What's the point of Barnes' Opening [1. f3], and what makes people think that Grob's Opening is bad?

So, I've been looking at openings, and I can't find the point of playing 1. f3. Every other opening does something, but Barnes' Opening only achieves a teensy attack on the centre, doesn't develop any pieces, blocks the kingside horse, and ruins kingside pawn structure. After Barnes' Opening, you shouldn't play g4 after Black moves their e7 pawn. Has anyone else found any uses? I currently feel like this is the worst opening in chess.

Second question:

What makes people think that Grob's Opening is so bad? I read some free samples of one of Henri Grob's books (it's in German so I needed Google Translate, but some of the translations didn't make sense), and for the most part, it was quite helpful. I don't see anything wrong with Grob's Opening other than that you shouldn't move the pawn on f2 after Black moves the pawn on e7. It fianchettoes the bishop, can lead to Grob's Opening: Fritz Gambit (where if Black accepts or denies by moving the pawn, Black loses a rook through White doing Bxb7), but removes most of the reasonability of kingside castling (which I usually don't do anyways). I actually use Grob's Opening often, so what makes its reputation so low?

I think it's the damned computer engines again is a big part of it. People now think the grob is "officially" the worst first move as if it's some kind of fact due to computers. Forget about personal knowledge, forget about instinctive feeling, forget about the ideas, the computer tells the objective "truth". I swear computer engines have caused as much harm, misunderstandings and wastes of time as much as they've helped. It's no coincidence that the youngest candidates winner ever Gukesh was strictly forbidden from using computer engines for many years.

Having said that it's an exaggeration to say that 1. f3 "ruins kingside pawn structure", many openings feature an early f3 to create a very powerful pawn structure. It's mainly the early queen attack that stops f3/f6 being a good move early on, of course it's far easier for white to ameliorate the situation than black playing 1. ...f6, which hardly ever comes up in black openings as black rarely has time, so should give white massive advantage.

@1

"I can't find the point of playing 1. f3" ++ It is playable: developing with g3, Nh3, Bg2, Nf2 it gets a firm control over central square e4.

"I currently feel like this is the worst opening in chess." ++ No. 1 g4 is worse.

"What makes people think that Grob's Opening is so bad?"
++ It loses by force for white and it is the only move that loses by force for white.

"Henri Grob's books" ++ Also Claude Bloodgood and the late IM Michael Basman.

"I don't see anything wrong with Grob's Opening" ++ It weakens the king.

"I actually use Grob's Opening often, so what makes its reputation so low?"
++ It loses by force with best play by both sides.

To Uhohspagettio1, through "[Barnes' Opening] ruins pawn structure", I mean that the diagonals immediately become weaker: after Black plays e5 or e4, the best move I see for White is g4, to block the immediate queen attack, or e3, e4, d3, and d4, to give an escape route to the king. I think it's a term-typo on my part. I mean that Barnes' Opening creates weaker initial pawn defenses, and can lead to some quick checkmates, not that Barnes' Opening dooms someone to a bad pawn structure. It can be repaired, it's just that it opens a hole in White's defences that needs to be covered quickly, which is possible in one move. Thanks for pointing that part out; it is an exaggeration in the way that I put it. Further readers and responders, please notice that I do understand that "ruins" is an exaggeration and I mean it only for the initial positions.

To tygxc:

1."I can't find the point of playing 1. f3" It is playable: developing with g3, Nh3, Bg2, Nf2 it gets a firm control over central square e4.

White playing Bg2 right after Barnes' Opening is illegal. I think you might be thinking of g3, the Hungarian Opening, instead, for this move. Also, Nf2 is illegal as well. The horse has to go to Nh3 first.

Although, I do agree, I did not say that Barnes' Opening was unplayable. Thomas Wilson Barnes actually had eight wins over Paul Morphy, so f3 is certainly playable.

Also, if you mean these moves as moves played by White, White will be behind in development if Black plays two moves such as e5 and d5, or c6 and d6. While this is being set up, Black could also set up a checkmate cannon with the bishop and the queen: Black can move a bishop behind their queen and shoot the queen down the board if White does not castle. If White does castle, Black can play h5 and then h6, to open up the defence structure enough for checkmate. This tactic I also use against the Barcza System: 1. Nf3 h5 2. g3 h6. While White is setting up the system, I tear it down before it finishes. But, yes, the tactic does include a debatable rook sacrifice. A similar tactic can be used against the moves you mentioned.

