It's really hard to program obviousness, or hard to see. There's also a balancing act in the process where if the definition is too strict brilliant moves will be so rare as to be non-existent.
Brilliant moves aren't all that brilliant
Sort:
Martin_Stahl wrote:
It's really hard to program obviousness, or hard to see. There's also a balancing act in the process where if the definition is too strict brilliant moves will be so rare as to be non-existent.
True, I completely understand that it would never be perfect. Only a human (for now) would be able to say if a move is truly brilliant. I just think using the definition of "a good sacrifice" is a little lacking.
Personally, I'd rather have brilliant moves be more rare than not, so that when I do get one I can actually feel good about it because I know there's no way I made this many "brilliant" moves in a week.
I average 0.1 to 0.2 % of moves as brilliants. Pretty rare.
moves that need to be precise, for example, when there is 1 good move/solution, need to be classified as brilliant in my opinion.
StephanVelChess wrote:
moves that need to be precise, for example, when there is 1 good move/solution, need to be classified as brilliant in my opinion.
Only moves are often forced or obvious.
I'm a very low rating player so I was very surprised on getting a brilliant move
After some analysis I think that another condition to get a brilliant could be a move that is not normal for things like the flow of the game or the rating of the players.
Meaning, a lot of people probably would see that and say "that's just an obvious move" but for players on my tier, maybe not...
MancoAjedrez wrote:
I'm a very low rating player so I was very surprised on getting a brilliant move
After some analysis I think that another condition to get a brilliant could be a move that is not normal for things like the flow of the game or the rating of the players.
Meaning, a lot of people probably would see that and say "that's just an obvious move" but for players on my tier, maybe not...
Rating plays into the definition as well. Som moves will be marked as brilliant for some ratings where they won't be for a higher rated player on the same position.
Brilliant moves are (most of the time) easier to get (for me). Like for example, I just got one today!
i disagree w/ him bc i think it's a brilliant move bc it makes me happy and happy is good
when i said i disagreed w/ him, i mean MasterMatthew52
squid wrote:
you’re not that brilliant either
💀
The reason for 6 moves to be brilliant should be if the king takes the rook, it is a draw, so the king basically cannot take the rook or else, stalemate!
I have only seen maybe 1 or 2 brilliant moves out of a few thousand games, but i dont analyze that often. It seems to me they could become a little more common. Id also like it to not be limited to ONLY be for sacrafices.
@8
"Rating plays into the definition as well. Som moves will be marked as brilliant for some ratings where they won't be for a higher rated player on the same position."
++ That makes no sense. Brilliant should be a property of a move, regardless of who played it.
@2
"It's really hard to program obviousness, or hard to see. "
Here are 4 easy to program criteria for a brilliant move:
- It is a move that wins. Losing or drawing moves are not brilliant.
- It is unique. When 2 moves win, then none is brilliant.
- It involves a sacrifice. Sacrifices are aesthetically pleasing.
- It is a quiet move: no check (+) or capture (x), as those are too obvious to be brilliant. That is the same convention as in problem chess checkmate in 2 or 3: the key move must be a quiet move.
An example of a true brilliant move that satisfies all 4 criteria is 30 Ba3
https://www.chessgames.com/perl/chessgame?gid=1031957
THEY ARE
I've heard some chatter as of late on how the new classification for brilliant moves doesn't seem right and I want to jump into the conversation by providing some context and examples from my recent games and add my thoughts on why I think the new classification doesn't live up to the term "brilliant".
Background
It's pretty clear that the term "brilliant" in the context of a chess move 99% of the time means a really good, yet hard-to-see piece sacrifice. This has been the accepted definition in many chess books. One such definition as taken from the web:
In chess, a brilliant move is a move that is unexpected, ingenious, and has a significant impact on the game. It is a move that is beyond the typical or expected moves and requires deep thinking and strategic planning.
And another taken from chessjournal.com:
A brilliant move in chess is refers to a move that is exceedingly rare and difficult to find that gives the player more of an advantage in the position from what it was prior to being made.
A brilliant move in chess can take many forms, but they all share some common characteristics. Firstly, brilliant moves are unexpected. They are moves that your opponent did not see coming and therefore have trouble countering. Secondly, brilliant moves are creative. They are moves that break the usual patterns of play and force your opponent to think in new ways. Finally, brilliant moves are highly effective. They achieve a clear and decisive advantage for the player who makes them.
