I played a few games recently that one of I won and the other I resigned, perhaps prematurely. I noticed something confusing to me with the analysis. Having said that, I am trying to not only take into account chess.com's computer analysis, but also the nature of chess theory.

I find it difficult to summarize what I am getting at with regards to the nature of chess theory and how it plays into this situation, but I have never seen games before these two that had no serious mistakes by either player right up to the end, before either it was over, or an advantage to so decisive was reached, that it wasn't going probably matter. I guess I am trying to decide if this is more due to chess.com's computer analysis or if it is more due to the nature of certain lines of play.

I guess it is seems illogical to think there are lines that are somehow considered to be basically about equal, as neither side makes much of a mistake, but suddenly it is over without a blunder? On the other hand, if the system calls the Sicilian defense an inaccuracy while calling the King's Pawn game sound, maybe it is just the system, but I can't help but ask is it somehow both, but to what degree ?

Here are the games:

You write so sophisticated I find it hard to summarize the nature of your question, but Ill try to answer.

In the opening take the engiens analysis with a pinch of salt. No matter how strong the engine will be, it cant compete with more than a century of developed opening theory. Human practice and experience in those lines will be a far better evaluation of the theory. If a computer calls an well known opening move a blunder, just ignore it and trust the theory.

In the middlegame, the engines evaluation is often very good, they will more often than not find moves that are stronger than humans, but still they offer lines that can be so absurdly complicated that they make no sense to a human, or they will offer lines that are just "strategicly" bad, because engines really dont understand the concept of strategy. That being said, a drop in evaluation from +0.6 to -0,2 is quite significant and will nearly always means you missed something in the position.

in the endgame humans will often have a better understanding of the position (well atleast GM's do). Computers can say who has the advantage, but they dont really evaluate drawish position very well. You might often see a clearly drawn endgame being evaluated as +1 or more. This might lead to the engine giving a certain move to be a blunder, when really it was just avoiding getting into drawish territory. This is how I see it anyway.

I guess what I was getting at is that there are the opening book inaccuracies, that of course I was throwing out, but that left so few mistakes of any degree. It just seems that the level of play was rated as high by the analysis computer, as compared to the elo of the players. I guess I am not used to seeing games with so few mistakes of any kind and the decisive moves not being blunders at the end. It makes me want to see how the Chess.com computer analyzes GM level play. It makes me realize that GM's probably laugh or just shake their head when they look at Chess.com analysis...

I don't understand your point about blunders. Both of those games ended with very big blunders indeed.

The game I resigned did have a blunder just before I chose to not play with my king so exposed without an material or time advantage, despite it, the only mistakes I made were really minor ones. probably a 1/3 were in the opening and don't count.

While I would have thought that black's last move in the first game that I won should have been called a blunder, it was only classified as a mistake. I think that mistakes shouldn't cost you a loss. I guess in general, in either direction you look at it from, many of the moves that were labeled by the computer analysis were questionable as to the true level of play that they represented.

The chess.com analysis engine is notoriously innacurate. Any move that invites an otherwise avoidable mate in one is a blunder.

I guess another thing I was trying to get at was that with such analysis, it makes it difficult to tell if you played well/correctly or if your opponent was bad and vice versa...

Hopefully they will improve the analysis service with time. My guess is that games submitted for analysis are not alloted proper processing time to get a good result.

For comparison, I ran your first game through the Critter engine running in Fritz 13. This is not a super deep setting, but it's vastly superior to what chess.com gave you. This run took about 10-12 minutes on my machine. All commentary comes from Fritz.

I just noticed that some of the markings are not present when pasting in from pgn here. The sidelines are sometimes marked with extra commentary, for example, it suggests 17. Bd3!? as a better move, and "interesting alternative." It marks the position as equal with that move, then shows that after 17. b4 Nac4, Black has the advantage.