I just started using the tactics trainer, and I find I'm enjoying it in general. One thing I don't understand, however, is that I often lose points even when I solve a problem correctly. I am assuming this is because I took "too much" time? I find that unless I answer within 60 seconds or less, I lose points, correct or not. Since it's not unusual for me to take 2-3 minutes on a problem, I find my rating continually dropping, although I have a 75% (9/12) success rate. This means that I am basically going to ignore my rating and just focus on my success rate. In my opinion, it seems ridiculous to lose points on a problem for taking 2-3 minutes on it, when 2-3 minutes is NOT an unusual amount of time to spend on a position in OTB time controls, and esp. if you're just thrown into the position (not in the middle of a game). I always take 20-30 seconds just to survey the pieces, pawn structure, material count, etc. to get a positional feel before even looking at tactics. I also follow up my decision with 15-20 seconds of "sleeping on it", just making sure I didn't miss something, looking for in-between moves, checks, etc. This is how I play OTB. Unless the focus is on speed rather than accuracy? I guess this might be expected, given no one online seems to have patience for a 15-20 "standard" game, (unless it's correspondence). Just my 2 cents.
The time factor is the likely cause, and it's based on the average time it's taken others to solve the problem so it's appropriate that it be included in the measure of success.
Incidentally, I'm pretty sure the problems are rated in the pool alongside players, and their rating adjusts every time an attempt (successful or otherwise) is made just as the attemper's rating is based on the outcome.
Grobe said: "it's based on the average time it's taken others to solve the problem"
Do you mean it's based only on the average time taken for those who SOLVED it, or average time of everyone? This is my point/objection. Let's say 30% of people correctly solve a problem. But that means 70% failed. If everyone is trying to make a choice in less than 60 seconds, because they're going to lose points otherwise, then of course the time it takes the average successful person will be less than 60 seconds, by definition.
Here's an analogy: Suppose everyone just played a random move in 5 seconds. If there are 20 legal moves, then only 5% of the people will "solve" the problem. But the "average time it's taken others to solve the problem" is 5 seconds! My point is, the people who miss it should be given a penalty like 5 minutes, which is then computed into the average.
I'm pretty sure it only takes successful attempts into account (and successful first attempts, at that) but someone from staff will have to confirm.
You can turn off the timer in TT options. Apparently TT is used to prepare you for blitz games, not analytical playing at leisure.
it deducts points when you spend too much time pondering about how a "tactical puzzle" can be so dumb.
MDOC said: "You can turn off the timer in TT options"
Okay, but will this affect the way ratings are calculated? If the time spent is part of how ratings are calculated, it would seem all the problems would be unrated then.
Yeah, that's a good question....
Seems to me your TT rating is fine, it's your online rating that is overinflated.
Um, I actually don't know... haven't tried it.
I have pretty much the same complaint about TT. I had no idea I could turn the timer off. I am going to try it that way.
Oops. Now I see that you can only turn the timer off if you are a premium member, and using TT without the time apparently sets the ratings off entirely. Might as well just ignore the ratings.
You can hide the timer - everything is still calculated as it would be if the timer was shown.
And you can turn off the timer, which is unrated. So random puzzles, I guess.
I wasn't aware either of them were premium features though.
This is a big reason prefer chesstempo.com for tactics training.
e4nf3 said: "Seems to me your TT rating is fine, it's your online rating that is overinflated."
Because we all know 5-minute chess is the best indicator of chess ability.
And I think you mean "inflated", not "overinflated", which is redundant.
I trust that more changes are forthcoming with Tactics Trainer.
Actually, there is a difference between inflated and over inflated.
A balloon needs to be inflated. Just don't over inflate it.
Many online ratings are inflated. But some "take the cake". This would be an example of over inflated.
It's worth noting, though, that it is possible to take a cake that is just regular-inflated: