|
jesus WEP posted:i found you also get one for literally first move out of book in the evans gambit, if your opponent takes a second pawn I don't know if this is actually true but I have heard it claimed that chess.com's "brilliant" scoring is weighted by your rating, so even if you're following a well known opening, it still really likes a good piece sac because I guess it assumed you didn't know that opening and discovered it independently, or at the very least it's just praising you for studying up. A funny one is that a lot of the traps in the Stafford Gambit will get a "brilliant" score from chess.com's analysis engine, but it really hates when you go into it in the first place and calls Nc6 an inaccuracy (to be fair it's pretty popular thanks to Eric Rosen so it's probably fair to call it a bad move when so many people know about it now).
|
#
¿
Nov 3, 2023 19:13
|
|
|
|
| # ¿ May 16, 2024 02:23 |
|
|
Mikojan posted:I know, and I don't, but I always feel kind of guilty when it happens. A way to think of it is that you're helping them learn just as much as you're learning. Learning to convert a winning endgame is an important skill and it's a hard one to practice at lower levels when people keep resigning from being down material.
|
|
#
¿
Nov 7, 2023 10:23
|
|
|
Leperflesh posted:lichess puzzles through the website show the history of the game so you can see if you really want to, but in the app you don't see it and the chess.com puzzles don't seem to show a game history. Really on any website it's not a big deal because if you want to know if you have castling rights you can just touch your king and see if castling is a legal move.
|
|
#
¿
Nov 18, 2023 01:02
|
|
|
The estimated elo thing is a bit weird, I think it has some kind of evaluation of how "hard" it is to see accurate moves, so just having high accuracy alone doesn't necessarily give you a high estimated elo (eg. a game where your opponent blunders a bunch of pieces and then a M1 doesn't require very high level play to be able to find the best moves), but at the same time it's never very clear why it thought a particular game was good or not.jesus WEP posted:the new elo estimate is very funny but not quite as funny as when you both had a queen hanging for 6 consecutive moves and the eval chart looks like a rapid heartbeat Look sometimes you just want to take the eval engine on a journey: 
The Cheshire Cat fucked around with this message at 00:53 on Nov 25, 2023 |
|
#
¿
Nov 25, 2023 00:49
|
|
|
regulargonzalez posted:Played a decent game just now but ultimately I'm frustrated because I just don't know how to develop better vision. I saw Bc4 rather than Bb3 which the engine seems to think is just as good (it likes the latter a little bit better though). I'm only 1021 rated on chess.com and don't play on lichess but I mostly play against bots rather than rated games because I always feel so intimidated against human players even if I actually have a pretty decent record against them. I didn't really have an idea of what I would do with that move but I did kind of have a feeling it might be the best one because of how forcing it is - the issue I see with Ne6 and Bxe8 is that both of them win the exchange but you also lose an attacker which makes it easier for the king to slip away. Nd7 doesn't have that problem and does have the advantage of only having one legal move in response, but it's also a "but then where do I go from here?" problem where it seems like the only thing to really do is make the same Bxe8 trade except now your knight is in a worse position. For any of Bb3, Bc4, or Bd5 on the other hand, you create a situation where there's only one legal move for the king (which leads to a pretty easy to see forced mate), and the only other option for black is to block with their queen (where every available option will end up losing the queen or being a forced mate). Although like I said I didn't exactly see all of that (although I did have a strong feeling that there were a lot of forced mate lines in there) but I had a kind of gut feeling keeping all my attackers on the board was the best move. There are a lot of situations like this where if you put yourself in black's shoes, even if you're down material what you want to do is try to trade off pieces because the current problem is how many pieces are threatening your king and you need more maneuvering room. Conversely, as white, you can take the trade as you'll still be ahead in material but it gives up your very strong positional advantage, so you want to look for moves that allow you to keep all the pressure on the opponent's king. Although all that said, you were already extremely winning so it's not a big deal to not find the most optimal move. It would have been just as viable a strategy to aim to trade off as much as possible to simplify into an easy to win endgame. There's a common problem a lot of players have at all sorts of skill levels which is "trying too hard to be brilliant" and ending up making a mistake because you feel like you should be able to win right now, rather than just sitting comfortably on your advantage and letting it win you the game in the long term. Playing it safer might be boring but it's a lot more reliable.