Also, is it possible that you could explain why it is necessary to firmly control e4? Wouldn't it be better to control the weakly defended f-file early instead of e4? Nothing is there, and Black will probably avoid it after anyways.

2. "I currently feel like this is the worst opening in chess." ++ No. 1 g4 is worse.

First, why do you believe that 1. g4 is worse? (that's what we're talking about anyways), and second, you can't make a definitive statement over something that hasn't been proven with definitive evidence. I said that I feel that Barnes' Opening is the worst opening in chess, not that Barnes' Opening is the worst opening in chess, just for the record.

3. "What makes people think that Grob's Opening is so bad?" ++ It loses by force for white and it is the only move that loses by force for white.

You did not provide proof that it is the only opening that loses by force (further proof will be accepted). And, I do not expect optimal play by anyone, not even computers. Chess engines are not yet optimised to play the optimal game. So, you did not provide definite evidence that it is the only move that loses by force for White, nor that it even does in the first place. (Further evidence will be accepted.)

4. "Henri Grob's books" ++ Also Claude Goodblood and Michael Basman.

Could you please explain the context? I don't recognise either name. Are these chess authors, and have they written anything reasonable against Grob's Opening? (Further evidence will be accepted.)

5. "I don't see anything wrong with Grob's Opening" ++ It weakens the king.

How? It does not move a pawn neighbouring the king. Sure, it does make moving the pawn on f2 a very bad idea, but after queenside castling, it'll be fine. Meanwhile, Barnes' Opening does weaken the king's area, although, as previously mentioned, it is reparable.

6. "I actually use Grob's Opening often, so what makes its reputation so low?" ++ It loses by force with best play by both sides.

First, are you expecting best play?, and how do you know that your mentioned "best play" is the real best play? Is the Chess AI you're using somehow perfect?

7. (Three attached games of White using Grob's Opening)

Going back to my first point, the chess AI you are using is not perfect, (though I acknowledge that it is a better player than I am), and, second, none of those games actually finish. I hope you noticed that the first game ends at turn 55, the second at turn 63, and the third at turn 77. I've continued each game on a version of Stockfish (in case you're wondering, I use Stockfish 11 Lite); each ends in a draw. Oh, and also, in these examples, White makes really unreasonable moves, such as moving pieces absently (as in absent play (in which pieces are moved but do nothing, almost like skipping a turn) or to random, unbeneficial positions, or trading higher-value pieces for one that's worth a lot less, or just sacrificing a piece for no apparent reason. I mean this as an honest question, and I mean it, honestly, like, not trying to insult you in any way, but Black in these examples is making better moves than White. Are you pitting two equal bots against each other, or are you putting, say, a training bot against the engine itself? Again, no insult, just a question. No offence intended. (Further evidence will be accepted.)

Also, you did not provide any examples of the engine playing Barnes' Opening, in other words, you did not show that Barnes' Opening fared any better. (Further evidence will be accepted.)

So, tygxc, with all the "Further evidence will be accepted"s, I would appreciate some further discussion and argument. Uhohspaghettio1, if you have anything else to put into the conversation, that would also be appreciated.

Barnes' opening is often used as a way for White to express contempt for and/or taunt Black, as it achieves nothing useful, while also blocking the g1 knight. As a result, it is often followed by Kf2 on the next move, violating typical opening principles. Black may be tempted to start a premature attack on White's king. Another factor is the embarrassment of losing a game to someone who moves the king out on the second move.

Grob's opening is evaluated by the most recent version of Stockfish (that is, dev-20240505-070e564c) as -1.31 at depth 30. This actually means that when playing itself from that position, it expects a black win around 85% of the time and a draw about 15% of the time. This is far from losing by force, but it is a significant disadvantage nonetheless.

Whether Grob's opening is a good choice or not is a matter of style at intermediate levels (at advanced levels, it is unsound). I personally don't like it, but that's just me.

To Quazkie, arguing that "none of those games actually finish": they end with 7 or fewer pieces on the board, and any position with 7 or fewer pieces on the board is completely solved. Just open the analysis board and find the "tablebase" tab, which will give you moves that guarantee the result of the game. Tablebases are literal godlike play, where there's no concept of evaluation - just win, draw, and loss, and exactly how to force the best result for you.

