Will chess.com ever fix brilliant moves?

I love the concept but most of the times the moves just don't make any sense as to why they are brilliancies and sometimes when you think you found a hard move it turns out it was just a best move 

would it really be that hard to fix the brilliant moves or is it more about chess.com's willingness for it 

There's probably nothing to fix, it's just the algorithm will occasionally assess randomly made moves as brilliant due to the outcome.

As a computer science professor with a fair amount of experience in chess programming, I can state pretty confidently that coming up with an algorithm for identifying moves that humans would consider "brilliant", is a VERY hard problem.

Brilliant moves have a pretty simple definition in regards to the review product. They are sacrifices that are good or best and can be influenced by rating.

The computer currently calls queen trades brilliant if you are the one recapturing. This is because any other move is 9 points worse. That is a bad definition.

To me, a brilliant move is a sound sacrifice that does not win back material or reach mate for a long time to come. It shows we saw far ahead even considering moves that seem counterintuitive in the short run. There are quiet brilliant moves, but changing the algorithm to that definition would be a good start.

You have to be able to program it. Now its like ”ohh I left a piece hanging but if my opponent takes it I win”. But how would you program a real brilliant move that people think is brilliant and hard to spot, that is kinda vague to make a code of.

Do you have an example game showing that? Because a brilliant has to be a sacrifice of material.

It was a while ago. I don't feel like digging it up.

In response to the other poster, what can make it more brilliant is how many moves transpire from the sacrifice to the mate or regain of material.

I can't make heads nor tails of how it classifies moves.  I played a daily game recently where an opponent made what game-review called a "Mistake".  This move dropped a knight with check, and on the next move dropped a rook, for no compensation whatsoever, and no exchange of pieces.  Now, if 8 points of material for free with a king out in the weather and my queen / knight all over it and no defenders nearby, doesn't qualify as a blunder, what does?  

PS. I've never had a move labeled brilliant that I can recall.  But something tells me I vaguely recall one where it was an only move, but I would have no idea how to find a game with one if I did have.    

The initial implementation of brilliant used to work differently but it was decided it didn't really work very well, so the sacrifice definition was created instead.

well clearly the current algorithm isn't working and needs to be improved even if it isn't a clear "fix"

As the local uneducated village idiot, I fully agree with you.

At some point, sacrificing the whole brilliant move classification altogether may make more sense.  Of course the community will not be able to gloat anymore.

I think 18. Nc6... was a genuinely brilliant move, forking four pieces whilst threatening checkmate.

@3
"identifying moves that humans would consider "brilliant", is a VERY hard problem."
++ A true brilliant move satisfies 4 criteria:

1. It is winning. Losing or drawing moves are not brilliant.
2. It is unique. When 2 moves win then none is brilliant.
3. It involves a sacrifice of material. That is aesthetically pleasing.
4. It is a quiet move: neither check (+), nor capture (x). Those are too obvious.

It should be easy to program that.

Here is an example of a true brilliant move
https://www.chessgames.com/perl/chessgame?gid=1031957 

How do you programatically determine non-obviousness? Engines don't think like humans and what's obvious for one person may not be for another.

In regards to the overall topic, people have differences in opinion on what brilliant should mean, and there will never be 100% consensus. 

If the site wants to categorize a move as brilliant, that have to make a choice and implement it, even if not everyone agrees.

It isn't. important because it ndoesn't mean anything in any case.

I wouldn't put too much stock in "great" and "brilliant" moves in the game analysis.  Those labels are based on you current rating level and often are just odd.  In a recent game, I got 3 consecutive "great" moves in a forcing sequence (basically, you only play the first move when you see the next 2 because the idea was all connected).  I have had other games when I got a "brilliant" or "great" move for literally the only possible move.

That would be easy to program.  However, it isn't a very good definition of "brilliant".  Consider the following position.... by your definition, 1.f7 is brilliant.