Yes, there are games where both players are 100% accurate. For example short games where a draw is agreed. I think that your question is answered by the OP very early on in this thread, on page 1 or 2.
True or False Chess is a Draw with Best Play from Both Sides

The answer to the main question of this forum is that it is obviously true that chess is a draw with best play by both sides. However proving this with math has not been done.
If it's obvious, then proving it should be very easy.

...
It's possible that as chess engines become better and better in our lifetime that they will reach a maximum ELO where it's no longer possible to play much better any more and that all the games between such highly rated AIs will end with draws.
That's a good way of putting it. It's been posited by at least one noted statistician that the highest Elo possible in chess is just below 3600. That is, today's top players can draw at least one game out of 50 against "God".
Well. Current top engine are getting near that elo. Not much room for improvement? But a top GM will still get 3 or 4 draws from ten. Not a win though.

The answer to the main question of this forum is that it is obviously true that chess is a draw with best play by both sides. However proving this with math has not been done.
If it's obvious, then proving it should be very easy.
I had a terrible time in geometry because the whole focus was on proofs for the obvious, such as the assertion that parallel lines do not intersect.

Wow, last time I posted here I was running that little experiment. I never really gave this thread much thought.
At a certain level, a draw in chess is easily achievable. In fact, I remember Fischer saying at one point after losing a tournament, "he was playing to draw!", or was that just the movie... So yeah, mathematically, that very fact increases the probability of chess being a drawn game under perfect play.
Fortunately that was never the purpose of the game or somebody would've surely introduced a rule to force a win out of every match. You see (again, I'm going to have to quote the man here), Fischer once called chess the search for truth. Let's think about that for a moment:
Chess has no opponents. There are only two partners, black and white, testing the very boundaries of mathematical precision. So a draw was never an option. Chess is to foresee the draw and avoid it all costs, it is to keep at precision until the equations break down at a certain point...
WSama

The answer to the main question of this forum is that it is obviously true that chess is a draw with best play by both sides. However proving this with math has not been done.
If it's obvious, then proving it should be very easy.
I had a terrible time in geometry because the whole focus was on proofs for the obvious, such as the assertion that parallel lines do not intersect.
Well there is no proof for that, so it probably wasn't.

Just because something is "obvious" does not mean it should be easy to prove. It was obvious for many
decades that checkers was a draw with best play before this was proven by math.

Cough coguh, you don't need 32 man tablebase to prove that chess is a draw or a win for either side.

To *prove* it mathematically, it may take a 32 man table base. But the evidence based on games played, engine and GM analysis, and so on all point to the game being a clear draw.

Which is? And is this one thing stronger than future computers?
no there is something stronger than the best chess engines today!
I will answer...A strong human player guiding a very strong chess engine is stronger than a very strong chess engine. [this has been proven]
And even that is a far cry from perfect play, the only perfect play we can do is mate in X, dead draws (KK for example). Anything else becomes unclear in what is the best move, hell, even top players disagree on that.

To *prove* it mathematically, it may take a 32 man table base. But the evidence based on games played, engine and GM analysis, and so on all point to the game being a clear draw.
Making a 32 man tablebase would solve the question, however, it is not a necessity by any means, ie, we know it won't be requiered. A Tablebase has much more information regarding the game state of a position and includes suboptimal moves information, mating distance, etc.
A good example is checkers, it is proven to be a draw but there's still no algorithm or tablebase to tell you the best move in X position.

Which is? And is this one thing stronger than future computers?
no there is something stronger than the best chess engines today!
I will answer...A strong human player guiding a very strong chess engine is stronger than a very strong chess engine. [this has been proven]
And even that is a far cry from perfect play, the only perfect play we can do is mate in X, dead draws (KK for example). Anything else becomes unclear in what is the best move, hell, even top players disagree on that.
I think you don't quite understand what top players are disagreeing about. They don't disagree that the chess is a draw with best play. They usually agree that there are many ways to approach a position. They usually agree that certain moves are riskier than others, and almost always agree when a move is just bad.
Top chess players are very objective. Take a look at Fischer's brilliant My Sixty Memorable Games. His objectivity was a hallmark of his chess writing. When he made a mistake, he owned up to it. It's sad that he wasn't objective in the rest of his life.

Which is? And is this one thing stronger than future computers?
no there is something stronger than the best chess engines today!
I will answer...A strong human player guiding a very strong chess engine is stronger than a very strong chess engine. [this has been proven]
And even that is a far cry from perfect play, the only perfect play we can do is mate in X, dead draws (KK for example). Anything else becomes unclear in what is the best move, hell, even top players disagree on that.
I think you don't quite understand what top players are disagreeing about. They don't disagree that the chess is a draw with best play. They usually agree that there are many ways to approach a position. They usually agree that certain moves are riskier than others, and almost always agree when a move is just bad.
Top chess players are very objective. Take a look at Fischer's brilliant My Sixty Memorable Games. His objectivity was a hallmark of his chess writing. When he made a mistake, he owned up to it. It's sad that he wasn't objective in the rest of his life.
Ermm, I didn't claim they disagreed on the result of perfect play, I only said that they can disagree on what are the best moves or the game state of a position.

What I am saying is that in the vast majority of positions, top players will agree on the evaluation. There is only a very small per cent of positions where they would disagree.

What I am saying is that in the vast majority of positions, top players will agree on the evaluation. There is only a very small per cent of positions where they would disagree.
Correct, my point was that perfect play from top players is not even a thing because they all play different things and their opinions can diverge as you comment.
are there any games where boths players are 100% accurate or does one player always make a mistake?
You would have to ask the "CAPS Masters" here.
The 1000 players that post their games with a CAPs score of 97.5, and ask if the played like a GM?