Confession of a 1700s Online player

When I started playing on chess.com, I was really surprised that my rating went over 1700 because I am really not that good.  But, I have learned that I can play at that level if I study every move very carefully.  However, I am also capable of making really bad moves if I slack off on my analysis of a single move.

The reason is that I have not really learned out to play at that level, I am getting by with intense analysis.  Which is okay.  That is the purpose of turn-based, online chess and it fosters learning to play at a higher level.  However, it must be supplemented with live chess, to really learn how to visualize the patterns that you can find through analysis in turn-based chess.

I have not been doing that much lately, and my live chess has gotten really terrible.  The more I depend on intense analysis, the rustier my skills for quickly assessing a position become.

Anyway, I mention this because, I suspect this is true of a lot of turn-based players at every level.  Not all, of course.  There seem to be many turn-based players who play pretty much the same as they do in live chess.  But the point is; do not be intimidated by a higher rating.  Just because someone is rated a lot higher than you are does not mean they are necessarily better than you are.  If they are like me, there is a really good chance that they will make some mistakes and you can get them if you are paying attention.

Right - it simply means that they perform better than you do. Then again, some might argue that performance goes hand in hand with ability.

I know what you mean, though - my online chess rating is much higher than my live chess rating, although it might just be because I don't play much live chess, and my live rating is what it was over a year ago, before a number of CC games that have increased my online rating. But I'm sure that a large part of the difference can be attributed to my poorer performance under time pressure. Still, I don't think that this implies that I'm "really" not that good a player; after all, they also have world championships for correspondence chess. Who's to say which time controls define "real" chess? Some people mock bullet chess by saying that it's not real chess; I might as well mock conventional OTB time controls and claim that "real chess" is when you can think for days about a single move, when the effect of blunders caused by time pressure is minimized.

If you don't play live chess that often, at least try doing some tactical trainer problems every day to experience the need to quickly assess critical positions under time pressure.  It might help...

So , just cause Carlsen, Anand, Aronian are a lot higher rated than me doesnt really mean they are better than me ?!  Thats good to know !  

Reb, 

I suspect that the relationship between skill/ability and rating is actually rather complex. 

First, you have the issue that rating is relative to prior performance, and someone who has changed the way their working, or who is improving their play in some way, will generally have a skill level higher than their rating anyway.

Second, you have psychological factors. Some players will play up because of their competative nature. Other players will play worse than they should because of a timid or uncertain nature. 

Third, there's normal performance variability. There are 1600's who bracket 2000 performance games with 700 performance games, and then there are 1600s who's performance range is from 1500 to 1700. 

Obviously, (at least I hope it's obvious) someone rated 1000 points above me is almost certainly better than me. But someone rated 300 points above me may well not be. I've been improving on 1700 performances lately. It isn't too much to assume that my actual skill level is above my rating given that my rating has increased regularly every month for the last 3 or 4. It might be that they were better than me 5 years ago, but have been resting on their prior history and slowly declining since then, and their rating hasn't caught up with their actual skill level yet.

But the question becomes what's a reasonable range to beleive someone could be rated above/below me and still be the case that I can know they are better than me? I expect 400 points above me and 200 points below me is probably a fair barrier for me. But someone else might have a different range.

Possibly, all rating systems are imperfect.  Now for you, as a master, I suspect the issue is less valid.

But my point here is that it is more dramatically true in turn-based ratings since the format allows for a highly divergent level of effort put into each move.  When you face an opponent that is rated a few hundred points higher, don't be intimidated.  I know when I play a higher rated player, I often play too timidly or am afraid to capitalize on a mistake because I am afraid I am getting sucked into some trap of some kind.

I am very capable of making a mistake that would allow a much lower rated player to beat me.  I have never played any real, over the board rated games, but I suspect my rating there would be in the 1200s.  When I am not busy and I do not have too many games going my play is much better.  If I get too many games going or I am busy at work or at home, then my play gets much worse.  My current rating is 1738 and I think when I am really focused and careful with every move, I can hang with the 1900s.  But if I am busy, tired, distracted, I can get beat by a 1200.

We will have to make a distinction between online ratings and established OTB ratings as the latter are certainly more meaningful as well as more reliable indicators of actual chess strength/knowledge  than the former. In my experience if someone has an otb rating 2 classes higher than yours ( 400 points )  they are most certainly better than you in chess . I have played a few 2600 GMs and I lose every game, not even a single draw to my credit. 