Chess.com has a much different, and simpler, definition for a brilliant move. To quote the Chess.com support page:
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.
While the Chess.com definition is a fair application of the term "brilliant move", I feel that many moves classified as "brilliant" on Chess.com don't live up to the expectation. I think chess.com is missing the "hard-to-see" and "unexpected" parts of the widely accepted "brilliant" term. Many of the brilliant moves I've seen here would only be considered great move (!) at most in any other context.
The following are some examples of why I think this definition is wrong.
Examples
In this first example, I was awarded a brilliant move for "sacrificing" a bishop. I wouldn't say this is a brilliant move by any means, but rather a simple back-rank mate threat. If white plays Rxb2?? I have a mate in 3. With that in mind, you can't say I'm sacrificing the bishop - I'm merely forcing white to respond to the mate threat before they can take my bishop. I don't think a brilliant move working out should be contingent on the opposing player blundering by taking the piece. Sure, this position is completely winning for black regardless of what white does, but for an ~1800 rated player, this type of move is not hard to find.
In the next example, I "sacrifice" a rook. Again, this is not a sacrifice. If I saved my rook, white wins a free bishop (-3 in material). If I save my bishop, I at least get the knight back in return (-2 in material). This isn't brilliant by any means, it's just basic math. I'd argue this is a "great" move (by chess.com definitions) because it was the only move that keeps an advantage, but even then, there are no tactics involved, no winning material, and no forced mate after this move.
This next move comes out of a well-known trap in the Fried Liver. White blundered by playing 7. Nxf7??. This is still a known opening line. Black takes the bishop, white played Nxh8 and black can take the c2-pawn and be completely winning (Yes, Qxc2 was the right move, followed by Qe1, Qxc1 Qxc1, and Ne2+ which I forgot was possible). Either way, black is clearly winning after taking the bishop. I'm not sacrificing the rook here, because I have Nxc2 (or Qxc2) and win my material right back. This again is a simple tactic and simple math.
If black played Rg8 last move to save the rook, white could respond with Nxe5 and win the rook anyways. So, if anything, I'm accepting that my rook is lost regardless, making this not that brilliant of a move.
In the next example, you see any other fairly basic back-rank tactic. Black just played Qxe1. If white takes the queen I have Rxe1, Qf1, Rxf1 and I win a bishop. If white plays Qf1 now, I win the bishop. I think this could be argued as brilliant in some context, but using the classic definition of "hard-to-see" this doesn't make the cut (that is, for higher-rated players). It was a very underwhelming brilliant move.
Changing it up a little here, this brilliant move wasn't even the best move in the position. I don't think any move that isn't the top move should be considered brilliant. But, going with the chess.com term of brilliant "nearly best move" this makes the cut. This is a fairly hard-to-see move and you can see the continuation below. White easily comes out on top. I will say though that white is completely winning regardless of finding Rxe6, so by definition, I don't think this should be considered brilliant even though it was a wonderful move nonetheless
In this example, I was given 7 brilliant moves in a row. In my opinion, this is just a basic tactic that everyone practices at some point, but rarely get a chance to actually play. While I do think this is one of the better brilliant moves I've shared in this post, this doesn't make the cut for me. What else am I going to do? I can't move any other piece but the rook, so it's pretty obvious that giving up the rook would mean a draw. Again, this is a game played between 2 1800s.
If anything, I can see how the first move here could be considered brilliant. It was after all my only hope at not losing a completely lost position. But after I played the first move, of course, I saw the whole continuation. I don't think I should be awarded 6 move brilliants. What's the limit? White didn't have to take the rook right there. I could have gotten way more brilliant moves.
1/2-1/2
12345678hgfedcba
1.-- g2+
2. f1
f2+
3. e1
e2+
4. d1
d2+
5. c1
c2+
6. b1
b2+
7. a1
a2+
8. xa2
1/2-1/2
So, the above example could be considered brilliant, for at least one move, but this next example is one that I feel is more closely in line with what people generally think a brilliant move is. I just sacrificed my rook and according to the analysis, this was the only move where white wasn't winning (mate in 6). You can argue this is hard to see, a sacrifice, unexpected, and the only winning move. This is what I believe the definition of brilliant is. This move did not underwhelm me when I looked at the game review.