|
|
#
¿
Dec 1, 2023 04:05
|
|
|
Yeah all the advice I've seen from higher level players is that it's generally a bad idea to concede the bishop pair so early in the game. In your specific example with the French, it might take away your immediate idea in the opening but it comes at the expense of a long-term weakness for white so the reason why the lessons on the opening don't cover it is that at that point white has stopped making the best moves and you're already out of the opening.
|
|
#
¿
Dec 2, 2023 01:05
|
|
|
CubicalSucrose posted:Lichess has a puzzle option where you can filter for specific puzzle types. Mate in 1, forks, pins, etc. I do find one problem puzzles tend to have as practice for real games is that the nature of a puzzle means that you know there's a thing to spot. Being able to look at a position and go "there's something tricky here" is really the hardest part. It does help to get better at tactics because some specific patterns do occur a lot in real games (man I used to get got by knight forks so often until I started doing puzzles), but I think it can be helpful to try uncategorized puzzles, to practice that ability to just find the best move, without the extra hint of what the best move might be. *edit* actually I have a pretty good example from a recent game, I had this position, where the tactic should be pretty easy to spot and I expect most people in here to see it, but I missed it in the game because I was more focused on avoiding getting back-rank mated and played Nd2 instead. It wasn't a super costly mistake since I was already up a minor piece at this point but it's a pretty crushing blow (white to move, although that should be pretty obvious since if it was black's turn it's just a M1):  Incidentally this wasn't even me just having a particularly dumb day; I actually had 95% accuracy in this game and this was my only serious miss. The issue I had was basically that I moved the Queen to g4 to defend the knight on d7, without even really thinking about how it would also pin the g7 pawn to the king. So in my head, f6 was still protected and if I took the bishop it would just be an equal trade and my knight was too well positioned to want to trade it off. Meanwhile I saw the threat of Qe1# and wanted to address that since blundering a M1 when I'm already up material is obviously a pretty bad thing to do. I just completely missed that Nxf6 wins the queen by force and I didn't even really need to worry about the back rank threat. The Cheshire Cat fucked around with this message at 00:40 on Dec 6, 2023 |
|
#
¿
Dec 5, 2023 23:48
|
|
|
Redmark posted:I think "no opening prep" just means that they check their own games or remember what not to do from playing the opening a lot (which not everyone can do depending on how talented they are). I don't think you can make it to 2k playing bad moves in the opening unless you're a sandbagging GM. Yeah while it obviously helps a lot to study known lines, with computer analysis you can learn a lot from your own games because even if you aren't consciously studying opening theory, the computer will just tell you what you should have done anyway.
|
|
#
¿
Dec 6, 2023 23:55
|
|
|
Bilirubin posted:Nobody is ever good at chess. I kind of wonder how the GMs of today feel about the game compared to the past. The age of chess bots has really changed how the game is understood because now we have an objective measurement of how far from "perfect" a game is, and in any tricky position analysts can just bring up Stockfish and find out what the player "should" do (although I know they do usually try to avoid relying too much on computer analysis when live commentating because it's just not very interesting to ask the computer what to do all the time). I mean there is more nuance than that, and often you can get quicker wins by playing suboptimal moves because you're gambling on the fact that your opponent won't always find the best responses and those suboptimal moves give you a stronger position than if you had played the top computer choice, but it is still a thing where if your goal is to maximize your win probability, no matter what your opponent does, the computer always has the right answer.
