Tactics and FIDE Rating

@Fezzik I was responding to posts 5 and 10.

I clicked on your link and it's not working.

http://chesstempo.com/

Or, if for some reason the link doesn't work, go to google and enter one word with no spaces......chesstempo.......and google will give you the link to the chesstempo home site.

chesstempo is a great site. When there are similar features at chesstempo, they are better quality at chesstempo -- and cheaper, too. Somebody already mentioned the additional options at chesstempo for tactics training, but they didn't mention that you can click the variations to move through them after the solution has been solved or failed. Chess.com just gives you the moves as a bunch of text, and you have to open a separate board and manually make each move to explore the variations.

Comparing the chesstempo game explorer to chess.com one, th chesstempo one is much better. It keeps track of all the variations you've explored, and it gives you a lot more information. You can see only games of 2500+ players or 2700+ players if you want. You can sort the results by date, by white playing strengh, by black playing strength, and many other options. You have the option of seeing the next moves for all the currently selected games, whereas chess.com requires you to open a separate window to see the game list, then click on a specific game to see the next move for that game alone. chesstempo keeps track of all the variations you explore in the chess explorer, so you have a record of the lines that you've investigated. With chess.com, every time you make a different move in the explorer, you lose the record of other variations you've explored. Chess.com makes the links to the next positions not be clickable when there's only one game for that line, whereas chesstempo makes those links clickable. On top of all that, it's faster too, as it doesn't do full page reloads for every little thing that you do. It just updates the parts of the page that need updating.

There are many other advantages to the game explorer at chesstempo, and for everything else that I've made a careful comparison of, chesstempo is far superior to chess.com, which is one of the reasons that I cancelled my diamond membership here a couple of months ago and got a membership at chesstempo.

For the record, I have no connection with chesstempo at all, except that I am a happy customer there for a couple of months now, after I got so frustrated with the support and the tons of little problems and poor design decisions at chess.com that they have no interest in improving.

Now, I am no sales rep for chesstempo.

I belong to many, many chess sites...and pay subsription fees to...um...more than 4. I really need to start a bookmark for the ones I have paid for!

---

Since the OP/me and readers of this thread have had the good fortune of one of THE top rated Tactics Trainer Chess.com members chime in, NM ozzie_c_cobblepot, I for one would like a tip or two on getting the most out of my tactics training sessions.

For starters, our own IM David Pruess and ChessTempo do not agree on how to get better at tactics.

From the Chess Tempo site: Tips for improving your Tactics Training using Chess Tempo

1) At the start only do Standard.  No Blitz.  Your goal is to get the problems right and improve your calculation, not to do them fast.

---

Chess.com has only one time control and, I think it is set to Blitz. This goes contrary to the thinking of ChessTempo. If chess.com still stays with only one time control - fast - then I would like to see a button added like the one at chesstempo - Give Up. Why should we need to make a random move when we don't know the answer?

IM David Pruess talks about tactics in post no.29, found here - http://www.chess.com/forum/view/general/chess-advice-most-chess-players-dont-like-to-hear?page=2

As a teacher, I do need to use a yard stick on Pruess's rear end for failing to  capitalize at the beginning of each sentence. ;)

"...or when i give players in the 1000-1800 range advice on improving their tactics, viz: 10-15 min per day of solving simple tactical puzzles. the goal is to increase your store of basic patterns, not to work on your visualization, deep calculation. remember that is your goal. you are not trying to prove that you can solve every problem. if you don't solve a problem within 1 minute, stop. it's probably a new pattern or you would have gotten it by now. (with private students i'll take the time to demonstrate this to them: show them through examples that they can find a 3-4 move problem in 10 seconds if they know the pattern, and that they can fail to find a mate in 2 for 10 minutes if they don't know the pattern). look at the answer, and now go over the answer 3 more times in your head to help the pattern take hold. your brain can probably take on 2-3 new patterns between sleeping, so you should stop once you've been stumped by 2 or 3 problems (usually will take about 10-15 min). there is no point in doing more than that in one day. and any day you miss, you can't make up for. a semi-random estimate on my part is that you need about 2000 of these patterns to become a master. so you need to do this for 2 years or more. ..."  - IM David Pruess

Thank you. I forgot to mention this finer point. You don't get a FIDE rating unless you pay for the Silver or Gold membership. I have the all you can eat Gold membership.

Below is from the ChessTempo site regarding FIDE ratings and membership. Please remember, I'm not a rep, just a crazy - will buy anything to get better - desperate chess player.

This site does say to not take the rating too seriously, which I highlighted in red.

---

 FIDE Rating Estimates

FIDE rating estimates are provided for both your blitz and standard tactical performance. The FIDE estimate is based on a statistical analysis of the relationship between the actual FIDE ratings of the users who provided a FIDE identification number and the solving patterns of those users (including recent problem performance and average time taken on recent problems). From the statistical analysis, an equation is produced that provides an estimate of all user's FIDE ratings based on their current performance. Note that the estimate is just that, an estimate, some users will have FIDE estimates well above or below their actual FIDE equivalent level. As such the estimates should not be taken too seriously.

Benefit

• Provides a rough estimate of your current ability compared to other FIDE rated players.

