Friday, July 24, 2015

Why Am I so Bad at Blitz?

No deep philosophical discussions this week, witty musing, perhaps self-motivational pep talk on the game…

No This blog is dedicated to my pretty-awful blitz game..  On I’m only around an ELO 1000 at it.   with the average being around 1200 this means I am significantly worse at blitz than the average user….

Blitz Chess kills your ideas” says the Great Bobby Fisher…. Yes I know!  Blitz is a poor way to improve’ says another guy in the back of the room.  I Get it!  Boy do I get it.

Blitz chess Isn’t, and won’t be a big preoccupation*…..   My purpose is to understand chess- and master it (to the degree that an amateur old guy can) and that means giving myself the benefit of some time to figure out what is going on at the chess board.

But it still bothers me, that the average user who comes along and fights the hoards on the world-wide web can get to about 1200 (near most peoples initial rating), and I who has worked for nearly a year now solving thousands of chess puzzles- pouring over my games for insights- paying for books and tutoring can barely crack 1000.

Anyway  I can complain- or give myself “good advice”- or even look up “good advice”  (though frankly, I’m really well read)… but that really doesn’t do much.

Instead I’ve taken some definite action!  I’ve done an experiment ….
“My Blitz Experiment”   earlier in the week, I played 5 games of blitz (G10) against comparable opponents- losing 2 and winning 3.   Getting about the same poor results I would expect.   Now the interesting part- I now go over the move 1 by 1 on a chess board recording my choice at 20 seconds, 1 minute and if needed 2 minutes.   I take this score to my PC engine and look to see where my choices are….  Where does my analysis rate compared to the game; and particularly comparing quicker and slower responses?

I have only done this (for now) for one of my first loss- but the results are interesting.
  1. First this game was really decided by time; and my analysis shows that I did struggle with time; even in the analysis “Blitz” portion of the study.   Of the 25 moves- I couldn’t come up with a reasonable choice is :20, with 5 moves.
  2. I came up with a better reaction to the opening given a longer period of time;  and its likely a good investment in blitz play to slow down after the first few moves of rote.   The opponent often makes odd moves in blitz that have hidden downsides (development, space, pawn structure)… Being aware of the consequences of their ‘creative’ play could win a lot of games of blitz I think.
  3. I expected that that the longer time spent looking over the position would  ‘carry over’ to the quick moves.  NOT True.   Despite the extra time looking on previous moves, A blitz move during analysis was as good during analysis as it was during the game.  About half the time the same move; sometimes better, sometimes worse.   This means I can’t spend the first several moves thinking deeply at the game; then play the rest of the game rapidly.  the game is too unpredictable for that (for me). 
  4. A very interesting 20 and 21st move.   Where in analysis; with the longer time I make a material losing blunder on the 20- that I don’t make in blitz.  And then on the following move, I avoid losing material on the 21st move with a the longer time.   ON both moves, during analysis- I have no blitz move. ( I knew the move was important and took the longer time to work it out)… which means…
  5. Taking longer times doesn’t guarantee I won’t lose to tactics.  Basically I tried to get tricky and ran into an unexpected refute.
  6. on the other hand,  in the blitz game, I lost material to the 21st move because of a pin (discovered attack) that had been in play for many moves. In the haste of blitz I forgot about it.  obvious errors increase in the quicker game.

This blog Could be called “IN Blitz, Time becomes an extra piece.”  And that is perhaps its only real draw for me….. oh and I want to quite being embarrassed by a class E/(F?) blitz elo.



  1. ok, I added the astericks for future purposes. lol.
    but never mind that- I was kind of hoping this week to get a little feedback from the land of the blogspot; what do you make of a 1000 blitz elo- embarrassing? irrelevant?

    should someone who is trying to improve in chess- make an effort to get to a certain online blitz rating? and do you think it points to important failures of the thought process or tactical recognition?

    part of the astericks is to give myself wiggle room. I cannot promise to never decide to dedicate myself to lifting the ELO and perhaps a sizeable investment of time and effort. but I don't think blitz ELO is a good indication of amatuer chess mastery.

  2. It is hard to judge (evaluate) how good/bad you are - unless you share the games (blitz ones - as you mentioned these at this post).

    There are many elements that may be important to make some kind of your games rating (mark).

    I play mainly at FICS - my blitz rating is below 1900. I have big problems against players rated 1800 or more. However I can outplay most of the players rated 1400-1600. The same may be with you - if you play really strong (and fast) players - your rating may stuck at certain point. And in contrast - you can play players rated 800-900 and beat these to achieve 1200-1400 rating.

    I would not focus on BLITZ rating that much. If you learn how to spot tactics and make best use of it - you would be able to do it at blitz games too. However it requires practice and self-discipline, not to mention a good way (system?!) of calculation.

    Low level of blitz may be a very good indicator of bad level of pattern recognition AND tactics. In the worst case - you will be losing ALL the games... just due to blunders.

    If you want to BECOME a really good player - try to learn chess WELL first. After that you can try to play good and fast(er). If you reverse the order (being fast before obtaining a good level of play) - you will end up like thousands of blitz players... 90% of these will never achieve A-level or more (rating 2000 or better).

    I have played about 100K chess blitzes so far. If you have more questions or comments - just let me know. I will try to answer and comment as much as I can. Hope it helps ;) :)

  3. At every site I play my blitz rating is remarkably lower than my standard rating! But I HATE blitz. It is my least favorite form of chess. I like slow games. They actually relax me.

    So my take on it is that your blitz rating is irrelevant. If you want to get a good blitz rating you have to play lots of blitz, do speed tactic training and get some wackadoo opening gambits because as you say blitz is all about time.

    At this point the slowest game I will play is Rapid (the slowest is 17 minutes and 17 sec increment added). Still too fast but a little thinking can happen.

    Again unless your goal is to be a great blitz player I wouldn't worry about it.

  4. so first of all, I really appreciate the quick feedback. thank you!

    my mind has been on blitz (off and on), since I got a 1700 rating in cc, a couple months ago. careful analysis and unrushed play had brought many victories.

    I think it is interesting that you are interested in the blitz games more than making any sweeping statements about the elo.

    it occurred to me... that I play so little blitz- that I am expecting to much for a rapid increase in elo in just a handful of games.

    on the other hand, I don't like blitz. blitz games make me nervous, blitz games feel rushed. even in standard time controls sometimes I see the big refute Just after I make the move. in blitz- sometimes these 'what the heck was I thinking' errors get really obvious.

    no. the point of the experiment wasn't just to address blitz. it was also to understand the effect on time on my game. am I using time well?
    anyway I often warned myself and others not to get all caught up in the ego of chess ratings and not to mistake what is a performance rating that can drastically vary from situation to situation, from some kind of insightful measure of what you know and how good you are at chess.

    its much better I think to have faith in the process, faith in the hours and the work; combined with a lot of sober reflection on ones games.

    I'll stick to that viewpoint; while occasionally trying to bump up both my standard and my blitz online ratings. Perhaps someday, when I am feeling a little stronger and more prepared for it; I'll go all- out a couple weeks and really look to lift my blitz score. but as you say-- There's the danger that your just going to reinforce your chess limitations... I agree that this makes more sense as you gain an expert chess rating.....

    though I want to add, for a guy who has achieved nothing higher than class E for most of his life-- becoming an Expert with a geniune uscf (or FIDE) rating of 2000 is Very impressive! I'm not entirely sure that any one really has gained that much ground as an adult.

    so for me.... I don't know. Class A territory might be near the glass ceiling where there's little chance for continueing elo gain-- and learning to play fast makes more sense.

    this is all a good discussion for 2027 of course when I learn to play strong chess. LOL.

  5. I think TommyG made a lot of good points.

    In general, I've observed that blitz strength increases with your standard chess level.

    However, if you want to become stronger in blitz, you need to invest more heavily in 3 areas :

    - pattern recognition (mostly tactical patterns)

    - an aggressive opening repertoire, or, alternatively, very simple openings which delay the fight

    - playing quick and making practical, if sometimes sub-optimal decisions

    The problem is that if your thought process in blitz isn't mainly intuitive, then you'll pick bad thinking habits that could hurt your long chess (ie. rushing). So if your aim is to become a better 'long games' player, I would focus on this objective, and let your blitz rating take care of itself.

  6. You need to learn the correct timemanagement for blitz, rapid and standard. It will take a few hundred games until you develop a right timing and thinking process for blitz. If you are used to check and double check your moves in standard games then it takes energy to stop doing this in blitzgames for example.

    At average a move gets better if you think longer about it. The faktor how much better you get if you think twice as long or how much you get worse if you think only half of the time was named by Empirical Rabbit : K
    Empirical Rabbit, Uri Blass, me and even Richard from Chesstempo did analyse K and it is somewhere between 80 and 200. I averaged the results to somthing near 140. Of course K is not the same for each player and as soon the player exceeds its visualisation abilitys K shrinks to 0.
    The problem with blitz is: that you exercise bad moves. You "learn" them. I analyse every single blitzgame with Stockfish for "blunders" but only 1 sec per move so that the analysis dont take a lot of time. Then i look for the biggest 3-6 blunders and analyse whats going on there. AND i analyse my opening in these games. Blitz is a good method to learn openings. So my aftermath is waaaayy longer than the whole game.

  7. I am back after a good weekend. I really appreciate all the comments, thank you.

    I sporadically feel bad about my blitz ELO but, I tell myself that no one number can exactly describe how good I am at chess or how much I know. and for that sure It would NOT be that number.

    Aox, I am intrigued by the factor, k. do you know of a good blog describing k? I have recently spent some time thinking about how much better my thinking gets with extra time. even in a long game; it is good to be able to spend time wisely.

    indeed I've been thinking an awful lot of whats said of grandmasters- how in apparently calm positions they can spend quite alot of time thinking while they rapidly move through some pretty tactical play.

    this is of course, because they have figured out all the tactical play before they started to make their decisive attacks. I need to THAT much better. I need to have better time management. You have good habits Aox. I'm thinking that serious analysis of ones game espacially with an impartial and strong PC engine is good for ones chess strength.

  8. The term "k" is from Empirical Rabbit: but thats mainly math. You find at this Site a lot! of material how to calculate K.

    I did analyse K since 2008. My first posts where often about methods to calculate K
    I had a lengthly discussion with Richard about K here you need to read until the end of the comments because it did take a while until he accepted the fact that thinking longer results ( at average ) in better moves.

    But there is no analysis about "how to improve" k
    I am working on that at the moment but i have no reliable datas by now.
    Never the less some factors are easy to see:
    K is higher
    - with a better thinking process, if you think over and over about the same for example than you dont make any progress so k is 0
    - with better visualization. Bad visualization makes it necessary to recalculate a lot to be sure its calculated right.
    - with more and or better patternknowledge
    - with better concentration ability
    - better memory , you need to remember: what lines have been calculated, which positions have been evaluated and how, what the correct move in this opening is...

    You may compare my K with your K by doing blitz puzzles and miced puzzles at chesstempo. My performance at mixed puzzles is something like 50 points higher than my performance in blitz puzzles

    So if your performance in mixed puzzles at ct is for example more than 50 points higher than your performance in blitz puzzles than your (ct.puzzle-) K is higher than my K...