By the way, Stockfish 11 is rather old software (released in 2020 January 15). The latest version of Stockfish is significantly superior at playing endgames. I'm going to run 20 Stockfish dev-20240505-070e564c self-play games from Grob's opening and report the results once they're done.

I believe f3 was some ideas such as preparing e4, or even using the Hammerschlag (2. Kf2) and trying to castle by hand. Of course, 1. f3 is objectively bad, but it is long-term problems and so white doesn't typically lose instantly.

As for 1. g4, this has a bad reputation because engines currently evaluate this as objectively the worst of the 20 possible opening moves. The Grob has several tactical motifs and attacking patterns which could easily catch an unprepared opponent. This gives it good value as a surprise weapon.

These two videos of mine might appeal to you:

The results are in!

Score of Stockfish vs Other Stockfish: 9 - 8 - 3 [0.525]

... Stockfish playing White: 0 - 8 - 2 [0.100] 10

... Stockfish playing Black: 9 - 0 - 1 [0.950] 10

... White vs Black: 0 - 17 - 3 [0.075] 20

Elo difference: 17.4 +/- 148.9, LOS: 59.6 %, DrawRatio: 15.0 %

SPRT: llr 0 (0.0%), lbound -inf, ubound inf

20 of 20 games finished.

In other words: 17 black wins and 3 draws, exactly as the evaluation had predicted. PGN below.

@4

"White playing Bg2 right after Barnes' Opening is illegal."
++ 1 f3, 2 g3, 3 Nh3, 4 Nf2, 5 Bg2, 6 O-O, 7 e4

"White will be behind in development if Black plays two moves such as e5 and d5"
++ White can hold the draw.

"Black could also set up a checkmate cannon with the bishop and the queen"
++ White can defend.

"If White does castle, Black can play h5" ++ White can defend.

"a debatable rook sacrifice" ++ There are no debatable sacrifices, only good, and bad ones.

"is it possible that you could explain why it is necessary to firmly control e4?"
++ White has e4 defended 3 times: f3, Nf2, Bg2, so white can push e4 when he feels ready.

"Wouldn't it be better to control the weakly defended f-file early instead of e4?"
++ The f-file is defended after Nf2 and O-O.

"why do you believe that 1. g4 is worse?" ++ 1 g4? loses by force with best play by both sides.
1 f3 white holds a draw with best play by both sides.

"something that hasn't been proven" ++ I supplied 3 sequences as evidence.
Feel free to use any engine and suggest any white improvement to draw.

"I said that I feel" ++ I said that Grob is the worst opening.
See also Figure 31

"You did not provide proof that it is the only opening that loses by force"
++ I failed to find a black win against 1 f3.
Feel free to use any engine and propose a tentative black win.

"I do not expect optimal play by anyone"
++ That is right.
Grob, Bloodgood, and Basman scored well with it at IM level in classical time control games.
https://www.chessgames.com/perl/chessgame?gid=1102936

For fast time controls and below GM level the opening does not matter.
Any advantage gets dwarfed by all the mistakes that follow.
'The winner is the one who makes the next-to-last mistake.' - Tartakower.
It does not really matter if you are the one who makes the first mistake.

"Chess engines are not yet optimised to play the optimal game."
++ Feel free to suggest an improvement for white with 1 g4?, or a black win with 1 f3.

"Could you please explain the context?" ++ Claude Bloodgood and the late IM Michael Basman also published books The Tactical Grob and The Killer Grob on the Grob.

"It does not move a pawn neighbouring the king."
++ The king goes to g1 after O-O. g4 has left holes on f4 and h4, where a knight can settle.

"after queenside castling" ++ That is even worse.

"Barnes' Opening does weaken the king's area"
++ No. 1 f3 2 g3 3 Nh3 4 Nf2 5 Bg2 6 O-O is solid.

"are you expecting best play?"
++ If you ask a theoretical question, e.g. for ICCF correspondence chess, then yes, that is the assumption. If you ask a practical question about a fast time control at low level, then no, the opening does not matter and any advantage gets dwarfed by mistakes that follow.

"Is the Chess AI you're using somehow perfect?" ++ Close to. Anyway, feel free to suggest any improvement for white in the lines I gave for 1 g4? or any line you think wins for black after 1 f3.