|
|
#
¿
Dec 17, 2023 00:27
|
|
|
The other thing to look for in scholar's mate is that the c2 pawn is vulnerable without the queen defending it; after chasing the queen away, you can jump your knight to d4 or b4 and probably win a pawn and a rook for a knight. At the very least this usually forces white to pull their queen back to defend it, wasting all the time they spent moving her around in the first place. *edit* they can also defend by putting their knight on a3 but usually the response here is to just take the knight with your bishop and force them to have doubled pawns on the edge of the board. The Cheshire Cat fucked around with this message at 22:29 on Dec 27, 2023 |
|
#
¿
Dec 27, 2023 22:26
|
|
|
GhostofJohnMuir posted:lol, some poor soul gifted me a 99.3% accuracy game. basically forced me to play the best moves by making them so obvious It's always funny getting these. You feel so smart looking at the game review and then you actually go through and look at each move and are just like "oh yeah I guess taking all these free pieces was probably easy to find"
|
|
#
¿
Jan 3, 2024 19:36
|
|
|
Huxley posted:As I've settled into the 2k+ puzzle range, the three "tricks" I've learned are: Yeah the last ones are often the ones I find the trickiest because they typically never look like they do anything on their own - they create a threat that seems trivial to defend so my brain just moves on but it turns out the position created by defending the threat is advantageous for me. Learning to look out for moves that are forcing is a good puzzle tactic since a lot of puzzles are built around "if they do anything other than this one move they instantly lose, so what can you do with that" The Cheshire Cat fucked around with this message at 17:15 on Feb 15, 2024 |
|
#
¿
Feb 15, 2024 17:02
|
|
|
Yeah under that kind of time pressure, if you don't immediately spot a forced mating sequence then it's better to just simplify and promote for an easy win. Trying to get too clever might result in you burning a lot of time to find a mate that isn't actually there, or thinking you've found one and then it turns out it doesn't work and having to push with even less time.
|
|
#
¿
Feb 17, 2024 07:58
|
|
|
Yeah I think that at lower skill levels, if you are going to learn an opening, just memorizing the correct moves to make isn't going to do you much good because if you don't know what the point is you'll likely blow it as soon as you're out of book moves. It's better to learn the general idea behind why those moves are being made - figure out what the objective of the opening is and then you'll have a better sense of what you should be doing when your opponent does something that wasn't part of the original plan. It's also a good idea to stay away from tricky/trappy openings because while they do give you a big advantage if the opponent falls for them, which at lower elo they often will, they tend to put you in a bad position if they know the trick and play the correct response, and you need strong fundamentals to be able to play your way back out of that hole, which you probably won't have if you've been relying on traps to rack up a lot of quick wins.
|
|
#
¿
Feb 23, 2024 21:34
|
|
|
multistability posted:There's a whole controversy in chess right now where the top grandmasters are saying they want to play chess960 or fisher random or whatever the gently caress you call it at classical time control because they're tired of having to memorise opening theory until say move 20 until they get out of preparation and finally start playing real chess and searching for tactics n poo poo. Now, I'm not sure how valid that is, but chess players seem to find more value in actually finding good chess moves in novel positions rather than working out if their memory is better than their opponents'. And those tactics clearly have more value at quicker time controls than classical. Something to think about I guess Yeah I could see GMs being frustrated by this in the age of chess engines. There was a random Hikaru video I saw a while ago where he commented that when you're playing against an opponent's prep, you aren't really playing against them, you're playing against Stockfish. It just kind of comes down to how willing you are to memorize all the best moves the computer tells you to play and trying to anticipate how deeply your opponent will play into your preferred line, which doesn't seem very fun.
|
|
#
¿
Feb 24, 2024 04:55
|
|
|
Bruce Hussein Daddy posted:That's a good example of something I don't understand about the engine. After black's blunder, it says +5. But after playing the next 2 moves, it settles back to what I would expect, +3. It depends on the search depth, sometimes making a couple moves into the best line allows it to find something new that it didn't quite reach from the initial position that changes the evaluation.
