Why does the engine only look for mistakes and not for good moves?

@GW-ISR said in #20:
> according to your previous post it will always be marked as an excellent move, beacuse even if it just gives away a rook for free instead of recapturing, it is better to be down a rook then down a queen.
IDK just check out chess.com & how it works. I'm not a robot, nor a programmer or coder, so I can't be perfect related to these things. Was just trying to give some input. If it hurt you (although that's nothing to be hurt of) then I'm sorry, alright?

@AyaanshGaur12 said in #21:
> IDK just check out chess.com & how it works.
Well, unlike on lichess, we can't actually check the chess-com algorithm. But the output it produces seems to support what was already said: the purpose is to pat users on their backs and make them feel well rather than provide meaningful information.

actually we do know and thing sais (in not a understandable way perhaps) but is really simple - and mean pretty much nothing:

"Brilliant (!!) moves and Great Moves are always the best or nearly best move in the position, but are also special in some way. We replaced the old Brilliant algorithm with a simpler definition: a Brilliant move is when you find a good piece sacrifice. There are some other conditions, like you should not be in a bad position after a Brilliant move and you should not be completely winning even if you had not found the move. Also, we are more generous in defining a piece sacrifice for newer players, compared with those who are higher rated. "

From : support.chess.com/en/articles/8572705-how-are-moves-classified-what-is-a-blunder-or-brilliant-and-etc

They did try something smart but automated something smart is very hard and not worth pursuing. I would not find any value on classification like that.

@petri999 said in #23:
> actually we do know [...]

That's interesting. They've acknowledged, then, that their earlier efforts were nonsense and laughable. (When I was a member there, "brilliant moves" were brought in for the first time - this was roughly 5 years ago - and defined as moves which the analysis had initially rejected as errors but were found to be correct on looking deeper. The results were that perfectly obvious moves to human eyes, the first we might consider, were being given "!!" annotations.)

So now it has to be a piece sacrifice. Well, what a miserable definition, ruling out so many genuinely brilliant moves from all the rest of chess. As if to play brilliantly you have to sacrifice a piece! And as if we don't already know that a piece is being sacrificed so we have to see the engine mark it "!!"

One could argue that patting users on their back and making them happy has some virtue.

In real life, if someone is repeatedly told he is a good guy, he is more likely to be a good guy. Or perform well. Or whatever.

If this works for chess as well, this is up for debate. But if you think you are a brilliant player by those .com standards, you might start looking for sacrifices that you missed before. A larger study about this would be nice.

Still, I don't think that's exactly their motivation, and serious players would probably prefer more meaningful annotations.

I'd just say we should be happy that Lichess looks for our mistakes to ACTUALLY make us improve at chess unlike chess.com which just praises their users that omg you played a brilliant move and just make them feel happy so they can keep paying for their diamond subscriptionship and keep game reviewing. I think this topic should probably go further than a forum thread and maybe on a blog (not just lichess blog). Again, some ppl might want to get praised and not improve at chess, nd they're just a special breed of chessplayers.

@mkubecek Damn, Kubecek - you make some extremely smart and informative comments. Bravo !

@Brian-E said in #24:
> So now it has to be a piece sacrifice. Well, what a miserable definition, ruling out so many genuinely brilliant moves from all the rest of chess.

Not only brilliant moves are sometimes just a pawn pushing, relocating a piece or bringing the King to a better place. These deserve to be rated as the most brilliant according to the difficulty to find them.

@Brian-E said in #24:
> As if to play brilliantly you have to sacrifice a piece! And as if we don't already know that a piece is being sacrificed so we have to see the engine mark it "!!"

To whoever/whatever is marking "!!"s: trust me bro, if I'm sacrificing a piece and I finally win, I don't need your approval.

The noodle is stuck on the ceiling. Did I already mention that? We want to reach the ceiling, if I gathered well over the past 5 years.

Afterthought, following, so that is the dominant theory of learning or teaching: "go fetch".

I use the "interesting move" myself all the time. Safe word.

more edit: scanning in mind for the world of Lichess internet audience in minds' ear and eyes and whatever lies beyond, the sensory experience and its recruitment thingies (where is that? @NDpatzer? If you have an idea, sorry to ping you, but I think the subjectivity of some annotations and imposed engine point of view might be a nice interlude from human psychology).

I meant same from engine automatic graffiti. The "interesting move". Also, I use that as visual bookmark of my own theories or stories I evolve (as I learn long term) over game fragments I encounter (either my own, or, sigh, others). It helps make breadcrumb in the move list forest of moves disguised as SAN syllables.

I suggest to the op, to grow 2 personalities of game analyses. The conformist one, engine ceiling noodle abiding one (which stangely is also perpetuating another ceiling based, older theory of learning), and then the pragmatic one, only human (maybe editorializing on the engine musings, in its unspecified cloud of engine tribe above the clouds looking down upon us).

Best get another human being to stimulate your analysis by listening to your evolving patzer verbalization attempts, even if a trickle of your board thoughts; that slowing down of your thoughts will help both your own mind, cricital constructive and hopefully still having time to use imagination mind and the others to weigh in, them too with their own pragmatic every changing (hopefully improving on average, yep, there too, averaging.. but not same foggy context as Elo pools). Ok, the likely acceptable, someone at least your "level" (for example, opponent that gave you some struggling in few games, even if you level, it is likely their sub-optimal experience set related budding expertise, would have complementary skill gaps with respect to yours. ( I would agree, if you can find one for free, on account, you may not view chess as financial life line, or do not have the means or time commitment choice).

That is, if one could pull their heads from the Elo narrow obscure 1 dimensional cavity (so not really a cavity, but style has its reasons, that the reason ignore, writing fun).