will be closer to the top of a true and accurate ratings leaderboard. +100

I have a membership to ICC, and the program (Dasher) one needs to download to play chess, watch videos and do other stuff is awkward and weak in several areas.

Chess.com has the best user interface of any chess site that I belong to.

I was once competitive in another online game, with a top ten ranking in the world, and in one category, I was ranked number one. Guys at the top really hated ANY kind of cheating, because we were the ones who were directly affected. Cheaters crush the spirits of the honest competitor, and some of the good guys end up quiting.

--------------------

Edit - I just read Erik's post.

$2.50 a month for a year subscription? $2.50!

Erik - You are too kind. Make the stubborn grumpy guys pay the couple of bucks to make their ratings legit. Good grief!

Maybe this new policy also calls for a ratings system that makes it harder for 1500 USCF guys to reach 2400+ on their own

"If 5 dollars will make ALL of the ratings at chess.com 100% accurate"

this policy only makes the shown ratings more unreliable

You probably want to please premium members who pay the money you use to pay your staff and other expenses.

"but it's not as big and scary as some chicken littles would have you believe"

Those chickens include all the GM's who regularly play on the other site, even though they are not payed? Or the staff member who plays his games for the live session videos in the same place? :)

Erik, if the group of strong players who do not want to pay anything is very small, then why not deal with them on an individual basis. You could even quietly grant them a free gold membership as sort of a bonus for their strong play and long commitment to playing on this site. 

Llawdogg

I think that a more reasonable filter would be to allow more veteran accounts to exceed 2400 regardless of premium/title status. It makes the idea seem less greedy.