I don't understand Chess.com's Computer Analysis

The "Accuracy" Claims that I played at 95%, yet in details, I only found the top move 52% of the time. This seems to be a contradiction to me. 

I don't believe that chess.com ever revealed how it calculates the CAPS rating. Accuracy is just CAPS renamed.

You had 23 "best" moves. I think that is the main factor driving the score.

However, six inaccuracies and one blunder cost you the game.

When CAPS came out, Chess.com posted an article telling some thing about how it is calculated:

https://www.chess.com/article/view/better-than-ratings-chess-com-s-new-caps-system

They haven't revealed more details that I have seen anywhere.   At any rate, you can't just take the number of good moves and divide by the total move count to get the CAPS accuracy score.

Thanks for the response. I am playing the computer at the top level so that I can properly measure my play at slow time controls as a point of comparison to someone that plays near flawlessly. I find the CAPS statistic confusing. Sometimes I will get a figure in the 90's such as the game I posted, and yet the % of finding the top engine move is only around 50% and sometimes even less. Other times, I will get a CAPS rating in the 60's where I think I have played very well and the % of finding the top engine is higher and avg pawn difference is lower per move as compared to high rated CAPS games.  

In other words, the fact that you scored a 95 CAPS rating (on a scale of 1 to 100) doesn't necessarily mean you played a good game!