"none of those games actually finish" ++ They end in 7-men Endgame tablebase wins.

"White makes really unreasonable moves"
++ Feel free to suggest any white improvement in the sequences I gave.

"Are you pitting two equal bots against each other"
++ Stockfish vs. Stockfish, with enough time / move.

"you did not provide any examples of the engine playing Barnes' Opening"
++ I found draws after 1 f3 and black wins after 1 g4?. A draw is harder to prove than a win.

Its funny that Quazkie is still on step 1 and thats where it ends.
Want to play g4 in a OTB game? Go ahead. Ive never seen it, but maybe quazkie knows something.But dont play f3...good tip...I was thinking of it..really lol.

Lets say he knows secret chess. So g4 without moving the piece by hand. Abra-cadbra. What would you do? Keep arguing that its impossible...but you lose.

1.f3 is a bad move, however the weaknesses it creates aren't of permanent nature.

1.g4 is another beast, as it does create permanent kingside weaknesses. After either 1...d5 or 1...e5 white is borderline losing IF the guy playing Black is skilled.

To Chess16723,

1. "Not only is it refuted by computers,"

Again, computers are not perfect, and neither are we. So just because computers think it's bad doesn't mean we should apply that to our own opinions.

2. "it is also hard to play practically as even the dxc4 exchange sac line is better for Black"

Could you possibly explain why dxc4 is still better for Black? Not saying that you're wrong, just saying that after Black accepts Fritz's Gambit, and White plays Bxb7, Black's rook is doomed (will be captured unless Black forks White and the bishop on b7 somehow on the exact next turn). That puts White ahead by three points, doesn't it?

3. "and as long as you realize you can sac the queenside pawns due to the queen’s weakness Black is easily better"

Three questions: "You" as in who, White or Black or me?, and what's the weakness of the queen in this situation?, and is the queen's weakness in reference for White or for Black?

To xiaolizhi24,

1. "Grob's opening is evaluated by the most recent version of Stockfish (that is, dev-20240505-070e564c) as -1.31 at depth 30. This actually means that when playing itself from that position, it expects a black win around 85% of the time and a draw about 15% of the time. This is far from losing by force, but it is a significant disadvantage nonetheless."

True. But, it goes back to two points I made earlier: First, people will not play perfectly. Secondly, computers are not entirely perfect either.

2. "Whether Grob's opening is a good choice or not is a matter of style at intermediate levels (at advanced levels, it is unsound). I personally don't like it, but that's just me."

True. As I've said, I use Grob's Opening often. But, could you offer an explanation on why Grob's Opening is unsound at higher levels?

3. "To Quazkie, arguing that "none of those games actually finish": they end with 7 or fewer pieces on the board, and any position with 7 or fewer pieces on the board is completely solved. Just open the analysis board and find the "tablebase" tab, which will give you moves that guarantee the result of the game. Tablebases are literal godlike play, where there's no concept of evaluation - just win, draw, and loss, and exactly how to force the best result for you."

Oh. Thanks, I never knew that existed. My bad again, I'm not used to playing online chess at all.

4. "By the way, Stockfish 11 is rather old software (released in 2020 January 15). The latest version of Stockfish is significantly superior at playing endgames. I'm going to run 20 Stockfish dev-20240505-070e564c self-play games from Grob's opening and report the results once they're done."

Alright. I'll try to find some newer version of Stockfish, or switch. Torch has been recommended to me. But, when I went to the "engine" option, as in the engine that Chess.com is using (Komodo, I believe) which is the maximum difficulty, I tried the Stockfish I have (Stockfish 11 Lite) using Grob's Opening against Komodo... and Stockfish won. I'm a bit confused about that part. (If you want evidence, I unfortunately didn't save the game, so I don't have any. I doubt that Komodo hasn't been upgraded, so I'm not going to get the same results ever again.)

To tygxc,

1. "White playing Bg2 right after Barnes' Opening is illegal." ++ 1 f3, 2 g3, 3 Nh3, 4 Nf2, 5 Bg2, 6 O-O, 7 e4

Okay, so it is a move sequence, in which case it is legal.

2. "White will be behind in development if Black plays two moves such as e5 and d5" ++ White can hold the draw.

