Will computers ever solve chess?

We can't KNOW this, and I emphasised this. However, the analysis of games using very powerful engines is consistent with this.

In response to your likely retort "these engines are not perfect", note that chess is primarily NOT a matter of what was considered good being seen to be bad with deeper analysis and then being reversed again with even stronger analysis, but more a matter of reinforcing and making precise, with occasional refutations. The Ruy Lopez has been viewed as sound for several hundred years and AlphaZero has the same, entirely independent "opinion").

As I have pointed out before, improvements in chess quality are largely about giving the opponent the opportunity to blunder by accumulating practical advantage rather than by refuting established conclusions.

 All of which have no idea what perfect looks like. And also, I REALLY doubt he averages choosing the engine move every move but one per game.

You’re looking at this through the lense of fallible players (humans and computers alike) playing stronger than other fallible players.

Until computers find a mating net (or a draw by force from an otherwise lost position), they merely give a numeric value on a position.

With solved chess there are only two types of “values” for a position - 0.00 or Mate in X. 

Looking 16 moves deep my analyzer assesses Ruy Lopez to be +0.35. You can’t just say, well that’s pretty close to zero so solved chess probably has it at a draw. It very well could be white has forced mate in 500+ moves. 

ponz111, that perfect chess is a draw may (or may NOT) be a fact. But the thing is, nobody knows. YOU don’t know.

Nobody takes issue with you guessing it’s a draw. We take issue with you claiming to KNOW it’s a draw.

And don’t even try that “we don’t really know anything 100% we could be trapped in the Matrix and none of this is true” copout nonsense. Under your dishonest word game we can’t “know” 1 + 1 = 2. You’re claiming to KNOW the outcome of perfect chess and your claim is unjustified. 

The opening phase is actually the most forgiving due to no contact between the two forces, almost zero piece mobility, and both sides being unorganized in general.

Once again, this is through the lense of fallible players who have no idea the ultra-deep ramifications of opening moves. They might make few *numeric* changes to current computer analyses, which don’t apply to solved chess.

We have ZERO knowledge which moves may or may not change 0.00 to Mate in X, or visa versa. 

Incorrect.

Again, no matter how much of a boogyman you make it out to be, on a fundamental level it's still about mobility.

 And I don't care about engine evaluations. I know the opening phase is forgiving without them

Great. Is there a list somewhere of the outcomes of mathematically perfect play - from all the common opening? 

I was under the impression chess hasn’t been solved.

False dichotomy right?

It's not that we either have "ZERO knowledge" or we have all the knowledge.

And sure, you could make some philosophical argument that we might be in the matrix, and everything we think we know is wrong.

Sure, you'd be right. Maybe we know nothing.

But that's getting a little silly IMO.

Great. What knowledge do we have about which opening moves “change 0.00 to Mate in X, or visa versa.”? 

Setting aside absurdities like fools mate or Legals mate. 

I don't know the true evals for each move.

But you said we had no idea which moves may change the true eval. Of course we have an idea. The moves that limit mobility (like moving the king towards the center early in the game) definitely change it for the worse (from win to draw or from draw to loss).

And with mobility as a basis, I think we can argue that the opening is the most forgiving. Each side is unorganized, no threats, etc. like I said earlier.

For starters that’s not what I said.

I said ”We have ZERO knowledge which moves may or may not change 0.00 to Mate in X, or visa versa.”

What we KNOW is that certain pieces more powerful than others, that active pieces adds to their usefulness, that overworked pieces can be exploited, etc. For practical play against fallible players we do know a lot. Computers can look much further ahead than we can. It assignes subjective numeric evaluations for these assessments and chooses the best (subjective numeric) value as its move. A clarification - from the computer’s standpoint it’s not subjective, but the calculation that was programmed into was. Just making sure you don’t try to exploit inexact language.

We DO NOT KNOW if after 3...a6 of the Ruy Lopez if black is trapped in a ridiculously long mating net with perfect play. In fact, we have zero KNOWledge about which normal opening moves may or may not change 0.00 to Mate in X, or visa versa (solved chess). 

"We have zero knowledge of which moves may or may not change . . ."

I disagree. We know which may or may not.

If you'd said "we have zero knowledge of which moves do and do not . . " then I'd agree.

Ok, maybe I sound like a prick, but for example "1.f3 may change the eval" is not the same as "1.f3 does change the eval"

For literally every move for which we don’t know, by not knowing they ALL “may or may not.”

So I guess you just have to decide if you’re interested in having a real conversation or if you want to pick apart semantics. Personally I think it’s more constructive to focus on what I actually mean.

"We have zero knowledge . . . " is a pet peeve of mine.

As for conversation, since we already agree on everything in this topic, there's no much else to talk about.

I know how my disagreements sound. I don't usually take it this far.

I’m curious what lengths you will go to defend a friend.

ponz111 has asserted:

- He’s PROVEN perfect chess is a draw, though not mathematically proven. You have already conceded that only a mathematical proof is sufficient to say it’s “proven.”

- That he can KNOW if a game is played perfectly, in fact- 

- He KNOWS that he personally has made all perfect moves in several games.

- That he KNOWS after 1. e4 e5 2. e6 perfect play leads to a draw. 

I like you but I agree with you on many things, but still if I think you’re wrong I’m going to stay true to myself and say so. You’re not his lawyer. I know you don’t think any of the above are true.