Conclusion
Out of all of the above examples, only the last 2 really made me feel like I found something actually brilliant. I used to strive for brilliant moves and eagerly look for them post-game. They used to mean something special. I remember seeing posts all the time in the forums showing off exciting, brilliant moves. Nowadays, when I see a brilliant move I often find myself underwhelmed because the move in question was just a natural move.
I think Chess.com is missing the hard-to-see, unexpected aspect of such moves. I don't want to be rewarded for playing obvious or natural moves. I want to be rewarded when I find something that's actually brilliant and takes time to find and calculate. I have many more examples in my archive that I didn't bother to pull in, but they involved sacrifices that aren't true sacrifices (I win the piece back the next move and improve my position) or sacrifices that aren't sacrificed at all, but rather simple math adding up the piece values.
I understand that the game review is not broken per se; I just disagree with the new brilliant move classification. The purpose of this post is to show why simply calling a "sacrifice" brilliant doesn't cut it.
Possible Solutions
Classifying moves differently for different rating groups is a great idea. Many of these could be considered brilliant by a newer chess player trying to break 1000 rating. But for more experienced players, these are just "good" or "great" at best. Maybe this is what's making these moves seem underwhelming. Maybe the bar just needs to be raised for all rating ranges when it comes to awarding brilliant moves.
Maybe it's the definition that isn't working. Maybe there needs to be a new category of classification above the current brilliant class. This new category could be used for those truly hard-to-see, unexpected, accurate moves. Maybe players can have something more to shoot for now that brilliant moves are so common.
I believe the best idea is to change the definition of brilliant (again) to better align with the standard meaning of the term. Is another full overhaul of the classification system needed again? Others have said these concerns extend beyond brilliant moves, so maybe the solution is another re-working.
In the end, this is the place to provide feedback on the game review features. So... here's my feedback. I'd love to get other opinions on brilliant moves or the move classification system as a whole.
Martin_Stahl
Apr 12, 2023 0
#2
It's really hard to program obviousness, or hard to see. There's also a balancing act in the process where if the definition is too strict brilliant moves will be so rare as to be non-existent.
MasterMatthew52
Apr 12, 2023 0
#3
Martin_Stahl wrote:It's really hard to program obviousness, or hard to see. There's also a balancing act in the process where if the definition is too strict brilliant moves will be so rare as to be non-existent.
True, I completely understand that it would never be perfect. Only a human (for now) would be able to say if a move is truly brilliant. I just think using the definition of "a good sacrifice" is a little lacking.
Personally, I'd rather have brilliant moves be more rare than not, so that when I do get one I can actually feel good about it because I know there's no way I made this many "brilliant" moves in a week.
Martin_Stahl
Apr 13, 2023 0
#4
I average 0.1 to 0.2 % of moves as brilliants. Pretty rare.
StephanVelChess
Aug 16, 2023 0
#5
moves that need to be precise, for example, when there is 1 good move/solution, need to be classified as brilliant in my opinion.
Martin_Stahl
Aug 16, 2023 0
#6
StephanVelChess wrote:moves that need to be precise, for example, when there is 1 good move/solution, need to be classified as brilliant in my opinion.
Only moves are often forced or obvious.
MancoAjedrez
Oct 23, 2023 0
#7
I'm a very low rating player so I was very surprised on getting a brilliant move
After some analysis I think that another condition to get a brilliant could be a move that is not normal for things like the flow of the game or the rating of the players.
Meaning, a lot of people probably would see that and say "that's just an obvious move" but for players on my tier, maybe not...
Martin_Stahl
Oct 23, 2023 0
#8
MancoAjedrez wrote:I'm a very low rating player so I was very surprised on getting a brilliant move
After some analysis I think that another condition to get a brilliant could be a move that is not normal for things like the flow of the game or the rating of the players.
Meaning, a lot of people probably would see that and say "that's just an obvious move" but for players on my tier, maybe not...
Rating plays into the definition as well. Som moves will be marked as brilliant for some ratings where they won't be for a higher rated player on the same position.
Chessbetterthancheckerz
Dec 16, 2023 0
#9
Brilliant moves are (most of the time) easier to get (for me). Like for example, I just got one today!