What do you mean by "hold the draw"? Do you mean "draw" as in "tie"? Or do you mean something else?

3. "Black could also set up a checkmate cannon with the bishop and the queen" ++ White can defend.

Of course White can defend, but how would you suggest White to do it? Black can play h5...

4. "If White does castle, Black can play h5" ++ White can defend.

Again, yes, White can defend, but how would you suggest White to do it? In all the ways that I'm thinking of, White playing h4 leads to Black questionably sacrificing a rook by Rh6 Rg6 Rxg3 and then White taking it with a pawn to open a diagonal through f2 to shoot the checkmate cannon through...

5. "a debatable rook sacrifice" ++ There are no debatable sacrifices, only good, and bad ones.

Let's say that I get time for only one game against someone. And I play Fritz's Gambit after Grob's Opening. I don't know how Black would respond. So, was it a good idea to sacrifice g4, or a bad one? I might get a rook, but I might not. I put it up to chance and hope that nobody has studied Fritz's Gambit yet.

6. "is it possible that you could explain why it is necessary to firmly control e4?" ++ White has e4 defended 3 times: f3, Nf2, Bg2, so white can push e4 when he feels ready.

Again, you gave me the "what": e4 is defended 3 times. But my question is "why": Why do you need to do that instead of attacking f7 with a queen and another piece early, forcing Black into counterplay or losing by checkmate?

7. "Wouldn't it be better to control the weakly defended f-file early instead of e4?" ++ The f-file is defended after Nf2 and O-O.

Clarification: When I say controlling the f-file, I mean attacking Black's f-file for checkmate or at least four points up in a defended queen attack. Not defending. My bad for being unclear with the word "control".

8. "why do you believe that 1. g4 is worse?" ++ 1 g4? loses by force with best play by both sides.
1 f3 white holds a draw with best play by both sides.

Again, are you sure that your best play is the real best play?, and how do you know?, and do you expect for people to play perfectly?

9. "something that hasn't been proven" ++ I supplied 3 sequences as evidence. Feel free to use any engine and suggest any white improvement to draw.

First, we'd've stopped talking if you've really proved the fact that Grob's Opening is worse and Barnes' Opening is better. Second, I know where White might be able to improve, but I'm not sure how. And xiaolizhi24 already pointed out to me that the version of Stockfish that I use, Stockfish 11 Lite, is old software, so I have no idea what I'd do.

10. "I said that I feel" ++ I said that Grob is the worst opening. See also Figure 31

I know that. But you made the statement before you provided definite proof that Grob's Opening is the worst in chess.

11. "You did not provide proof that it is the only opening that loses by force" ++ I failed to find a black win against 1 f3. Feel free to use any engine and propose a tentative black win.

I do not have any other engine than my previously mentioned, outdated Stockfish 11 Lite, so I can't propose a win. And anyways, no engine is perfect, but they can still help us. There is at least one way that Black beats White after 1. f3 when White is a chess engine.

12. 'The winner is the one who makes the next-to-last mistake.' - Tartakower. It does not really matter if you are the one who makes the first mistake.

Savielly Tartakower was right. But doesn't that mean that it doesn't matter what opening you use, but the rest of your play matters? If you are good at chess and you know what you are doing with your opening and your strategy, you can win like Henri Grob did with g4, Thomas Wilson Barnes did with f3, Adolf Anderssen did with a3, and Magnus Carlsen did with the Bongcloud Attack.

13. "It does not move a pawn neighbouring the king." ++ The king goes to g1 after O-O. g4 has left holes on f4 and h4, where a knight can settle.

Why would you kingside castle after Grob's Opening? It's not that Grob's Opening is a bad decision here, it's that castling would be the bad decision.

14. "after queenside castling" ++ That is even worse.

Why is queenside castling worse? Kingside castling will land you in open territory; queenside does not. Doesn't that make it a better decision to queenside castle than kingside castle? Sure, it takes longer to set up, but after sacrificing the d-file pawn, White can move their queen somewhere on the d-file and then castle, connecting the rook and the queen, which can become a helpful endgame tool. Meanwhile, kingside castling requires White to sacrifice the f-file pawn (another advantage of Bird's Opening, but that's another topic).

15. "Barnes' Opening does weaken the king's area" ++ No. 1 f3 2 g3 3 Nh3 4 Nf2 5 Bg2 6 O-O is solid.

