Tactics Trainer Math?

In the Tactics Trainer, I typically solve about 74% of the problems, yet struggle to keep my rating just above 1000.

Then I see that some people solve only, say, 50% of their problems, yet maintain their ratings way above 2500.

Can anyone explain the math that makes this possible?  The way the Tactics Trainer scores things, it seems that a 50% solution rate would definitely make your ratings decline.  What am I missing?

Thanks for your help !

Amy

With many of the harder problems they are many moves. Say you get 5/7 moves correct. You will be marked as have gotten the problem wrong, but probably have lost minimal if not gained some rating pts. 

Thank you so much, bondiggity !

this should be about the time you spend on your problems. those 50 people are probably very quick and get close to full points for each problem they solve correctly.

out of topic, but I think trying to ignore the rating and going for accuracy is the best way for developing tactical skills. but for that to work, I mean to comfortably ignore the rating, you need to pull the accuracy up to 90% something.

Yes, I am amazed by the ability of some people to look at a position and then more or less instantly find the solution, especially since the Tactics Trainer presents the position out of the blue and not as part of an evolving game.

I would dearly love to move my accuracy rating higher, but I can't seem to stop myself from just "jumping in" in an attempt to grab those bigger points.

that's why I think the balance between time and accuracy in such rating systems are way biased towards time. Having a 50% will simply not improve anyones tactics if not make it even worse. that's having one problem wrong in every 2. To be honest, I don't even understand how people enjoy such practice. you must absolutely hate making mistakes if you want to get better.

and between you and me, the "standard" training mode in chesstempo.com only measures your accuracy so you can forget about the time you spend on the problems. shh.

@philidor: The top players on the Tactics Trainer do not get 50% completely wrong. Like bondiggity said, the solutions often consist of many moves. You can get 80% of the solution right, but it will count as 0% correct for the pass rate stats.

Furthermore, the higher you get, the harder the problems get. It's similar to playing chess. Would you say that players like Ivanchuk perform badly? He plays against 2700+ players most of the time, and scores like 50-60% in total. But that doesn't mean he doesn't play amazing chess. The same happens with these problems: if the server throws only hard compositions at a 2800+ player, he may not get them all right, but solving 50-60% of them perfectly is still a great accomplishment.

Also, the tougher the problems get, the more time you get to solve them. So then it's not about speed at all, but about being accurate. Yes, the "simpler" problems involve less time, since they should also be solved faster than those harder problems. Of course accuracy matters more than speed, but is it really an accomplishment to find a simple fork after 5 minutes of thinking? If you want to improve, you should see those moves instantly.

sorry phobetor, I see your point but I don't agree with you about several things:

I don't think there's a difference in having 80% of a problem and 0%. the way I see it, you should have a perfectionist attitude towards tactical training and if you don't force yourself to figure out everything before making your first move and have a "well let's play this and we'll take it from there" attitude, it's going to be less efficient. much less actually.

I could never say having a 2800 rating with ANY accuracy is not an accomplishment. of course it's a huge accomplishment. I also don't think the case of Ivanchuk is very relevant. in a tactical problem, you KNOW there's a winning line. in games however, it's an equal fight. so having a 60% consistently in top tournaments will make sure a spot in top 10, and is amazing.

I never argued about such thing. let me rephrase my point: I think throwing several rating points (maybe a few hundred, even) and pulling your accuracy up to 90% from a high rated 50% will be more beneficial for overall tactical improvement.

Amy, play a correspondence game with me , please !!!!