Why does TT deduct points when I solve problem?

http://support.chess.com/Knowledgebase/Article/View/53/0/how-does-time-affect-my-tactics-rating

Here it says that it is the average time to solve the problem. That would mean that it does not include the times of those who failed.

diogens said: "This is not exact. If you got your 20 problems wrong, instead of earning 3 points you would have lost, let's say, between 150-200 points."

Well, that's not the comparison I made.  Of course, if you miss 20 out of 21, you'll lose lots of points.

I was thinking about comparing to the "average player".  This "average player" over these 21 problems, would answer 56.8% correctly, or about 12 problems.  Since each of these would be done in "average time", each would generate a little over 0.8 "points" (the score put into the Glicko formula, which is given by a graph on the link above).  (This is an approximation, since the function appears to be slightly concave down.)  Altogether, this average player would earn roughly 10 points in this "21-game match" vs. the TT.  Since the ratings of the TT problems are on par with the average player's rating, what do we expect?  Well, what do you expect if two players with roughly similar ratings split a match (11.5-9.5, or 11-10, or 10.5-10.5)?  You expect their ratings to remain practically unchanged.

Well, this is what happened with my results.  So, according to the TT ratings, my performance is on par with the average player.  In other words, taking an additional 27 seconds per problem is as bad as answering half the problems wrong.  (Or equivalently, saving 27 seconds per problem is as good as getting all the problems right.)

"Earning 3 points will allow you to have a similar roster of problems next time until you get quicker.... I would say that chess.com TT ratings are focused to longer (than blitz) time controls. "

Maybe longer than blitz.  Not longer.  Certainly not OTB standard time, which gives 2-3 minutes per move.  I think the reason chess.com TT ratings are focused to blitz (or rapid, at least) time controls is because that is the way the players are reacting to them.  Without a penalty (or a significant penalty) for incorrect answers, there is little motivation to focus on accuracy, so "average times" (however computed) go down, and pass rates go down, and these both reinforce the same reaction.

The reason I go on and on about this, is that (besides wanting to offer constructive feedback to the site), I find it very interesting that two of the most fundamental and well-accepted principles of educational psychology are completely ignored in designing such a training system:

1.  Students will alter their behavior to maximize rewards and minimize punishments.  Whatever is rewarded, that's what you'll get more of from the student.  This is the whole idea behind grading.  Anyone who's taught for any length of time knows this.

2.  "Practice makes permanent, not perfect".  In every other endeavor of learning, stress is always put on accuracy first, then speed.  Music students don't fly through their scales and exercises at top speed month after month, year after year; instead, they practice very slowly until their technique is mastered, then they gradually increase speed.  The same applies to other games, from poker to basketball to academic subjects.  Only people who design chess tactics trainers seem not to know this.

When you ignore both of these simultaneously, you create an environment where old habits and mistakes are reinforced and repeated, and true improvement stagnates.

Have you noticed how many pathetically easy tactics problems (hanging piece, obvious mates in 1, etc.) are rated 1400-1600 and yet still have 40% of the people miss them?  What do you expect?

"They are also telling you that if you don't resolve faster, you mostly get into time trouble or just miss the tactics."

Every expert or master I've talked to, and everything I've read written by serious chess teachers, say the same thing: blitz is detrimental to your game.  If you want to improve, don't do it.  A local expert told me, "Unless you're a master, anything less than 20-minute time controls is at best a waste of your time.  You're reacting completely on instinct and reflex.  You can never progress tactically or strategically at blitz time controls."

iused said: "Here it says that it is the average time to solve the problem. That would mean that it does not include the times of those who failed."

That's an inference.  However, they do not say if there is a penalty given for incorrect answers.  Someone on another thread alleged there was such a penalty.  So, I don't think it's clear.  If this is the case, however, there's a real problem with that, for reasons I've said above, namely the measure is an artifact of player behavior, without some type of penalty for incorrect answers.

I don't really agree with that.  And anyway, it's just to have fun (or can be), so I wouldn't be too worried about it.

I do think blitz is detrimental...for those who are not high level players.

I think that blitz is neat for really good players...of which there aren't that many (proportionally speaking).

Most blitz players that I've known play like crap with standard time controls. So, with blitz/bullet...well what is the superlative for crap? Crappier, I suppose.

But...they are having "fun".

Shoot me, if that is fun.

