How good is too good

Originally posted by nimzo5
Misssing the point that we don't know Carlsen's CC strength- he could be 2600 or 3100 who knows.

As an interesting aside, GM Ulf Andersson is a rare "dual GM", and the second paragraph below is (to me) an interesting insight as to how OTB and CC play differ in the hands of GMs. I haven't found an example of his CC games yet, but the fact that he plays them in a very different style than his OTB games speaks volumes about the differences in the forms.

"Sweden's Ulf Andersson is well known to the chess world as “The Endgame King”. He has won plenty of games in rook endgames, these normally are extremely long games. He is known as one of the toughest players in the world to defeat, but ironically his best-known game is a loss: a crushing defeat in 30 moves to 18-year-old Garry Kasparov at Tilburg, 1981. After the game he reportedly said, "I will not play against Kasparov any more!" However, in reality they have met in about 20 subsequent chess tournaments, and close to 75% of their games ended in a draw.

At his peak, Andersson reached number four on the FIDE Elo rating list. In recent years, he has begun playing Correspondence chess, quickly becoming a grandmaster in that form of chess as well. In 2002, he reached the top of the correspondence chess rating list. His correspondence games tend to be very tactical, which contrasts sharply with his style of play over the board. Very few players in history have been as successful as Andersson at both forms of chess."

Link: http://previews.chessdom.com/magistral-casino-de-barcelona-2009

No one has really answered the question, lots of people have gone of on different tangents saying a top player wouldn't play in here. If one were to, would he be that good he would come across as an engine user? Would he get banned or would RHP notice the difference between his/her moves and an engines moves.

Many thanks

This is only speculation but I suspect that a GM would have a low matchup rate to a computer.

To test this perhaps one of the computer kinda wizard folks could examine a small collection maybe 6 games of a GM from before the computer era and check match rate i predict it will be kinda low.

I Predict like maybe 60 percent matchup rate.

One thing I wonder is what matchup rate would be normal for a weak FM just 2300.

I predict only about 50 percent in that case.

Disclaimer...This is only speculation but a computer wiz could find this out easily.

Originally posted by Goshen
I don't think one has to be one of the world's top 10 players to be able to beat everyone here every time. I'm sure any grandmaster can do that.

On the contrary. I think even top GM's can be held under control by top RHP cheaters. We have excellent skills on here 😕

Originally posted by brit commando
If one of the worlds top 10 came on here using an alias, would they get banned? I mean because they would come across as an engine user? Know this might get some slating but genuinely curious

There's a difference between strong human play & engine play.
A strong unassisted GM won't look at a position and think "well at depth 20 this move outscores the others by 0.15 of a pawn". They will play human moves following human plans & making the odd human error.

Unfortunately we'll never know just how well Kramnik or Carlsen would do in long time control CC because you can never be certain that they haven't consulted an engine!

What you can do is look at the CC World Championship finalists from the pre-chess engine era & also the best modern OTB players & see how they match up with engines.

If you find a member on this site with little or no real-world credentials who massively out-performs these greats whilst moving daily in many games, you have to ask what reasonable cause is there for them to be able to do this!

Originally posted by brit commando
No one has really answered the question, lots of people have gone of on different tangents saying a top player wouldn't play in here. If one were to, would he be that good he would come across as an engine user? Would he get banned or would RHP notice the difference between his/her moves and an engines moves.

Many thanks

I answered you already: no, they wouldn't be banned, unless they used engines.

good players don't play like engines, not even the best ones. getting banned is not about playing well, but playing engine-like.

Originally posted by Zygalski
What you can do is look at the CC World Championship finalists from the pre-chess engine era & also the best modern OTB players & see how they match up with engines.

Those statistics are useful but it is a mistake to then regard them as a perfect comparison. What we need is statistics for a modern top player playing CC without engine assistance. Ok, such statistics are unfortunately not available but that doesn't make other stastics more valid than they actually are.

The bottom line is that we don't know what a top player of today could achieve in CC with unassisted engine play. Having said that, I do accept that in order to prevent blatant cheating, practical measures have to be taken. And if that means making certain assumptions then so be it - afterall many of the assumptions are indeed reasonable in my opinion. But don't lose sight that they are still assumptions.

Prior to the computer era, a top CC player had a lot of paper work to maintain. Imagine modifying pages of variations with pen and paper. Or imagine trying to find similar positions from previous games. The amount of manual labour has been significantly reduced by software such as ChessBase, etc. So with such changes like those, plus the Internet for resources, I find it hard to believe that a pre-computer CC player wouldn't play better today.

Another big change in the last 30 years is the importance of dynamic, concrete play in chess. Again, I don't believe that this hasn't impacted CC play too. A big factor here is the ability to calculate concrete lines. Hence, the luxury of being able to do so by moving the pieces and maintaining in ChessBase is even more of a benefit.

I agree that many players have been rightfully banned for cheating. Some cheating is simply too blatant for other explanations. But for more borderline cases, we should keep in mind what assumptions are being made.

Let's not forget that these "RHP greats" banned for cheating also play far more engine-like chess in large non-database move samples than modern Super GM's who have been tested.
If anything this site & others are deliberately conservative in their approach. It's just not too good for business banning borderline cases where legitimately strong players would be deterred from signing-up.
For each player they do ban, I reckon they're only skimming the surface & getting rid of the most idiotic cheats.

By the way, on another site someone (not me!) posted the following, which pretty much sums things up:

It's clear that there is not a methodology available to us that is perfect:

* Using pre-computer era correspondence games exclusively as the baseline, it can be argued that the theory is stale, and allowable tools such as databases have developed to the point where comparing current C.c games to old correspondence games is not an apples to apples comparison.
* Using OTB games exclusively as the baseline has a similar counter-argument, plus the argument that additional time may lead to higher matchup rates (despite the evidence to the contrary from the pre-computer era correspondence baseline).
* Using any non-OTB game after the computer era as a baseline is simply subject to contamination from computer assisted games and, rightly, should be avoided.

Some of these issues simply can't be overcome, but the argument that in the absence of a perfect methodology no conclusion should ever be drawn is a non-starter.

There has to be a pragmatic middle ground if you want to tackle this problem. Because of these challenges, simply jumping up and down and yelling "Imperfect! Imperfect!" is not at all constructive. What would you suggest as an alternative to help overcome some of these issues?