|
|
#
¿
Feb 25, 2024 22:15
|
|
|
Yeah that looks pretty winning for white but probably not a trivial one since while you're way up on material, black's rook gives them enough counterplay to be annoying and if you don't deal with those pawns they are going to be a big problem for you. It's the sort of position where my gut would be to try to mop up their pawns (which seems easy to do by saccing the bishop for them) and then look to try to force a rook trade so all that's left is black's king vs. white's king + connected pawns. It's also the sort of thing where I would probably blunder one of those pawns during the "trying to trade rooks" phase of that plan though and that can easily end up turning into a drawn endgame.
|
|
#
¿
Mar 2, 2024 00:28
|
|
|
Salt Fish posted:There's something broken with the puzzle because if you "play against computer" its black to move. However, even in that position the eval is +2. The eval with white to move is like checkmate in 32 after the obvious move, so I think chess.com is broken right now. No this is just how chess.com's puzzles work. They always start with an opponent move, so the position shown isn't actually the puzzle, the puzzle is the position resulting from that first move (which is preset by the puzzle-maker). I'm not sure why it's like this, my assumption is that it's to allow en passant to work correctly if that's part of the puzzle.
|
|
#
¿
Mar 6, 2024 03:49
|
|
|
Stalemating is a great rule because of how many people hate it. Everybody always thinks "I trapped the enemy king, why is that not a win for me", because they always think of it from the side of the winning player rather than the losing player. If stalemate didn't exist, then what's the point of even playing out an endgame where you don't have sufficient material to win? Most people resign there anyway, but the existence of stalemate at least gives you something to play for rather than a resignation being the only thing worth doing. It also means that you can't win a game just by taking all the opponents pieces. The objective of the game is to checkmate the king, not to win material. If you can't checkmate the king, you didn't win.
|
|
#
¿
Mar 13, 2024 21:27
|
|
|
Positions constructed to confuse chess engines are usually very weird like that. Generally the point is less "this is a position that could occur in a real game" and more "there's nothing in the rules that says this position is impossible so the software has to account for it".
|
|
#
¿
Mar 13, 2024 22:19
|
|
|
I don't know that exact position but it might be a 50 move rule thing where Stockfish would call it a draw because of that. There are some funky positions that are very long forced mates that can't actually be done because your opponent is allowed to claim a draw before the line is finished.
|
|
#
¿
Mar 25, 2024 22:15
|
|
|
Huxley posted:I have played like, 2500 games between the two big sites in the past 18 months and have never in my life played a game over the board. When playing over the board as a beginner make sure you're playing against someone better than you, because in my experience I really need someone to tell me when checkmate has happened.
|
|
#
¿
Apr 23, 2024 21:18
|
|
|
Exercu posted:it's true - the main counterplay to wayward queen is just going "ok, good for you" and developing your knight like you probably should anyway, and then he's just got a queen hanging out putting pressure on gently caress all until he eventually has to move it somewhere else. Yeah wayward queen is bad even if they don't blunder their queen because it's very easy to chase it around the board while developing, so you win a lot of free tempii while they're constantly wasting moves just getting the queen out of trouble. You do have to be careful about not blundering a fork since the queen is going to have a lot of freedom to move around the board in the early game and can attack any pieces you've left undefended, but so long as you don't get over-aggressive about trying to chase her down and just develop normally you should come out ahead.
|
|
#
¿
Apr 30, 2024 23:02
|
|
|
|
| # ¿ May 16, 2024 02:23 |
|
|
PerniciousKnid posted:If you keep falling for the same opening trap then you should probably figure out how to avoid it just from reviewing your games. If you're hitting different traps every time then which traps would you even study? A lot of traps tend to have common patterns so you should look for those and see if you can start recognizing them. Learning tactics can help a lot, since usually the idea behind traps is that by falling into them you are exposing yourself to a tactic later down the line, so by being able to spot those tactics in advance you can figure out moves that will shut them down before they get played. There's a lot of opening traps that simply stop working because you pushed an extra pawn and are now defending a crucial square they needed to put a piece on at the end of the sequence.
|
|
#
¿
May 1, 2024 04:32
|
|