Your rating goes down so that you are rated closer to problems that you can solve within the average time... this should be in some F.A.Q... somewhere. Your tt rating may seem to go down but it is all in the interest of actually improving your chess

I found some interesting articles on the pros and cons of blitz chess.  I guess this is a rather old chess controversy.  I didn't know it was such a heated area of debate!

http://www.onlinechesslessons.net/2011/05/28/confession%E2%80%99s-of-a-blitzaholic-%E2%80%93-will%E2%80%99s-opinion-on-chess/

BTW, I love that pic of an "alcoholic drink chess set"... classic!

e4nf3 said: "Most blitz players that I've known play like crap with standard time controls. So, with blitz/bullet...well what is the superlative for crap? Crappier, I suppose.  But...they are having 'fun'.  Shoot me, if that is fun."

That's pretty much how I've come to feel.  I don't really find it fun anymore.  Now I just find it frustrating.  At the same time, I've really grown to enjoy correspondence games a lot, where I feel I'm thinking about positions in a way I never have before.  I wonder if some of it has to do with attention span.  A lot of my students can't seem to focus more than 5 minutes on a problem without their mind wandering or checking facebook on their phone.  If you can't do that, it's hard to sit still for a 2-3 hour game.

The main reason I can see why someone would want to play blitz for improvement would be to try out a lot of opening lines.  But overall I think it just reflects instinct and reflex, whatever is already imprinted in your subconscious.

I think you're right that the tactics trainer encourages speed over accuracy.

From this graph,

http://support.chess.com/Knowledgebase/Article/View/53/0/how-does-time-affect-my-tactics-rating

the score decreases the longer you take (which makes sense), but also decreases more and more rapidly (which I think is odd). Once you have exceeded the average time your score decreases ever more rapidly towards 20%.

Also, 20% is not much more than 0% so the penalty for taking double the average time to solve the problem is only a little less than failing it completely. This means that as your time runs down it is better to gamble on half an idea, rather than taking the time to see it fully. You stand to gain more than you stand to lose. I would like to see the scoring tweaked so that it costs a bit less to take the time to get it right. 

On the other hand I'm not entirely sure how much of a problem this is. I hide the clock, and don't move until I think I see the end of the problem. Sometimes I get negative scores for taking too long to find the solution. My rating would probably be a bit higher if I adopted more of a blitz mentality, but I have no shortage of challenging problems at the rating I'm at. 

Thinking a bit more about how to tweak the scoring. As it is now, it looks like you always hit 20% at twice the average time and that the standard deviations determine the shape of the graph. The graph shows 100% at 2 standard deviations less than average, 90% at 1, 80% at the average time, 60% at 1 standard deviation over and 40% at 2. 

I don't think that standard deviations are the best way to do this given that the distribution of solution times is probably skewed. I think a better way would be to use the percentile value to calculate score. If you were faster than 40% of people you get a certain score, faster than 60% of people, you get a better score. I think that using twice the average as the lowest score is a bad idea because it doesn't reflect the spread of peoples times. I think this should also be linked to the percentile value.

No doubt about it...lots of very fast time pressure on the tactics training.

I did get a reply from staff. It answered some of my questions but raised more. I also disagreed with some of the assertions made by the staff who replied (who is on TT team) and tried to explain to him why. I appreciate the earnest response I got and hopefully it's the start of a fruitful discussion. If there are especially interesting parts of the discussion people might be interested in, I'll post them here. I think I just might have fundamental disagreements with the design team. If that's the case probably little will change. But I hope to have given them something to think about. When someone answers 32 of 33 consecutive TT problems correctly in rapid time control average times, and his rating stays essentially constant, it seems hard to argue TT has accuracy as its main goal.

TT isn't bad...but it definitely could be better.

For example...

Players should be made aware that in the rated mode, the time allotted is very tight. If you are not a speed chess player...this could be more frustrating than it is worth.

Maybe there could be an option for the TT. Do you want to play rated, speed, rapid or standard? This would be a great option!

Then there is this thing of black and white. You either get the machine-perfect (and...I, for one, am not a machine!) or you lose everything. You know, in real chess, very few players get "the ultimate" move on every play. How about if the move was exactamente you get the full rating...but if your move was not that, yet still quite good -- then you'd get a partial rating?