I already mentioned making a checkmate cannon with the bishop and the queen, playing h5, and sacrificing the rook. After that, White's king's area is more open to the checkmate cannon. And, the sentence in reference is incomplete. The complete sentence is "Meanwhile, Barnes' Opening does weaken the king's area, although, as previously mentioned, it is reparable.". I don't think you intentionally removed that part.

16 and the rest:

You cannot say that your chess engine is close to perfection if you don't know where perfection is. I now understand 7-men endgame tablebase wins, but it doesn't explain how and why White decides to make certain moves. I can't make any suggestions since my Stockfish is outdated. Why are draws harder to prove than wins?

To nighteyes1234,

1. "Its funny that Quazkie is still on step 1 and thats where it ends."

If "step 1" refers to the opening that White uses, White's decision changes how Black plays. If White uses the QPO, Black is less likely to play e5. So, if I play Grob's Opening, does the usual response give me the advantage or Black the advantage? That's the question. And, anyways, there's been more than one step mentioned.

2. "Want to play g4 in a OTB game? Go ahead. Ive never seen it, but maybe quazkie knows something.But dont play f3...good tip...I was thinking of it..really lol."

I play g4 in over the board games a lot. And I do know quite a lot about Grob's Opening. I actually have won quite a few games with g4.

3. "Lets say he knows secret chess. So g4 without moving the piece by hand. Abra-cadbra. What would you do? Keep arguing that its impossible...but you lose."

First, I don't know "secret chess". Second, it is impossible to move a piece in person without interacting with it in any way, whether directly or not, so of course I wouldn't argue for something impossible, nor do it in the first place. Third, it's not possible to lose a game, or, well, anything, that you've never done in any way. Fourth, what point are you trying to make? Is it about the topic, or me? And are you expecting me to "magically" move g4 not by hand, but by my foot?

To Mazetoskylo:

1. "1.f3 is a bad move, however the weaknesses it creates aren't of permanent nature."

True. If you're ever planning to beat someone who plays like Morphy, and you'd like to use Barnes' Opening against them, tygxc mentioned some pretty good moves that might surprise someone enough into losing positions.

2. "1.g4 is another beast, as it does create permanent kingside weaknesses. After either 1...d5 or 1...e5 white is borderline losing IF the guy playing Black is skilled."

It is true that Grob's Opening has major weaknesses, some of which are more obvious. But every opening has a weakness, every attack has a counterattack. So then every weakness has a fix. And I would argue that the weaknesses created by Grob's Opening are easy to fix by moves such as Na3, Nf3, and h3 (separate move choices), just like how Barnes' Opening's weaker diagonals can easily be fixed by moves like g3, which tygxc mentioned. It is also true that if White plays g4, Black can quickly win with solid play. But if White has some skilled play of their own, they can quickly trick Black into a losing position.

To KeSetoKaiba,

The videos were helpful. I'm not sure if I agreed with everything the first one says, but the second video had some strategies I believe I'll find helpful later on.

Anyways, using Grob's Opening offers good surprise advantages too. It's just like how Henri Grob called 1. g4 the Spike Opening: if you step on a spike barefoot, it hurts. But if you're prepared and you're wearing thick boots, it might not make a difference to you.

@12

"What do you mean by "hold the draw"? ++ There is no way for black to force a win.

"Do you mean "draw" as in "tie"?" ++ Yes.

"how would you suggest White to do it? Black can play h5" ++ Give a move sequence you think wins for black and I will give a defense for white to hold the draw.

"White can defend, but how would you suggest White to do it?"
++ It depends on how black plays.

"White playing h4 leads to Black questionably sacrificing a rook by Rh6 Rg6 Rxg3 and then White taking it with a pawn to open a diagonal through f2 to shoot the checkmate cannon through" ++ It is a naive approach that backfires.

"I play Fritz's Gambit after Grob's Opening." ++ So a bad response to a bad opening.

"I put it up to chance and hope that nobody has studied Fritz's Gambit yet."
++ Chess is not about hope, or study, chess is about logic and thinking.

"Why do you need to do that instead of attacking f7 with a queen and another piece early, forcing Black into counterplay or losing by checkmate?"
++ You cannot attack right away out of the opening, that only backfires. You first have to fight for the center. The center is like high ground: it facilitates attack and defense.

