Jonathan Tait wrote on 12/18/22 at 09:17:25:



There is no world where an engine beats a centaur; that has never happened and never will happen. However, many decisive games are lost due to human errors.

The easiest way to think about this is type I vs type II errors. By making one more likely, the other can't occur.

Kerangali wrote on 12/09/22 at 16:25:11:

Always computer helped by human or commonly referred to as centaur.

cathexis wrote on 12/10/22 at 01:34:09:



What a weird boast given many more made the IM Title in pre-engine days when it was far easier to achieve titles than the engine era. Case in point, Jon Edwards earned not just the WC title but the GM title at the same time (as this is practically the only way to get the GM title these days).

bragesjo wrote on 12/10/22 at 10:10:46:


This is actually correct; a sign Tim Harding isn't with the times I suppose.

It has definitely gotten way way harder to win games but ultimately interesting ideas can still be found as I am often finding myself in endgames up a pawn or an exchange but there still is no clear path to victory because the engine is so strong. Still, these games are interesting and keep correspondence fresh.

I would argue that it is not the draw rate which will kill correspondence chess but the rating deflation and the rudeness of player's using deadman defense. Right now I have a game against a 2469 which has been known as a draw since the opening and I have offered him two draws over the last two years. He just keeps playing in hopes that I will die and he can be declared the winner when I flag.

About Engines and horizon effect, in the past 3 years the horizon effect has been pushed back by more than 30 ply in standard position and the double in some positions.

But like I said before unexplored lines are key to win.

Well, the thread has been interesting. 

I am most grateful for the discussion regarding the latest World Correspondence Chess Championship 
That was a clear example of the thought I had 
Unless outside factors are involved, all games will be drawn. 
And it is exactly how it happened. 

Stockfish vs Stockfish , version 15 is going to be a draw. 

Ruy Lopez - draw
Giucco Piano -  draw
Queens Gambit declined - draw
English 1.c4 e5 draw

Many lines do not need the human at all. 

To me, that marks the death of this. 

Now, regarding OTB... 
I have known that these devices exists and I know they are not detectable by the screening they have now

Those who know how to cheat very well, they will be careful
They will wisely know when to use them and when not to use them 

Nieman was a case of greed, too much greed, and too much desperation to make the climb a bit "Faster" in the ratings 

Such carelessness got him caught or under the vast amount of radars and scrutiny now. 

Those who are cheating and are wiser than him will be content with making a climb that seems more normal 
They will know to make the climb while at the same time be aware of their performance percentage 

Niemen was either ignorant or he did not care one bit that his performance was going to stand out and be a red flag

He did not care that his performance rating got hither than Fischer's and Kasparov and let himself have a game with 100% rating (perfect) 

Those who know how to cheat better will use those devices under better calculated times and conditions. 

I am saying this now --

If FIDE does not implement the much more effective means of screening, this is only going to continue 

Nieman was just an example to the others who are  using devices of how not to do it , how not to cheat is important for them.  

I am stating it like it is

Please excuse me for painting such a gloomy picture but I call it and it is important that I speak without regard if the picture is gloomy, if the goal is to state how things are instead of how I wish they were. 

It is time for FIDE to recognize and be much smarter than this

It is time to do these "uncomfortable" checks or else there are going to be more victims and more perpetrators 

World Correspondence Chess is a joke and a waste of time now for me

The last event was the strongest reason for me to feel this way.  The results were exactly what I thought was going to happen
Again, if you get rid of outside factors, all results will be draws.

I can add a slight comment about Correspondence Chess titles. Most higher titles are taken before computer Era or at last years before todays standard. The lower CCE and CCM titles can still be taken but higher titles the norms reqeured 3 or 4 wins vs extremly high rated oppositionen so norms reqeurements has not followed the increased draw rates. Its still doable since players still takes norms but its far more difficult.

The elo system is designen how game results was in 1970-1980 and drawrates between all intervall are higher today sort of inflation and deflation at the same time.

The last "passing of chess" was in Lasker and Capablanca time, when they figured each game could be diminished to a sort of QGD with play and counterplay decreased to nothing and dead-draw rook endings.
Presently we're confronting the following neighborhood ideal: when old style PCs absolutely overpower people in all pieces of the game, human impedance will be, best case scenario, pointless. Without a doubt corr chess will be inconsequential, however OTB chess can continue essentially until it's typical for people to have inserted chips or point of interaction.
Anyway PCs will propel chess hypothesis: today none has demonstrated that any chess opening is a draw, just that their down to earth strength permit them to draw against human and friends. Additionally people have incredibly profited from PC disclosures.

Further development can be anticipated if one day quantum PCs capable of running Grover's calculation transform into the real world. This will decisively extend search and assessment. Desire to see that eventually.

The last "death of chess" was in Lasker and Capablanca time, when they thought each game could be reduced to a kind of QGD with play and counterplay reduced to zero and dead-draw rook endings.
Now we're facing the next local optimum: when classical computers totally overwhelm humans in all parts of the game, human interference will be at best useless. Indeed corr chess will be pointless, but OTB chess can persist at least until it's commonplace for humans to have embedded chips or interface.
However computers will advance chess theory: today none has proven that any chess opening is a draw, only that their practical strength allow them to draw against human and peers. Also humans have greatly benefited from computer discoveries.   

Further advance can be expected if one day quantum computers able of running Grover's algorithm turn into reality. This will dramatically expand search & evaluation. Hope to see that in due time.

cathexis wrote on 12/10/22 at 01:34:09:

At Usain Bolt's preferred distance of 100 meters this is true. At ultra-marathon distances, since you said _never_, it's not true. The horse usually wins, but not every time. 
https://ultrarunninghistory.com/man-vs-horse/