I asume: they will reach 4000 ELO.
But there will be the Question: what is such a '4000 ELO'-value.
It is a calculated value, based on a fiew thousand Engine games. OK.
And no one will be able to say, what this has in detail to do with the strength of strongest humans.
"Much stronger, anyhow!" OK!
"We had a strong storm, much stronger than normal Beaufort 12. It must have been wind-strongness 15!"
It's just a time problem .
So, for a 4000 Elo perf., a player would need to score 96% against opposition with ~3500(!!) average, or 99% against 3323 :-)
Against good opponents, even for a vastly superior engine it is unlikely that such percentages can be achieved, due the draws. For example, 96% requires total results like +23 =2 -0. For comparison, In 1105 CCRL games, Quad-Rybka 3 has drawn 403 games. And that is already a small share of draws, relative to the next competitors.
Engine rating lists have many opponents with very high ratings (far beyond the top human ratings, already). At CCRL, Quad-Rybka 3 scored 76.9% against an opponent's average which was 190.1 points lower rated =3035.9, and got 3226. (The rating system there apparantly assigns slightly different preformance differences, but very similar.) If the opponents became clearly stronger, like 3100, then 85% would give ~3400. So, both with even stronger opponents AND an even better percentage which a new version might achieve, 3500 seems out of reach, let alone 4000...
Sincerely, I could get many times draws(3-6% in past tense against Rybka 3 and this is too much finally for this kind of experiments of dreams of machines in a way of tottal perfection), against these super machines. I dont have any elo. then .......... This (4000 elo) looks not possible. Only we could answer to the next sincere and pragmatic question: If could I search 3-6%, playing Rybka 3(what a nice machine I really love it !!!!), ...then.....What Rybka 3 elo is really? ...just as example.... not simple.....
Nevertheless: If your estimated strength is somewhere like ~1900...2000, it can be said that by playing effectively against this particular engine weakness (almost disregarding which particular engine, with few exceptions), you perform several hundred Elo points above your general strength, in these specific games. From my viewpoint, this is a fact.
Maybe we shouldn't talk this further, to avoid the usual uproar, rage and hysteria :-D
................. and the 3%-6% target as score for human being side, my own case, probably could be better, why not be at least 10%
or more (I would like to have time to seartch the 51%) ? In fact I have not worked hard in this last way, after Rybka 3 came to the market.
Of course than Rybka 3 team have done an excelent work, improving the Rybka performance fighting against the stone wall technique
in man versus machine. What I woulld like to show us, is simply: A perfect chess machine looks as a perfect dream now. It is only into
the imagination. I dont know any machine in the Earth, thaty could have a perfect score in a long long challenger in man versus machine.
Then, never we could find a machine over 3.550. But I would like to go far away: There is not a computer machine that could have a perfect
scorfe playing against a simple amateur as I am, or as "Javier Rubio is" an Sapain man, with an elo, not better than 1.300. Then there is not
a machine that could have an score over 2.050. First of all, it is necesary to have a chess software that kill the anticomputer style. I
suspect, it is not possible of all.
> Even if a 32 EGTB chess entity was available it would not score 4000 ELO
Would there be a draw between a 32 EGTB chess entity and, say, a 3200 Elo engine? :-)
Wihle I agree there must be a wall and that many possibilities may occur to draw a game the number of 3400 seems low to me.
Already today on a simple Q6600 Rybka 3 should have an ELO of around 3200.
If we take next Nehalem 8 core machines with 2 sockets, that is 16 cores.
I assume such a machine OC is around 5x faster than the Q6600. If doubling the speed is around 60 ELO => around 150 ELO gain => 3350
Do you think now if Vas improves Rybka 4 by 50 ELO he can stop programming and retire because the wall is there ?
I dont think so and dont think we should underestimate chess.
my figure would be around 4000 ELO because we today have no clue what does it mean...
Is just a guess though.
PS. I forgot to mention in 40/40
PS: Vas can retire even today! :-)
r3 almost certainly isnt 3200 against humans.
"If doubling the speed is around 60 ELO..."
going fwd, the 2x of speed will almost certainly not produce gains of 60elo, at least not at 40/120 time controls.
"Do you think now if Vas improves Rybka 4 by 50 ELO he can stop programming and retire because the wall is there ?"
no, what we'll see is that the top progs will all soon reach a death draw at 40/120, but there are still huge gains to be made at faster time controls.
"Q: What is God's ELO rating?
A: Several threads inquiring about the ELO of God have been created during the life of Rybka forum, even, the topic has surfaced on unrelated threads because this information is very valuable, so most people want to know. To stop the madness once and for all, possibly because He is very busy making miracles or creating/punishing things and doesn't have much time left for practicing chess tactical quizzes, God's ELO has been measured to about 1600 by Alan."
Anyway, Uri Blass's arguments were very solid, on that iteration of these discussions I even got to switch sides and be the bad guy. Later on we proved logically that God can change the future but not know it (just like us!) or know it but not change it (e.g. He can know that he's going to lose the game but not be able to do anything about it), but he can't know it and change it (because then he didn't know it in first place), a requirement for Magic Moves®, the only way he can ensure 100% score against non-perfect opposition, without the 100% we don't know how easy it's to draw against him, if God is a she perhaps her interest on chess is very small and we could even beat God, so ~1600 sounds about right. I also like that I can say I have Godly strength at standard time controls :)
I still propose we post this image each time the topic pops up:
It is easier to use his trademark phrase: I disagree... :-)
> It is easier to use his trademark phrase: I disagree... :-)
what about the following?
...that is not correct
\'jer-ō-nd\ (listen) http://www.merriam-webster.com/cgi-bin/audio.pl?gerund01.wav=gerund
What you just were, when finding yourself lost after your opponents opening book ended.
Blassing, Blass, Blasser
\ˈblas-d\ (listen) http://www.merriam-webster.com/cgi-bin/audio.pl?blast001.wav=blast
Feeling of self-doubt and slight resentment after being caught by inexactness pointed out by another.
> he's a very valuable member of the community and one I haven't been able to win an argument against.
I hope master Blass gets well vey quickly, we miss posts.
As for inherent playing strength (which is probably at the very least, a highly debatable idea); In my opinion it begins to asymptote rather strongly at 3200 fide.
Ideas on 'god's chess':
1. Perfect chess vs. Optimum chess
2. What is needed for the chess "solution"
3. Issues with the elo system
4. Different elo systems, like for example, some people like to quote ccrl numbers as opposed to fide numbers.
Common topics with useful information that has more or less been established include:
1. When Rybka behaves weirdly, slower hardware or during same system engine matches
2. Issues involving engine vs. engine ratings/matches, that is, conclusions reached from forum discussions and/or other sources.
3. Perhaps some information for a buying choice between Aquarium vs. Chessbase GUI.
4. Positions that Rybka 3 tends to handle/evaluate incorrectly
5. How to access the "vas" cam which gives you live video of where vas is and what he is doing at all times*
*This suggestion may only be for comedic value
this is assuming 40/120, 20/60, 30-min time controls, w/ no opening book restrictions.
You now know something hard to believe for all of us, right?
Okay, go on, keep it in the secret room.
But please release it before 2012 at least a year!
3400 CCRL is the limit (maybe 3430 on a good day)...
Or do you think if CCRL offers an opponent average of 3100, a future Rybka can score more than 85%?
> On which scale?
On all (reasonable) scales.
> Or do you think if CCRL offers an opponent average of 3100, a future Rybka can score more than 85%?
Easily - that's not even a tough question.
never mind the fact there are ridiculously huge (private) books that pretty much have everything under the sun covered.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill