Upgrade to Chess.com Premium!

Relationship Between Specific Chess Skills and ELO


  • 3 years ago · Quote · #1

    Isiah_Nougat

    I’m an experienced poker player who has just begun to learn chess.  One tool that helped me advance early in my poker career was a simple table put together by a smart and experienced player that associated various poker skills with advancement points in one’s career.  The table listed the various stakes of play along the poker advancement ladder and then listed which skills the player should be proficient in before graduating up to the next level of play.  For example, if the player wanted to have a reasonable chance at success playing $1/$2 then he ought to be proficient in skills A, B, and C.  If the player wanted to play $2/$5 he would need all of the previously listed skills plus skills D and E.

    I’m wondering if a similar table could be written for Chess, replacing the gambling stakes with basic ELO levels, 800, 1000, 1200, etc.

    For example, if a new player has learned to rarely hang pieces anymore, what ELO level could he be competitive at?  Now take that same player and add an ability to identify, use, and avoid the most basic one-move tactics like forks, pins, and skewers—at what ELO would he have a reasonable chance at victory?  Or, to use an example at the higher end of the ELO spectrum, at what ELO would the player need to have engaged in an extensive study of openings to be competitive?

    Is it possible to create a meaningful index like that for Chess?  If so is anyone willing to take a stab at it?  

  • 3 years ago · Quote · #2

    Arctor

  • 3 years ago · Quote · #3

    Isiah_Nougat

    Thx Arctor, that link is helpful.

    The link seems to show 1400 ELO and 1800 ELO as two key development milestones. I was hoping someone would take a crack at laying out the most important specific skills to reaching 1400, to reaching 1800, to reaching say 2200 etc (the link broadly categorizes skills and doesn’t appear to address anything above 1800).

    I guess I was hoping for a more granular treatment.  It’s probably wishful thinking to ask anyone to do this since it would likely require a fair amount of thinking and work.  But if anyone’s willing I’d love to see it.

  • 3 years ago · Quote · #4

    waffllemaster

    I started to try, and believe me, I'm not one to shy away from long involved posts, but this one is pretty tough for a few reasons. 

    One being the continuity of chess skill is fuzzy for all of us I think... the skills across each class are essentially the same, just separated by degrees.  For example I've seen world champions talk about games using "beginner" language... they're just applying these basics at a very high level that isn't necessarily verbalized differently from what many amateurs do during a game.

    Also not all 1600s (for example) are equal in the same areas.  Opening, middlegame, endgame, calculation, and other general skills can each compensate for the deficiency of another.

    -------------------

    So getting that out of the way, there is one general idea that respected author Dan Heisman posits and he draws a line right at USCF 1600.  He says players under 1600 are not consistently checking to see if their intended move is safe from all of the opponent's checks and captures (forcing moves).

    Of course we all learn to look for immediate threats, practically from day 1... but to check and catch ALL of our opponent's basic forcing moves 99.99% of the time puts a player into 1600 category according to Heisman.

  • 3 years ago · Quote · #5

    Skwerly

    yup, i gotta agree with waffllemaster here.  not all 1600s are created equal.  although, in the grand scheme of things, they are going to *most likely* be generally worse in all areas of the game than an 1800. by the time you reach 1800 uscf, you have a good grasp of general chess and know a few tricks. then, the next hurdle is 2000, which is expert level.

    to reach 2000, you are an extremely solid player who has more than the basics down and some good endgame skills.  2000-2100 in real life isn't easy to get and maintain.

    the next jump is 2200, which is master, and from there the world goes sideways.  to hit 2200 you gotta have some real study time behind you, and a little talent, if i do say so myself. 

  • 3 years ago · Quote · #6

    AlCzervik

    Nougat, I completely agree with waffles 2nd part. Even though I haven't read any chess literature, it certainly seems sound.

    However, there is no comparison between chess and poker. Or, any other endeavor. Your poker table works for poker. Learning chess is a pursuit that can drive one mad!

    Consider: At my current chess.com rating, I am at about the 85th percentile. To get to the 90th percentile would take a massive amount of study and play.

    The reason there is nothing in the link that goes beyond 1800+ players is because attaining a level of 1800 is difficult enough. Players at that level have studied the game. Some for years. The link gives basics.

    Matches between players at, say, 2000 level, is difficult for some to understand. Why did he make that move? It's for later, but we don't see it!

     

      

  • 3 years ago · Quote · #7

    waffllemaster

    @Skwerly

    But then you hear IMs make comments like "I'm lacking fundamentals."

    I think it was Silman who said to beat a master is rather simple for a GM, because masters are fatally deficient in two areas of their game, and the GM just has to wait for it to pop up.  Likewise IMs have 1 deficiency, and to beat them you just have to wait for their blunder in whatever area it may be.

    Of course in my perception, Masters are very very solid all around players and very consistent players who I can't realistically hope to beat.