Piggypiggypiggypiggy
21 days ago 0
#10
i disagree w/ him bc i think it's a brilliant move bc it makes me happy and happy is good
Piggypiggypiggypiggy
21 days ago 0
#11
when i said i disagreed w/ him, i mean MasterMatthew52
squid
18 days ago 0
#12
you’re not that brilliant either
spacecatchess2007
18 days ago 0
#13
squid wrote:you’re not that brilliant either
💀
AMMAR1ISCANOR99SAMA7
15 days ago 0
#14
1.
e4
e5
2.
Cf3
Cf6
3.
Cc3
Cc6
4.
a3
a6
5.
Fc4
Cxe4
6.
Cxe4
d5
7.
Fxd5
Dxd5
8.
d3
Fc5
9.
O-O
Ff5
10.
Te1
Fxe4
11.
dxe4
Dxd1
12.
Txd1
O-O
13.
h3
Tfd8
14.
Te1
h6
15.
Fd2
Cd4
16.
Cxd4
Txd4
17.
Fc3
Tc4
18.
Fxe5
Txc2
19.
Rh1
Txf2
20.
Fxc7
Txb2
21.
e5
Fd4
22.
Rh2
Tb5
23.
Tac1
Fb2
24.
Tc2
Fxa3
25.
Td1
Te8
26.
Fd6
Fxd6
27.
exd6
Td8
28.
d7
Te5
29.
Tc8
Txc8
30.
dxc8=D+
Frank_Franklin
3 days ago 0
#15
The reason for 6 moves to be brilliant should be if the king takes the rook, it is a draw, so the king basically cannot take the rook or else, stalemate!
DavidMa123
1 day ago 0
#16
SOmetimes brilliatn moves gives you some advantage, but not a totally winning position like the way stockfish crushes humans. Sometimes brilliant moves only win a pawn or so.
Woofsalot2
4 hrs ago 0
#17
I have only seen maybe 1 or 2 brilliant moves out of a few thousand games, but i dont analyze that often. It seems to me they could become a little more common. Id also like it to not be limited to ONLY be for sacrafices.
tygxc
18 min ago 0
#18
@8
"Rating plays into the definition as well. Som moves will be marked as brilliant for some ratings where they won't be for a higher rated player on the same position."
++ That makes no sense. Brilliant should be a property of a move, regardless of who played it.
@2
"It's really hard to program obviousness, or hard to see. "
Here are 4 easy to program criteria for a brilliant move:
It is a move that wins. Losing or drawing moves are not brilliant.
It is unique. When 2 moves win, then none is brilliant.
It involves a sacrifice. Sacrifices are aesthetically pleasing.
It is a quiet move: no check (+) or capture (x), as those are too obvious to be brilliant. That is the same convention as in problem chess checkmate in 2 or 3: the key move must be a quiet move.
An example of a true brilliant move that satisfies all 4 criteria is 30 Ba3
https://www.chessgames.com/perl/chessgame?gid=1031957
DrWolfeKrishnasis060814
0
#19
THEY ARE
Something that surprised me was I played a 1100 game review said played like a 900 and I played like a 1800 in the game and got a brilliant move do you think if is brilliant also I rated 1095 and feel overrated anything on that would help. Here the game
The brilliant move is the rook sack
@diamondsalamander The fact that Elo is even being weighed in the assessment of a solid state board position (or series thereof) completely disqualifies any extrapolated performance figure they attach to it. Precedent performance can be factored only to produce a "chance winning percentage" from X position- not the value of an arrangement of chess pieces. Example: If I play Capablanca, my chance is a billion to one. After he suffers a fingerslip on move 12, I now lead Capablanca +2.50. My chances improve to a million to one LOL. Elo is fine when applied this way. It has no business in assessing his misplay as still superior. His -2.50 is what it is.
I've heard some chatter as of late on how the new classification for brilliant moves doesn't seem right and I want to jump into the conversation by providing some context and examples from my recent games and add my thoughts on why I think the new classification doesn't live up to the term "brilliant".
Background
It's pretty clear that the term "brilliant" in the context of a chess move 99% of the time means a really good, yet hard-to-see piece sacrifice. This has been the accepted definition in many chess books. One such definition as taken from the web:
In chess, a brilliant move is a move that is unexpected, ingenious, and has a significant impact on the game. It is a move that is beyond the typical or expected moves and requires deep thinking and strategic planning.
