Training Tactics Revisited

Well after reaching over 2000 rating and puffing out my chest I got shoved off my high horse and landed all the way back to the 1200s.

Today I just broke 1400 and will let it rest for a day or so until my next level (1500).

Someone in another thread mentioned that you can game the TT system by not accepting (making) the first move on the difficult ones and only gain points on the sure things. I suppose that would work, but the program isn't called "Trainer" for nothing and by not taking the hits you will never see the path to the best move or learn from it.

One aspect of TT, I was impressed with, was the statistical matrix. It is so comprehensive, it had a tactic called "opposition", which in 20 years and several books on chess I had never heard the term before. Had to Google it.

Once I found out it was an end game tactic I realized why I did so poorly at it. I have played nothing but 1 minute for 20 years online.

My best performance was on "en passant", but whenever it presents itself I naturally assume that is the key move.

Curious as how how others use the Tactical Trainer. 

Your rating in may was just inflated because of the various changes they had applied to TT then. The whole system was unstable for about two weeks, with the ratings of the puzzles themselves being insane. Most TT users reached a rating about 300 points avove their normal rating then.

So did I, which is why I reset my rating. - And now comes the funny part: I found out that it is way more satisfying and effective to reset one's rating every two weeks or so. Then I climb steadily again, strictly keeping  my solution ratio above 80%, no matter how long I need for the puzzles. This is good to improve in accuracy. After a week or so I reach my limit then (currently about 1500), when I'm confronted with puzzle I need multiple minutes to solve (the above 1800 stuff), which why I  can try/solve way fewer puzzles, and after a couple of days I realize that my pattern recognition abilities deteriorate significantly. This is when I reset my score to start all over with the quick-and-easy stuff.

I've done this 3 or 4 times now, and with each iteration my rating is about 50 points above the limit value of the previous iteration (like 1200->1400, reset, 1200->1450, reset, ...) .  I get a very broad band of puzzle types and difficulties, and my rating is always rising, so there is much 'positive reenforcement' and no frustration.

Perhaps I should mention here that I started with chess only in march, things may be different for more the more experienced addicts.

Well you are doing phenomenal compared to most people who start the game after 4 months. Although I base that on what you say your TT rank is. If your actual play is that strong after 4 months since learning to play chess I would find it hard to swallow. The nice thing about TT is it is like a lie detector test on how good a person really is, because it is impossible to cheat using a computer to solve the puzzles, due to the time constraints.

Edit: I just checked your profile and notice you only play computers, which is news to me. Didn't know they existed here. Well I am still stuck playing 1 minute and I am getting frustrated, as more of my losses are due to time than skill. However after playing 1 minute for 20 years it is hard to break out of the mold.

> If your actual play is that strong after 4 months since learning to play chess I would find it hard to swallow. 

Until a couple of days ago I really focused on TT (about 3 hrs/day avg), so my 'actual play' is much worse. If you'd take the time to inspect my games (not recommended!), my 'endgame techniques' would give you chuckles ... - For the time being, I prefer to play the bots, until I stop panicking at the mere sight of the ticking clock ... and this at 20/20 ...

Now , seriously: I think in order to improve, you should play slower games.  I think these really short time controls may make sense for really strong players (above 2000 or something) only. And if they cheat you?  - So what? At the end of the day, they're only cheating themselves. Besides, I believe this cheating at higher time controls is a myth. If someone really wants to cheat, he can do it in quick games likewise, using a some software.

Well I just managed to top 1500 today. Starting to puff chest out again. Like they say in AA "One day at a time" (or so I have heard) )  

Actually, I was so paranoid of comp users I never played anything but 1 minute for 20 years. A little 3 minute, but 1 minute was definitely a hard game to cheat on, especially because the site I played on has a watchdog program embedded in it. Now you sound more like a beginner, but how did you find out abut playing the comps here? I joined because they used an old drug dealer trick on me. I was given free samples of the Tactical Trainer on Facebook and quickly became addicted. I then signed up for the cheapest unlimited plan. So far I quit my job, my wife left, the mortgage company is foreclosing, but I just broke 1500 on TT! Okay I'm kidding, but that was the hook. 

Seriously, I need to play slower too, I am stepping into it by doing a daily each day and now I started the 3 minute. I think they call it blitz vs 1 minute being bullet. Doing surprisingly well there, but for some reason 1 minute players seem stronger and harder to raise your rating against. I don't know how old you are, but if you really like chess, you have to learn the opening trees to get around 2000 and over. You can be at 3500 in Tactics, but it will not do you any good if you can't get there. The problem there is, it isn't fun. It's tedious and time consuming. However, if you are young and do well at the TT, you should study openings to get on the porch with the big dogs.

Yeah I have been slowing down on the puzzles too and it seems to be working. I still screw up like today and thought I saw an easy mate and threw my Q in front of the King only to get the red "incorrect". I will keep trying for 100 a day, but it's like losing weight. The better you do, the harder it gets!

I wasn't looking for advice, but I would like to offer some. If you feel compelled to reply on the forums in spite of having such a tremendous rating, you might want to reword that first sentence. Some folks might get the wrong idea and label you a narcissist!

The way I see it your rating allows you to be given problems of an appropriate difficulty. My approach is to take the rated puzzles more slowly and seriously, and if I wish to blitz through dozens of easier puzzles I'll do so using the unrated training mode.

I didn't even know you could do that. 

Thanks for reminding me. I have yet to try those. I have seen them.

I lied! Didn't realize the Fischer one was a daily. That was not easy peasy, although I think castling may have done the same result. However after checking with the bishop I had to stare at it for the longest time to figure the mate.

What do you think is a correct alternative solution? I only see one solution to that problem (https://www.chess.com/tactics/527076).

I had a similar position today: https://www.chess.com/tactics/586670. 

I like the tactical trainer, but I feel like problem rating is a bit messed up. Today, for example, I got one puzzle wrong that was rated at ~1900 and had a very unusual and counter-intuitive tactical idea with trapping the queen with a knight and pawns. Then I got one right after a bit of thinking that was just a usual elimination of the defender 3-move puzzle, and that one was rated at almost 2500. 

I agree, I studied it and there is NO alternate "equally good" solution. Yes many have alternate solutions, but TT always wants you to zero in on the optimal solution. I wish I had a nickel for every time I grabbed a hung piece instead of looking for the forced mate.

One thing I noticed about players like you (the elite, over 2000) is that the tactical trainer more closely matches your actual rating as does others. My theory on that is you made it over the "openings hurdle" and have had hundreds of games to use your mid game skills. Since TT doesn't cover openings, those of us who suck at openings may be good at tactical thinking, but never get there in actual games, due to our lack of opening skills. Thus the skewed ratings on players under 2000. That's just IMHO though.

Baffles me that people do this. What is the point of cheating at something where there is no opponent. Tactics trainer is a tool we use for our own improvement, if you game the system then the only one who gets cheated is yourself

Just cleared 2300 for the first time using tactics trainer. A couple of years ago I struggled to keep over 1900. Been doing 20ish puzzles a day since then. I don't know whether ive actually improved or if its because of the changes... I hope its the first