ELO Rating of Chess programmers

[elcabesa]

decide yourself whether my program , Vajolet , is original or not but my Elo rating is now lower than 1500 point

[Henk]

If you cannot understand the moves of your chess program anymore maybe it is better to give the chess program less thinking time when testing manually.

[Uri Blass]

One example on the other side: Vincent Diepeveen's Elo is around 2300 and it's hard to gauge Diep's strength because it's not publicly available, but it did well in many tournaments (e.g., 3rd place in WCCC 2004).

Diep did not do well relative to other programmers with clearly lower rating(programmers of Junior or Shredder).

I don't understand your logic. If I had said that Vincent is a strong chess player and his engine beat all the engines written by weaker chess players, your criticism of my statement would make sense. But I made a very limited statement (there is one strong player that had a strong engine), which I think is well supported by evidence.

My point is that
I do not define diep as a strong engine and I think that it was an average engine.

[velmarin]

I do not consider myself a strong player,
30-year federal license and elo 2000 (FEDA, Spain),
I think to move in this world is very interesting hobby and have some knowledge.

Although great programmers have contributed a lot for your programming knowledge. Great technical solutions but with some lack of salt ...
I guess you can make a good program that provides a good code and then a good tune, all from a mathematical point, scientist, etc..
Then ask yourself:
if businesses how Microsoft put its programmers and machinery in a project,
that elo would get a engine.

[cdani]

Hi!
I think it helps a lot to understand more about chess. It's easier to evaluate the games of the engine. Of course you can simply use statistics and sure will be enough, and specially taking into account the incredible strength of the engines. But I like a lot to understand and enjoy that it's playing better.

It is not unusual for a strong player hold his reasoning on what has given him the strength. Is as usual as anyone to lean on what he knows. Hence the merit of either to see something different.

So for me the raw statistics and also the knowledge of chess are very valuable in developing the engines.

[Henk]

The ELO rating of the programmer has nothing to do with the ELO rating of the program. I am a weak player, yet DiscoCheck is much stronger than me.

In fact, I would go further and say that being a strong player is an impediment more than anything. Strong players always have this arrogance to believe that they know what is right and what is wrong. They don't have the humility to understand that everything is a trade-off and a probabilistic approach. When testing results contradict their dogma, they prefer to trust the dogma and discard the evidence...

The advantage of a weak player is that he doesn't get demotivated when his chess program is playing bad or awkward moves. For he doesn't know it.

[PK]

Speaking from experience: even a 2750 (CCRL) engine can make moves that make a 1850 (FIDE) player squirm. Noticing it is an easy part. The hard part is to turn this impression into something that doesn't decrease strength, while improving "visual quality" of the moves. I had to resort to a couple of funny solutions that I wouldn't come across otherwise (doubling isolated pawn penalty on 3rd rank; 64 bitboards of pawns that make a bishop bad).