More? I could go on for hours. But, I am the quiet type.

e4nf3,

I don't think TT is bad either.  The interface, the fact there are ratings (although I disagree with the specifics), the analysis of each problem, the ability to have comments, are all very good.

I think the choice of what time control to play could work well.  Each problem would have different ratings for each time control.

As far as being "machine-perfect", I did get confirmation that "partial solutions" (i.e. getting 2 out of 3, or 3 out of 5 moves correct) are considered in the computation, so it's not just "pass/fail".  Which is a good thing.  My only issue with this is that, as you point out, not all imperfect moves are bad moves, and many are nearly as good as ideal.  For many problems for instance, the first forcing move can be very difficult to see, but after that, it is easy to mate, although it can be done in say, 2, 3, or 4 moves.  To my mind, finding the first difficult move is the most important part of the problem, and whether you mate in 2, 3, or 4 is something to consider but not paramount.  This would not be that difficult to measure by say, measuring the "relative error" of the player's move from the ideal move (as given by the engine they use).  So, if the ideal move wins 150 centipawns, and the player's move wins 130, no big deal.  Or maybe the ideal move forces mate in 3 or 4, but the player's move wins a rook or more, making a practically won position.  This could be computed.

For better or worse, I'm not the "quiet type".

Well, this "quiet guy" was a joke...as I'm sure those who know me would agree.

Yes, if you get a partially right answer, you do get some credit...unless the clock over rules you.

I was thinking along a different line. Let's say in a real game...I win the Q, or maybe an R. This isn't too shabby. Maybe the exactamente (I've never used that word in my life...but, oddly...second time today) move was a mate.

I'm not saying that tactics should be watered down. No...no...I like a tough fight. But, is it realistic that good moves are worth nothing and only the...here we go again...exactamente move is lawful?

All things being equal (ceteris paribus), in a real game...if I get your R or Q...I'm gonna win.

BTW...I wish they would allow us to reset the rating (which they did up until a couple of weeks ago). That would make me much more happy than I am now.

Also, if anyone hasn't used TT lately...they'll find that it is much tougher. Loaded boards more frequently (the higher the piece count, the more difficult it is to narrow in on the problem...meanwhile, the clock is ticking). Cheese...lots and lots of cheese now (rat bait...slows you down because you have to go sniff it and the clock is ticking). Also, you play at 1200 and it is like what 1600 was...four or five step intricate combos (this is 1200?!)

Standard deviation chicanery? Sure. But these are some of the nuts and bolts of the problem.

Oh...I could go on forever. But, as I said, I'm the quiet type.

I believe over inflated only applies to our opinons of ourselves.

ModularGroupGamma:  why do any of us care about your personal problems? 

ChazR,

Are you trying to put me on tilt or something?

I use TT as a tool for learning tactical patterns in chess. If I take a short time on a problem, it is likely that was because I knew the pattern well, so award me points to try more complex patterns. If I fail to solve the tactic in the short period of time alloted, then there is something to be learned.

The short period of time is essential to the TT, even though my goal is not (limited to) improvement in blitz chess. Your points on the pedagogical failure of TT do not apply to me. When I run out of time and quickly choose a candidate move, I faithfully replay the tactic and run through the calculations regardless of the outcome earlier.

And, it beats eating potato chips and just sitting and staring from your couch at the boob tube.

Also, if for example you don't drink alcohol, it is a good alternative to turning off and tuning out of reality for 15 or 30 minutes.

In fact, it may even be better than sex. No...forget I said that. That's too much of a stretch. Pushing the TT as an alternative to whiskey is about as far as I want to go with the absurd.

It beats exercise when you first wake up.

Now here is a fact check...

If I want to see how good someone is, I look at there tactical training rating.

There is no way that I can comprehend that someone can use an engine,  database or analysis board on rated TT.

So...someone has a 1900 correspondence chess rating but a 1310 tactical training rating. Tells me all I need to know.

And, yes, I am aware of the argument "I am a brilliant chess player but I need time to think." Sure. Would be nice to have plenty of time.

And that's the worst thing I've been fighting...the time. I hate it when I get the right answer but am 15 seconds over the time limit, the buzzer sounds and I get strapped into the dunking chair. Hate it. But...I'm getting faster and better.

High ratings on the other stuff but very low on the tactical? To me, I smell some bull dung. I have no doubt that that is one of the criteria for exposing the flimflammers (they don't like you to use the word "cheaters" on the open forums...so I didn't).