Chess.com Tactics Calculations

Anyone know how point calculations in the Chess.com tactics trainer are done? Or perhaps a link that explains them?

I only ask because I just did a problem that took me almost 2 minutes to get the answer to and I still got 97% even though the tim I took was much longer than the average time.

Then the very next tactic was a 1 mover that I got in about 25% of the average time, yet I only got 87%.

This is not a complaint. I like the tactics trainer.

I know the answer.

The rating and average time are based on past member's attempts at the problem.

They are just like game glicko ratings, with the puzzle itself as the "opponent."  If you fail the puzzle, the puzzle "wins" and its rating increases (based on your current tactical rating).  If you solve the puzzle, you "win" and your rating increases, then the puzzle's rating decreases for the next guy that attempts it.

Here's how it did work. There was talk of changes so not sure how much is still correct.

Firstly the total clock is 2x average time. Take longer than that and your score is capped at 20%. If you take the average time I believe you get 80% but not completely sure on that number. This score is then multiplied by the percentage of the problem you got correct. So 1 out of 2 moves in the average time will get you 40%.

The change your percentage score makes to your rating depends on your rating relative to the problems rating. So 50% on a tactic at your rating and your rating won't change. But for a lower rated problem you will lose points, or gain for a higher rate problem. The exact formulas for this depend on a rating deviation similar to your chess rating so your rating will change more quickly at the start.

The problem rating will have the opposite change but not necessarily by the same value.

Thanks.

The time part does not help me understand the scores I got in this case, but that's OK.

But, as you said, there may have been other changes that could be affecting the points.

Yeah the percentage you got certainly wasn't possible before if you took over the average time. I think you have misremembered some of the detail but it may be different now.

Agree with chessgg about alternate winning moves. There are tactics where I can go a queen up but it wants me to find a mate. That's just stupid, a win is a win.

Strategy is not 10%, it is less.

Tactics : 90%

Attitude : 5%

Intangibles (game conditions, venue, lighting, etc.) : 3%

Strategy: 2%

You are totally wrong.

A mate is a win.  It's a guaranteed win.

Being a Queen up it not a guaranteed win.

It makes a win more likely, but still anything could happen.

Pretty simple really.

Na queen up is a win. I'm not going to play anyone who wouldn't resign at that point.

Lots of people who are vigorously defended on those "why won't my opponent resign" threads.

I see people make comments like this about TT (An awful lot of Q sacrifices, Too much "cheese", etc..)

I understand what you mean, but I think that people who see it this way may not be aware of some basic techniques in training for any sport. (And, yes, I do consider chess a sport)

If you run 400m races, you do not spend most of your time running 400m. You spend most of it running 600m or 800m. Why? Because you want to practice harder than is necessary for the actual event.

I did the triple jump in high school and we had to wear 3-5 pounds of weights when we practiced.

I had an actor in a film I was making do the something similar. When he practiced his lines, he practiced them much louder than when we shot his scene. When I asked why, he said that, even though he was speaking normal to softly, being louder first made his voice stronger at lower levels.

I hope this does not turn into a "TT sucks" thing.

I like TT. I like the "cheese" and the seemingly imposible positions. And, I for one, have gotten better from it.

So you're in Thailand and make films?  That's cool. 

At the same time there are broken puzzles. I've reported dozens that have been deleted. Alternate mates etc, it doesn't cope at all with alternate moves.

Yeah, it's possible I am misremembering, but not by much. I did these puzzles just a few minutes before posting here. But now that i have some idea of how the calculations are done, I will be better prepared for these types of things.

About the go for the queen idea. I think the point there is "find a stronger move". Sure, you can get a queen, and probably you will eventually win. But if you can win now, THAT is certainly the best move.

If its an alternate mate with more moves then it still loses.  Less is more.

Mate in two is better than mate in seventeen.

On standardized tests, you are instructed to choose the best answer.  Many of the questions have two "correct" answers.  Your intelligence and aptitude is determined by your ability to choose which is more correct.

I do live in Thailand, but I don't make films any longer.

I made a few short films when I lived in S.F.. And I made a few with my students here. But I no longer do filmmaking.

I would like to though. It is fun.

But if you find a win, why should you keep looking? It doesn't make sense in reality, spending extra time when there may not be anything else to find. Also using up mental energy when you may have more games in a tournament. If you find a checkmate, play it. Don't waste time looking for a quicker one.

Tactics trainer is best for quick exposure to patterns. Leave the complete accuracy for tactics books that require it.

That is also a good point.

I totally see your point.

But, of course, TT is not meant to be reality. It is just a tool to help in the development of an important skill. Which, despite the occasional problem(no put intended), it does a good job.

It has given me many new ideas about tactics and I now see myself looking at positions a bit differently than I did a month or so ago before I started using it.

I never asked for advice about anything. Please go back and actually read the post.