    Then you hear that Kasparov calls Akopian, rated ~2700, a chess tourist and you're wondering if every last bit of it is subjective.

  • 3 years ago · Quote · #8

    beardogjones

    Computers have shown that search depth correlates with rating -

    and supports D. Heisman's statement.

     

    Also, some computers like Shredder claim to give an accurate elo 

    after only a few games of play - but I do not know what algorithm 

    is used. In fact, i hate when I win and my rating goes down...

  • 3 years ago · Quote · #9

    waffllemaster

    @beardog
    Computers have shown that alpha-beta pruning and fine tuned algorithms that mimic human positional evaluation correlate with computer chess strength.

    By comparison the thing you said about computers is quite a leap. 

    Also, what I call one of the two main misconception among non-players is that calculation correlates to skill when in fact it's evaluation.  Everyone can calculate 20 moves deep, it's incredibly easy.  What's hard is making these moves relevant, in which case 20 moves is nearly impossible.

  • 3 years ago · Quote · #10

    Elubas

    "Everyone can calculate 20 moves deep, it's incredibly easy."

    What is your definition of calculating 20 moves deep? I would have a pretty hard time visualizing (correctly) simply what a position would look like 20 random moves into the future, even when ignoring all the tactics. Unless of course they were just a bunch of queen checks or something.

  • 3 years ago · Quote · #11

    waffllemaster

    Elubas wrote:

    "Everyone can calculate 20 moves deep, it's incredibly easy."

    What is your definition of calculating 20 moves deep? I would have a pretty hard time visualizing (correctly) simply what a position would look like 20 random moves into the future, even when ignoring all the tactics. Unless of course they were just a bunch of queen checks or something.

    Exactly, here's an example.  Of course I went on to say that to do this in a meaningful way in a real game is nearly impossible.

     
    Not only should people who know only how to set up the board be able to perfectly visualize this sequence, most should also be able to set it up at any point given only the move number.
  • 3 years ago · Quote · #12

    hicetnunc

    It's an interesting question - I think the core skills to reach the 1st milestones would be :

    • 1300 elo : avoiding putting your pieces en prise (ex. putting your queen on a square where it can be captured immediately by a pawn), meeting your opponent's direct threats, knowing basic mates (attrition wins), applying basic opening principles, being able to use basic tactics 1-2 moves deep, counting to evaluate trades
    • 1600 elo : meeting your opponent's 2-moves threats on a fairly consistent basis, basic opening repertoire, knowledge of most common tactical patterns, ability to calculate 2-3 moves deep in some positions, activating your pieces, knowledge of basic pawn endgames, knowledge of most basic middlegame checkmates,

    elo scale would be USCF/FIDE rather than chess.com's

  • 3 years ago · Quote · #13

    fburton

    Dan Heisman has already been mentioned. He gives a series of milestones - the sort of thing OP is looking for, but only up to 1400 ELO - in his book "Everyone's 2nd Chess Book". You can read that bit in Look Inside at:

    http://www.amazon.com/Everyones-2nd-Chessbook-Dan-Heisman/dp/1888710268/

  • 3 years ago · Quote · #14

    Estragon

    If you can make a proper Margarita without getting lime pulp everywhere, you can do anything.

  • 3 years ago · Quote · #15

    madhacker

    I think it's a very difficult question to answer because every player is an individual human being, and the ELO rating is only a very high level abstraction of their playing abilities. You can find two players of the same rating who have completely different styles and different strengths and weaknesses. I remember Kasparov once saying he was surprised at how little theory Carlsen knows.

    It's probably more useful to try to characterise at the lower levels, in the sense that a 1200-1400 probably wont be dropping pieces anymore like a 800-1000 would, but at the higher levels I think it's much harder. So it's no surprise to me than Heisman only goes up to 1400.

  • 3 years ago · Quote · #16

    fburton

    Good point, madhacker.

  • 3 years ago · Quote · #17

    PatzerLars

    It might also be that in countries with a higher overall chessic standard like russia  for example you will need more skill per rating point since you don't play in international tournaments right off the bat.

  • 3 years ago · Quote · #18

    rickdaniels52

    there is no luck in chess as in poker  as a patzer i find i tend to be myopic when i make my biggest blunders and i dont see smaller weaknesses or strengths   im also impatient

  • 3 years ago · Quote · #19

    madhacker

    Yes, rating pools is also an issue. This is what Lars just described - there are only a certain number of rating points to go round in any chess society, no matter how strong or weak the players are. Only rarely do ELO points pass between geograpical locations, as most people you play are in the same area as you. I'm sure there are players in Russia who don't even have FIDE ELO ratings who would get into most national teams elsewhere in the world.

  • 3 years ago · Quote · #20

    madhacker

    Rick, if you think there's no luck in chess then you clearly haven't played very much chess!


Back to Top

Post your reply: