Are "Brilliant moves" in computer analysis just any decent sacrifice now?

Ex. Here

The computer calls this a brilliant move even though I only have 4 choices and this is the only one that saves the bishop, so this is kind of unreasonable but here

Bd1 is in fact a brilliant move, because Rf1 is difficult to see in that line

It gave me a brilliant move for a bishop sac in a daily chess match. Maybe the engine doesn't focus on sacrifices first so they are more likely to be brilliant?

It gave me a "!!" for a simple move. Perhaps b/c most people would just try and check with the Q? And this was a Daily Chess game! well this one actually makes a lot of sense Xe4 the knight has to move any other square than d2 causing the player to lose the night because Kd2 is a blunder as, after Qf2 it is an instant checkmate.

j

I have only had one (that I've noticed anyway) in all the games I've analysed on here.

Apparently 30 rc8 is brilliant. To me it seemed like a reasonably clear way to try and clamp down on the pawn promotion but I think I missed some of the follow on moves that made it better than I thought it was. I certainly missed a forced mate later in the game although I did manage to win in the end. I've set up a library for any games I get with brilliant moves in future. I don't expect to be adding to it very often !!

As an engine programmer, we use both depth weight and risk to code brilliancy, including especially sacs or risk weights that initially score high as blunders but which are the opposite, which is how we use metrics to code counterintuitive, just like regret, contempt and respect also are calculus in chips and neurons.

A related question might be: why don't engines play more brilliancies? We're working on that at high levels of Turing standard coding right now in avatars that play more like humans, not just in adding obvious tree weighted inaccuracies  and blunders but less obvious brilliacy heuristics.

https://www.chess.com/analysis/game/daily/322955790?tab=report

Would 27. Rxh2 here count as a brilliancy?

It's pretty clear that if white takes there's forced mate, but it took me a while to realize how much it turns the tables of the game.

I got one here!

I have gotten a few. Actually two in one game. I'm about 2000 on Daily, but have gotten a few in blitz, about 10 I say

I think you get them when you make a move that the computer didn’t even think of that’s “brilliant “. Sort of leading to mate in 260 moves lol

Not necessarily. They can be critical moves that save the game, even if the computer sees them.

Never seen one 

Brilliant is when the best move is different at depth 18 to depth 15 and if that is so then it is considered "Brilliant"

I’ve got like 400 games to my name and I’ve gotten one. I spent a lot of time thinking about the move and it turned out to be brilliant, but from my point of view it forced checkmate with different units than the computer found. So the computer thought it was brilliant for another reason.

my next goal is to get a brilliant move and have my moves match what the computer would’ve done

Brilliant moves are ones that are better than what the computer saw so not super common

Hey everyone, I created brilliancy in a 30 second bullet game. Is this stockfish being dumb or am I actually brilliant? I thought this sacrifice was pretty obvious. I calculated pretty much every outcome of the sacrifice in a couple of seconds :l https://www.chess.com/analysis/game/live/22625406025?tab=analysis

Wow

Idk I honestly thought the brilliant I made was actually kind of brilliant. The move was a bishop sacrifice that would have led into a rook sacrifice that would have led to mate unless the opponent sacrificed his knight.