And another taken from chessjournal.com:
A brilliant move in chess is refers to a move that is exceedingly rare and difficult to find that gives the player more of an advantage in the position from what it was prior to being made.
A brilliant move in chess can take many forms, but they all share some common characteristics. Firstly, brilliant moves are unexpected. They are moves that your opponent did not see coming and therefore have trouble countering. Secondly, brilliant moves are creative. They are moves that break the usual patterns of play and force your opponent to think in new ways. Finally, brilliant moves are highly effective. They achieve a clear and decisive advantage for the player who makes them.
Chess.com has a much different, and simpler, definition for a brilliant move. To quote the Chess.com support page:
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.
While the Chess.com definition is a fair application of the term "brilliant move", I feel that many moves classified as "brilliant" on Chess.com don't live up to the expectation. I think chess.com is missing the "hard-to-see" and "unexpected" parts of the widely accepted "brilliant" term. Many of the brilliant moves I've seen here would only be considered great move (!) at most in any other context.
The following are some examples of why I think this definition is wrong.
Examples
In this first example, I was awarded a brilliant move for "sacrificing" a bishop. I wouldn't say this is a brilliant move by any means, but rather a simple back-rank mate threat. If white plays Rxb2?? I have a mate in 3. With that in mind, you can't say I'm sacrificing the bishop - I'm merely forcing white to respond to the mate threat before they can take my bishop. I don't think a brilliant move working out should be contingent on the opposing player blundering by taking the piece. Sure, this position is completely winning for black regardless of what white does, but for an ~1800 rated player, this type of move is not hard to find.
In the next example, I "sacrifice" a rook. Again, this is not a sacrifice. If I saved my rook, white wins a free bishop (-3 in material). If I save my bishop, I at least get the knight back in return (-2 in material). This isn't brilliant by any means, it's just basic math. I'd argue this is a "great" move (by chess.com definitions) because it was the only move that keeps an advantage, but even then, there are no tactics involved, no winning material, and no forced mate after this move.
This next move comes out of a well-known trap in the Fried Liver. White blundered by playing 7. Nxf7??. This is still a known opening line. Black takes the bishop, white played Nxh8 and black can take the c2-pawn and be completely winning (Yes, Qxc2 was the right move, followed by Qe1, Qxc1 Qxc1, and Ne2+ which I forgot was possible). Either way, black is clearly winning after taking the bishop. I'm not sacrificing the rook here, because I have Nxc2 (or Qxc2) and win my material right back. This again is a simple tactic and simple math.
If black played Rg8 last move to save the rook, white could respond with Nxe5 and win the rook anyways. So, if anything, I'm accepting that my rook is lost regardless, making this not that brilliant of a move.
In the next example, you see any other fairly basic back-rank tactic. Black just played Qxe1. If white takes the queen I have Rxe1, Qf1, Rxf1 and I win a bishop. If white plays Qf1 now, I win the bishop. I think this could be argued as brilliant in some context, but using the classic definition of "hard-to-see" this doesn't make the cut (that is, for higher-rated players). It was a very underwhelming brilliant move.
Changing it up a little here, this brilliant move wasn't even the best move in the position. I don't think any move that isn't the top move should be considered brilliant. But, going with the chess.com term of brilliant "nearly best move" this makes the cut. This is a fairly hard-to-see move and you can see the continuation below. White easily comes out on top. I will say though that white is completely winning regardless of finding Rxe6, so by definition, I don't think this should be considered brilliant even though it was a wonderful move nonetheless
In this example, I was given 7 brilliant moves in a row. In my opinion, this is just a basic tactic that everyone practices at some point, but rarely get a chance to actually play. While I do think this is one of the better brilliant moves I've shared in this post, this doesn't make the cut for me. What else am I going to do? I can't move any other piece but the rook, so it's pretty obvious that giving up the rook would mean a draw. Again, this is a game played between 2 1800s.
If anything, I can see how the first move here could be considered brilliant. It was after all my only hope at not losing a completely lost position. But after I played the first move, of course, I saw the whole continuation. I don't think I should be awarded 6 move brilliants. What's the limit? White didn't have to take the rook right there. I could have gotten way more brilliant moves.