Two Sites, Different Rating

my tactics rating here is also about 200 pts. below my turn-based rating.  i think that it the "best move" thing in tactics that makes it lower.  time is also a factor, but in a game if you miss a mate in 2 for a mate in 3 instead you still get the points.  of course, between this site and the other site, how knows how the ratings compare. 

I think chess tempo is fairly accurate -- of course the problem's ratings are based off of pass or fail of users, but at least for me it's pretty close to my actual strength.

I don't use chess.com's (I don't like timed) but the one's I've heard talk about it say they're inflated here.

My two cents.

In general ratings strongly depend on the players pool. Just for arguments sake let's assume we have two chess sites using same rating system and with the same number of players but the players on the other site are mostly masters and on the other site mostly beginners. Because the sites have same number of players you would expect the rating distribution on both be about the same. But if you play on the master site you'll probably loose most of your games and your rating will be very low whereas if you'll play on the beginner site you'll often win and your rating will be high.

I understand that we are not speaking about games but tactical training here. However, in order to rate your training performance there has to be some estimate about the difficulty of problems. This estimate is most probably obtained by observing how well users perform solving a particular problem. So, the players pool will have similar effect to ratings as with normal game ratings.

Now, it seems to me that chess.com has larger percentage of inexperienced players and even complete beginners compared to many other sites. Therefore it's not too suprising really that many people have higher ratings here than on other sites.

I have done 9800 on this site in5 month and I am starting to see more and more. before a 1600 task was difficult to me. Now I am at 1808 my record was 1975. I have 47.1% correct.

I guess this site is pretty accurate.

I think the reason why you are higher on this site is that chess.com TT will give you +40 for solving a puzzle in the start.

Now I only get something like +12 and -18 if I am wrong.

Actually I take back what I said about chesstempo -- compared to my real rating, chesstempo is inflated as well... not a big surprise or anything that the two don't compare :)

1. Tactic trainers are meant to be a study tool. The rating there has the main purpose of selecting the right problems you will receive.

2. Is the other site emrald? If so, your 1500 rating there is very good!

i agree that TT is only a tool and has nothing to do directly with your chess rating.  at the same time, the chess.com rating works the same way and has little to do with your OTB strength.  i would think that live would be a closer indicator, but as many have said, it is really only to compare to people on this site and not to FIDE or USCF or any other site. 

One of the main purposes of the online or live ratings on this site is also to allow finding suitable opponents or selecting suitable players for a tournament etc. But that doesn't mean that you can't use them to gauge your progress or general standard relative to other users. It's the same with tactic trainer ratings.

The important thing to understand is that the rating only measures how well you do compared to other users inside the same rating pool. Therefore you cannot really expect that your rating here for example would be same as USCF, ELO or whatever as both the players pool and, in these cases, also the rating system are different.

Finally, tactics ratings only measure your ability in one (altough very important) part of the game so it's only natural that for some people they correlate more closely with their game ratings than for others.

Sure there is a correlation between your playing strength and your TT rating, but that is not the main purpose of the TT ratings. For the site I told you before (emrald), the rough conversion would be: >1800 -- FM, > 1900 -- IM, >2000 -- GM. Compared to the ratings of other players of similar strength with me, I can say that the rating I have there is quite a good indicative of my strength.

For the TT on chess.com I didn't solve too many problems to get a stable rating and I don't know what would be the correlation with real world. It's also annoying that some of the problems here have multiple solutions, forcing you for instance to find a tricky mate in 4 when there is a simple mate in 8 let's say. But the TT here has more difficult problems, even studies like this one:

http://www.chess.com/tactics/server.html?id=86411

I was told by the staff that saying "I play on  ..." or "I read ...", with specific names, is not considered advertising. However, "go to play at ..., it is the best site in the world, it has these features: ..." is advertising.