Reb,

I expect that what you say is pretty true. However, I do think at hte lower levels the range can be much wider, particularly with kids. I've seen kids with 1000 ratings regularly trounce people with 1600 ratings. Adults of course are a different story and for them I think 2 classes is about the right range -- except for older folks who are playing infrequently and who are trying their best to make it to their rating floor.

I dont agree at all with this and I have done a lot of both . When you go to an otb event there are real expenses involved, entry fees, travel expenses, food , hotel to name just some of them so you are going to take such an investment a lot more seriously than online play. There is also the problem of cheating , which is easier and far more common in online environment than in otb . 

The fact that I'm 2750 in bullet demonstrates everything anyone might need to know about the ratings here... :)

Online, you can  1. cheat, 2. choose your opponents, 3. avoid certain opponents, 4.  be (and so can your opponent) more easily distracted (if live),  5. Play one color more often if you choose, 6. consult books, etc., if turn-based .... all things that can skew one's rating.

Why can't people accept their ratings?  This self denial, delusion, of being "underrated" is laughable.  Play more games, prove it.

If "kids with 1000 ratings regularly trounce people with 1600 ratings" they wouldn't be rated 1000 very long, would they?

It would seem like points 2 and 3 could be very important.  I play essentially only tournaments and team matches for my CC games, which means no direct control over opponents.  That should give me a reasonably realistic CC rating, right?  Except my rating is quite a bit higher right now than it "should" be because of a recent unfortunate streak of more than a dozen abandoned games (by my opponents) that shot me up 70 or 80 ratings points, and it will be a while before I lose myself back down to a rating more indicative of my ability.  That sort of thing can't happen OTB.

They usually aren't...

Actually, that's one of the problems with OTB ratings: the amount of games one can practically play is much smaller, at least without being independently wealthy and/or (more likely "and") living in either NYC or Moscow, so random outliers like a very underrated kid or a severely overrated Life Master can have a pretty profound effect (especially while still provisional).

I think OTB is generally far more serious, more controlled and better monitored.  Playing only tournaments online can eliminate or neutralize some of the more questionable aspects, but I suspect there will probably always be differences bewteen online and OTB chess. As NM Reb pointed out, just the expenditure in time and money to play in OTB tournaments is sufficient to ensure they are takenmore seriously.

Just to clarify, I think OTB ratings are still much more meaningful than online but, again, the fact that it takes time and money to play OTB means the number of games is drastically limited and thus one's typical sample size is massively smaller. This matters quite a bit!  To be honest, I suspect the most accurate rating system might someday something like the "pools" on ICC if cheating can be sufficiently minimized, since you can't pick opponents, color, or anything else that might game the system. Of course, minimizing cheating is a pretty big "if"...

 I did make that distinction in the first post of this forum.  Online, turn-based chess is what I am talking about here.  You could have a whole different discussion about ratings in general.  But the point I am making here is that Online turn-based ratings can be extremely inflated even without considering the possibility of cheating due to the extremely different levels of effort that can go into each move.  I have sometimes spent days on a particularly complex position going back to it several times and working through every possibility I can think of.  Other times, I have the computer on my lap while watching television and make a move after 10 seconds of review while glancing back and forth between the television and the computer screen.

My point is don't be intimidated by higher rated players in turn based because they may not be as good as you think they are.

Neither of them is compatible to what the real theater of chess is : OTB play because you have plenty of time to think(usually) at major events like the world open so its not compatible to live chess. But you dont have an anyalasis board like you do in online chess. A good drill would be to say okay i move there he moves there... in your head for two moves before moving on the anyalasis board to improve calculation. Than (gradually) try going up to 3 moves than 4...

But drastically limited doesn't mean drastically inferior - in fact, it may just imply the opposite., especially since the players who do invest so much into the game to play in tournaments are more likely more serious, and better players on the average than online players . . .  But I could be all wrong as I've never played in tournaments.

I agree.  I think you could make the case that other than OTB ratings based on 25 or more games, standard live ratings would be the most representative of true chess skill.  Blitz and Bullet have too much ephasis on speed, which is, of course a skill in itself, but not representative of knowledge of the game.  I think of my online rating as my potential rating.  Representative of how good I could become with improved calculation and board vision abilities.