Solving chess

Edit: Oh whoops, that wasn't actually Elroch... Sorry Nytik! (noticed just before your following post). But I'm going to let my post stand as a testament to my silliness and because I make some good points in it despite my complete brain-fail.

Well thank you for at least trying to reread my post! I'm sorry if you have trouble with reading comprehension. I didn't honestly think I had to simplify everything down for you to something a kid can understand. I think you finally get it now, but let me just make absolutely certain.

1. With the game of chess solved, and with perfect play, there would be no "advantage", merely win, lose or draw, and indeed, mistakes are the only way these evaluations can change.

2. When you said advantage (and specifically not "forced win") to ruby7, clearly either you were talking rubbish, used the wrong term when you meant "forced win" (which would be a strawman, since ruby7 wasn't talking about that), or you didn't mean this type of solved game concrete analysis, but the human/imperfect-computer variety of advantage. I'm assuming you used the wrong term and attacked a strawman (partly because you seem to like beating up the poor things), which yet again provides some lovely irony, since you're saying I'm getting my terms mixed up ;)

3. The human (and computer) variety of advantage can indeed evaporate with what we have no reason to doubt is perfect play. In fact, many positions where humans evaluate an advantage for one side may well turn out to be forced wins for the other side, but that's beside the point.

Get it? Got it? Good.

Now, I have no idea why you're yammering on about 10 rating point bands, since the data I gave you was for 100 rating point bands. Give the strawman a break! Try actually addressing the points I raised - may the flying spaghetti monster forbid that you might actually agree with them, or even worse, have to admit fallibility. Or hey, point out a flaw in my logic, if you can find a real one!

EDIT: Seriously, stop making posts while I do!

EDIT AGAIN: Points re-edited back in, OK.

Don't know if you caught this, my post was made at the same time as yours above it and hence your bands were different then. My apologies.

My bad - I was in the process of editing my post, apologising for that, when you posted!

(yes, deleted by accident, managed to go back in my browser to when that comment still existed and recopied it back in, I don't shy away from admitting when I make a fool of myself :P )

@Schubomb, I repeat that in a finite game of perfect information (eg chess), the concepts of game theoretic advantage and forced win are identical. Fully understood examples are the game of noughts and crosses or an endgame tablebase, where (I hope) everyone would agree on this point. The reason I used the term "advantage" is because for a hypothetical  computer that could solve chess the evaluation function and the result with perfect play would be the same (just like the evaluation function in an endgame tablebase).

However, I understand your confusion, because the term is one used in a less precise way in practical chess (nothing to do with game theory). But since we all understand the concept of a perfect game and the theoretical result of a chess position, and no further knowledge of game theory is needed for this discussion, let's not waste more time.

With regard to the statistical significance of the sort of results we are looking at, experience gives me some intuition (and some useful rules of thumb) for seeing when something is more than a fluke, but we are already on very dangerous ground doing any sort of significance testing on the data we have looked at. The reason is because in statistics, a very important principle is that you create your hypotheses before you look for patterns in the data you will use to test them. (To see why, imagine there are 100 patterns each of which would occur 1% of the time by chance. You look, see one of of them, find out it would only occur 1% of the time by chance and then claim it is statistically significant). So we need to propose a hypothesis and use some independent data.

We might be able to use computer-computer games. There is a fairly large public domain database of these, and I have not ruined its usefulness by noticing any patterns. Unfortunately the reason I have not done so is that the database contains no ratings for the computers concerned, and entering the ratings of 3000+ computers is something I am not going to be doing.

So best to make a hypothesis now, and wait for the next 10 years of top level chess games to provide fresh data.

The opening position is one of millions of balanced positons where whoever plays first has a very slight but not winning advantage. Computers have demonstrated this fact, as they play by position evaluation, always striving to gain at least an even postion.

In the majority of balanced positions their are a good number of possible moves, just as in the opening itself.However some positions are knife- edge and can be lossed with one slip.

Who can say what the best opening move is ? Answer there are a number of equal choices.

No Elroch, it was not my confusion about the word "advantage". Read the post you jumped on from ruby7. Did ruby7 say "advantage, in a game theory sense"? "Advantage, in terms of concrete, win/draw/loss calculations with perfect play"? No. Just "advantage", and you misinterpreted that. No-one will judge you for admitting fault, you know, it's ok to be wrong sometimes (especially someone as ancient as yourself) - as you can see from this thread, it's sometimes even funny!

As for statistics, well, beneath all that terribly interesting text, I'm sure there's a tacit admission of error there. You're welcome :)

And you don't even have to ask, I forgive you for being a pretentious, pompous, patronising... um.... person!

OK, I admit it, this is mainly to poke fun at Elroch and rile him up, and see if he has a sense of humour at all. I'm betting: no.

@Schubomb, you are being an annoying child. That's not patronising, it's a statement of fact. Imagine having to deal with a 6 year old who refuses to believe you know more than him about something, and you might see things from my perspective.

I was rather blunt with ruby7, who had not deserved it. But his post had a number of factual inaccuracies which were worth discussing (eg "Advantages cannot be maintained", "Theoretically, there is only one possible perfect game" etc.)

I was right - highlight at the bottom of my last post to see why ;) You need to take yourself less seriously, Elroch!

I like that analogy - see if you like mine! Imagine having to deal with a senile 90 year old who refuses to believe you know more than him about something, and you might see things from my perspective.

Actually, since I'm over 4 times older than your analogy, maybe I should proportionately return the favour: imagine having to deal with a 220 year old! I can tell you, my grandma has dementia but an indefatiguable ego, and she's nowhere near 220... it's a real problem, man!

Oh, and yes, the other parts of your post to ruby7 made sense mostly, which is why I didn't make a comment about them. *shrug* You have a knack for non sequiturs.

Unless chess is solved, it's impossible to know whether perfect play results in a draw, or a win for white.  Personally I think it is more likely that the result would be a draw.  For example, if a computer was limited so that it could only see 2ply into a game of noughts and crosses, then the first move would appear to be an advantage, however we all know that this is not the case.  I think the computer is most likely unable to see deep enough into a game of chess to realise that white has no advantage with perfect play.  Also there are countless examples of dead-drawn positions in which computers give an advantage to one side or another.

Of course this is pointless and unfounded speculation.

Anyway, I believe we will know the answer within 50 years or so, judging by the rate at which processing power is increasing.

This is crazy! Dude, I saw you were from Adelaide, looked at your profile, recognised your name but didn't know where from, checked my emails for your name - I sang your composition in that workshop, with Lumina. Though your piece was annoying for me, I was on bass and only had a low G for the entire piece, I think.

I don't think 50 years will be enough, though, or if we will ever solve chess. I'm open-minded to the idea, though. And I would laugh my head off if it's solved and it turns out white is in fatal zugzwang, and it turns out to be a forced win for black!

Amazing, I've run into people from Adelaide on these forums more than once now, and never anywhere else on the net.

Sorry about the low G by the way, I have a drone fetish. :)

It worked for that piece well, so I can forgive it :) and it's my own fault, if I was better at vocal harmonics, then I would have had the more interesting part!

@Schubomb, did I really misread your year of birth as 1995 and not suspect otherwise for the whole discussion? If so, I have made exactly one mistake during this discussion. It seems unlikely that you would change it...!

I agree that 50 years of improvements in computing power are unlikely to be adequate to solve chess. There are various reasons for this. Even if the trend in speed continued, it would probably not be adequate for the task. In addition, radical new technologies are necessary to make much progress in speed - silicon is reaching its practical limitations. Recent processors have become increasingly parallel rather than much faster, and parallelism can only go so far. I've speculated that quantum computing will solve chess in this time scale, but people more knowledgeable than me on the subject have said there are reasons this cannot be achieved.

Hm. Well. Dammit, this is what happens when you get a cold, can't breathe at night and can't go to sleep: the desire to show the magnitude of someone's mistake. I prefer the gentle mocking approach, but clearly you need the sledgehammer.

1. Failing at basic date reading. Probably some associated prejudice based on that. Maybe you have some kind of "must put that blasted smartass teenager in his place!" complex. Whatever.

2. Assuming without any justification that I'm not the most knowledgeable person here. I'm not saying the mistake is your conclusion, which may well be true, but the mistake is your lack of valid reasoning. Even were I 15 years old, you've shown no valid reasoning.

"At the moment you are a very long way from being the most knowledgeable person here or the best at chess"

3. Telling someone that they misused words when they were used appropriately.

"I suggest you look up the meaning of the words irony and hypocrite in the dictionary, as you have misused them."

4. Saying that someone insisted they were right when they did not. This one may not be your fault really, since you seem to have extreme difficulty in understanding nuances in text, and I gave the appearence of insisting I was right unless one "read between the lines". Have you been tested for Asperger's Syndrome? Or maybe you have had delayed social development.

"Note that my bluntness was a response to you insisting you were right when you were not."

5. Ignoring basic logic: someone is the most knowledgable here, and someone has to have the most chess skill, except in Elroch-world!

"I am confident that all four assertions in the paragraph you have emboldened are precisely true for everyone here not just you." (regarding one of the previous quotes about my knowledgability/chess skill).

6. In your (admittedly fascinating) observation that 26xx players seem to score fewer draws than the "the better players are, the more draws" theory would indicate, you failed to account for the even higher rated players who resume the expected trend (or actually got worse as white), and that the 26xx players, while interesting, are clearly only a minor anomaly. And refused to budge on this position until it was proven undeniably invalid, and then only tacitly.

7. Failing to notice that ruby7, when talking about the advantage, clearly didn't mean in a game theory, concrete calculation sense. Admittedly, I made the complementary mistake soon after. Not in such a way as would make my post incomprehensible though:

8. I think once Elroch saw this mistake, he couldn't bear to wait another second before hammering away on the keyboard, typing up an explanation of game theory and why I was wrong. Couldn't wait even until reading the end of my post when I make it clear that I understand perfectly well that in concrete analysis of a position, there are wins, losses and draws, and only mistakes can change that, when I post:

"It comes down to semantics. Advantage means nothing ultimately: as you said, with perfect play, every position is either a white win, a black win or a draw. But until we can demonstrate perfect play and have a 32-piece tablebase, advantage is all we have, and those do indeed seem to often disappear even with play of which there is no evidence of its imperfection (and may well be imperfect, but that's just my point: until we prove it, the while win/draw concrete evaluation falls flat on its face."

I guess Elroch just got hung up on terminology and refused to budge from being hung up for a while.

9. Failed to lighten up when I poked a bit of fun at his complete stodginess and unwillingness to budge. Ok. I admit, this mistake is subjective, but come on, everyone needs to be able to laugh at their own silly mistakes and pride (both of which Elroch has talent in exhibiting) every now and then. I mean, I thought at least my alliteration would provoke a smile, but nothing! What a tough crowd.

10. Underestimated the amount of mistakes made during this discussion by an order of magnitude. (ok, fine, only by a factor of 9 if you ignore the self-referenced mistake, though I've been very generous in combining mistakes into one, and letting some questionable ones slide).

Amazingly, this line probably was their attempt at humour! I give some points for that.

"If so, I have made exactly one mistake during this discussion."

Dammit I need a prescription of pseudoephedrine. It's making my style inconsistent, it's a crazy mishmash between talking directly to El, talking to some kind of generalised 3rd person audience and talking to myself. All nine of me. Blurg.

@Schubomb (I can't imagine why anyone would choose a username with such associations), ad hominem attacks (even fantasy ones) are considered a very low form of behaviour in forum discussions, and are banned by chess.com. You won't get a second warning.

There is no purpose to continuing this discussion, whose factual content was interesting before it got waylaid by your defense of your ego. The sad thing is, you are not arguing to deepen understanding, you are arguing because you felt snubbed. I am sure you are an excellent musician (a subject which I respect): you shouldn't be concerned about not having studied mathematics or game theory to the extent that I have. Any more than you should be concerned by having an online chess rating 600 points lower. That's not a snub, it's just the truth.

This is very long so I did not read.  I am so sorry!

You are wise, AndyClifton.

Thank you very much Mr. Elroch!  And I also do not understand what is Shoe Bomb.

Really? Perhaps google would help.