Seeing my games analysis, I presume that is one of the best, but not the best, move according to computer analysis. Like having -0.2 compared to best move. My biggest question is about brilliant move. Isn't the brilliant move not the best?
What's am "excellent move" in analyze feature?
Sort:
Jul 8, 2019
0
#3
Hey guys. Basically an excellent move sits between good and brilliant.
The scale goes (with the highest rated move at the top of the scale):
Brilliant
Best
Excellent
Good
--------
Inaccuracy
Mistake
Blunder
A brilliant move is even better than the "best" move the chess engine would normally propose at that position. It is an extremely incredible unbelievable mind boggling Einstein type move 
. I personally have only seen brilliant moves come up when analysing say games played by famous grand masters.
The moves are classed depending on the computer score given for that move and it's difference from the score for the move made before that (up or down), measured in pawns and 100th of a pawn. So if you make a move and it results in you loosing 2.00 points that will be classed as a blunder. A loss of 0.9 is a mistake and a loss of 0.2 is an inacuracy.
Jul 8, 2019
0
#5
I saw that in my rush to explain I clean went and left the "excellent" move off my scale. Sheesh....not very excellent of me. Sorry for that. @JDCannon there used to be an article in the chess.com help files (I am sure it is still there) showing in centipawns what was an inaccuracy, mistake and blunder. Maybe whoever is responsible for writing the help docs can write a new one explaining the new metrics to include the excellent and brilliant metrics. It could maybe also be part of the article I believe you are putting together on CAPS that gives a better understanding on that as it's still a bit of a "black box" measurement for many.
I can see that there are probably many people who don't understand the reasoning behind the way the engine determines the classes of moves it classifies (brilliant best excellent good, inaccuracy, mistake, blunder) plus probably CAPS as well as best move % etc. A well written article would go a long way to demystifying it for many. While the stats for CAPS is probably heavy, I am sure the basic principles and stats frameworks can be explained in greater depth than it has in the past.
I think this notation is misleading. In my point of view a "brilliant" or "excellent" move would be a movw rhat:
1, It's the only move that easily wins an otherwise unclear position or draws an otherwise lost position.
2. It's a move that is very hard to find. A natural move that is easy to find can never be called excellent.
3. An "excellent" or "brilliant" move has to be also the "best move"
"Brilliant" and "excellent" are not superlative adjectives.
Their definitions leave an open end to how many things in a set of things they can describe. In a list of placement times in a footrace, there's one "best" result (the person who came in first), but there can be multiple or even many "excellent" or even "brilliant" results. I don't believe a change in terminology is necessary.
If you are asking for the analysis board to tell you when you found a move that wins or draws an unclear or lost position, you would be asking it to look at all the moves, determine if all of them except one are strictly bad, and then report back that you chose the only good move, which is time-consuming from a software perspective. The end-user expects an answer as quickly as possible on whether they chose a good or bad move, and would not want to wait minutes or even precious seconds for it to practically judge all the possibilities first. Even the "best" move described here is not actually waiting for the computer to decide whether it's actually the greatest move, which is where "brilliant" comes into play when the computer gets it wrong, I suspect.
Thanks for your detailed reply but you have missed my whole point to be only focusing on details. I respect your way of reasoning but now I will try to shortly state what I mean:
I think the notation "Brillian" and "Excellent" should be abandonned because it's very unclear what they refer to.
Yes, this is what I think too! A move that gives +5 higher evaluation score than other moves doesn't need to be amazing, brilliant, excellent or genial. It might be a natural move that is very easy to find.
What does the analyze computer mean with an "excellent move?" At first, I thought it was a move better than the "best move" but looking at the ciomputer evaluation the "best" move is better than the "excellent" one. Why is then a move called "excellent?"