Quite possible Fezzik, since I havne't played a rated game since October.  I'll also add that my USCF rating is only based on 25 rated games thus far.  I consider my playing strength somewhere between 1650-1750 USCF, unfortunately I don't often get the chance to play tourneys.  I tend to destroy D and C players most of the time, score about even at the local club against B players, and some wins but more losses than wins against A and expert players (prob around 20% wins, 70% losses, and a few draws etc.)

You pay less for Chesstempo.com, because you get less, i.e., specific stuff like tatics training and endgame training. In this area of chess, after using both Tactics Trainer here and tactics training there, Chesstempo does a better job. 

That being said, the instructional videos alone are well worth the extra bucks for a diamond membership.

I didn't feel this way about the diamond membership a few months back, but Chess.com was good on their promise and have cranked out several high quality videos for beginners - stuff I would place in the must know category for those wanting to go beyond the beginner level.

Last thing before bed. I just did a few tactics at both sites. Each site has a computer feature. At chess.com you can play against the computer using the same starting position of the tactics problem. It's pretty much a repeat of what was already done in the problem. I'm not getting anything useful from that. Maybe I'm doing something wrong.

At chesstempo, you can run an engine analysis of the tactic with an engine of theirs or one you choose from your personal collection. I use Rybka.

What I find useful is checking my conclusions with a chess engine right after the problem when I'm not sure about something. It helps me with calculation. I can move the pieces all over the board with as many variations as I want, checking what would happen if Black did this, or White did that, for example. I learn a lot from this exercise.

I suggest that Chess.com take a look at chesstempo's tactics trainer. It's simply better for teaching tactics - in many ways.

That other site - their rating system seems to converge too quickly. If you make an early error it seems to pigeonhole you. My rating after 100 trials is about 1935, with 72/100 correct. Seems to me it should boost the difficulty of the problems presented as you get them right more so than it does.

Excellent observation. When I make an error over at ChessTempo, my rating takes a 5 point hit. Correct answers are only worth 1 point, on average. 

At Chess.com, after an error my rating usually takes a 15 point hit. But, I can recover much faster, since correct answers are worth much more than 1 point.

---

Tactic Tip -  I need guidance on how to practice tactics and, maybe other readers are in need of some advice. Currently, I will spend up to 20 minutes on one tactic, here or at Chesstempo, striving for accuracy. My average deep think is 5 minutes.

What do you think of David Pruess's tips for how to practice tactics? Should I only take one minute per tactic or, go for accuracy, taking more time on each problem as recommended at Chesstempo?

"The goal is to increase your store of basic patterns, not to work on your visualization, deep calculation. Remember that is your goal. You are not trying to prove that you can solve every problem. If you don't solve a problem within 1 minute, stop."  - IM David Pruess

I don't doubt that. As a teacher who finds interest in learning theories, I find what he said of great interest.

"The goal is to increase your store of basic patterns, not to work on your visualization, deep calculation."  - David Pruess

This has me thinking. If I go with this plan, I will be spending far less time on each tactic puzzle. Will it work for me in the end?

My rating at Chess Tempo is 1600. I have solved about 50% of my last 100 attempts in about 2 minutes. The rest took me longer, some 10 to 15 minutes. I also strive for accuracy. I am learning new patterns and trying to apply them in each problem I attempt.

If I needed to solve each problem within one minute, I would have to be doing much easier problems. Perhaps I need repetition of more basic patterns? There is not an option to keep working lower level problems.

A 1600 rating is better than my 1465.

Excellent post. A 50% success rate just doesn't make sense, but then, David Pruess says that we should not be focusing on calculation or accuracy.

Here's a thought -

To speed up calculation, so problems can be solved with greater accuracy at a faster time control - say no more than 5 minutes per problem - look at the ChessTempo stats for the types of problems missed most often. Select a type and do many of them. That's what I am going to start doing. Second, I'm working hard on checkmate patterns because I find myself not finding a mate in one, two or three fast enough.

My tactic stats clearly show that I am weakest in double check, overloading and interference. This tells me that I really don't see those patterns and need to work on those separately under unrated. At the top of the screen next to my name is preferences. I need to click on that more often and select one of my weak areas, like interference.

---

Does anyone know the relationship between chess.com tactics and otb ratings, I am 2300 chess.com tactics, but obviously far lower otb.

I've always been a believer that learning something "sticks" better when you've put some thought into the issue.

For example, if you are investigating why the following shell command does not find whatever file you're interested in, first spend some time looking.

$ find . -type f -name '*.cpp' | xargs grep -l 'Hello World!'

After looking for awhile, you're much better equipped to process an answer from someone who knows.

So I find myself disagreeing with my esteemed colleague Mr. Pruess. If you want to learn a new pattern, he is suggesting that you either see it or you don't, and pretty quickly. If you don't see it quickly, then look at the answer and go on to the next one. My view is that if you don't see it quickly, you should spend some additional time getting to know the position. Only after spending a sufficient amount of time with the position will you be able to appreciate the solution, and process it into your long-term memory.

 The does seem to make more sense.

I gave the Pruess method  - blitz through the tactics problems and flunk most of them -  a trial run a few hours ago and after the session wanted to toss my computer through a window and quit chess forever. I felt completely stupid and found myself too grumpy for company.

My personality doesn't mesh with this method, even if it is the correct one, the most efficient one, for cramming bazillions of chess patterns into one's own brain.

Your method is the one that I use and will continue to use. Thank you for your input.

You refer to '*.cpp', rather than ".cpp", oh villainous one ?