"I mean attacking Black's f-file for checkmate or at least four points up in a defended queen attack." ++ That is a naive approach.
You cannot just storm a defense, you first have to occupy high ground, i.e. the center.

"are you sure that your best play is the real best play?"
++ Yes, but if you contest that, then use any engine for any time and try to find an improvement for white in any of the 3 losing lines I presented.

"do you expect for people to play perfectly?" ++ On lower levels and in fast time controls not. In ICCF correspondence on high level: yes. In classical time control and on high level: almost: average 1.1 error in the Toronto 2024 Candidates' Tournament.

"we'd've stopped talking if you've really proved the fact that Grob's Opening is worse and Barnes' Opening is better." ++ I did.

"I know where White might be able to improve"
++ Then present your improved sequence where white does not lose.

"the version of Stockfish that I use" ++ Does not matter, you can use any engine, any version, any tuning of it, for any long time per move. Just present an improved sequence of moves.

"But you made the statement before you provided definite proof that Grob's Opening is the worst in chess." ++ I made a true statement and I presented evidence.

"I can't propose a win."
++ Because there is no forced black win after 1 f3, like there is after 1 g4?

"There is at least one way that Black beats White after 1. f3 when White is a chess engine."
++ There is none if both play perfectly.

"doesn't that mean that it doesn't matter what opening you use, but the rest of your play matters?" ++ Indeed: at lower levels and in fast time controls the opening does not matter. There will be so many mistakes, that you can just deliberately make a mistake 1 g4? on move 1 without severe consequences.

"Why would you kingside castle after Grob's Opening?" ++ Stay in the center and white gets checkmated. Queen's side castle after you play c4 is bad as well.
The best option is to support g4 with h3 and castle O-O. That is hoiw Basman played it.

"castling would be the bad decision" ++ No 1 g4? is the mistake.

"Kingside castling will land you in open territory; queenside does not."
++ It does too: you need c4.

"after sacrificing the d-file pawn"
++ Pawns are valuable, never sacrifice a central pawn for nothing.

"kingside castling requires White to sacrifice the f-file pawn" ++ No.

"I already mentioned making a checkmate cannon with the bishop and the queen, playing h5, and sacrificing the rook." ++ A naive approach that backfires.

"your chess engine is close to perfection" ++ It is if given enough time per move.

"you don't know where perfection is" ++ I do know. Here for example are 102 perfect games.

"it doesn't explain how and why White decides to make certain moves" ++ Engine.

"Why are draws harder to prove than wins?"
++ A win is straightforward to prove. Generate a sequence. It ends in a win.
Now try to improve for the losing side. If all attempts fail, then the win is proven.
That is why Losing Chess was proven to be a white win:'my whole involvement in the project was a sort of “gamble” that White would win in the end, as else there would be little chance of completion'

maybe just dont play either of them? they are the bottom of the barrel. do you really need a reason? you are just arguing with an engine at this point

With best play, 1. g4 hangs a kingside pawn as WHITE on the first move.

For example (best moves by both sides, you can analyse for yourself):

To tygxc,

1. "how would you suggest White to do it? Black can play h5" ++ Give a move sequence you think wins for black and I will give a defense for white to hold the draw. / "White can defend, but how would you suggest White to do it?" ++ It depends on how black plays.

I asked how White responds to Black playing h5. Right now, your moves versus mine: [1. f3 e6 2. g3 h5 3. Nh3 h4 4. Nf2 hxg3 5. Bg2 gxf2+]. White loses castling rights and three points. Feel free to change an earlier move.

2. "White playing h4 leads to Black questionably sacrificing a rook by Rh6 Rg6 Rxg3 and then White taking it with a pawn to open a diagonal through f2 to shoot the checkmate cannon through" ++ It is a naive approach that backfires.

Nope, it's my bad. After [1. f3 e6 2. g3 h5 3. h4 Rh6 4. Nh3 Rg6 5. Nf2 Rxg3 6. Bg2], the rook isn't even in danger and is therefore not sacrificed.

3. "I play Fritz's Gambit after Grob's Opening." ++ So a bad response to a bad opening.

White offers Fritz's Gambit. Do you even know enough about it to say that it's bad? Because you saying that it's a bad response to Grob's Opening without knowing how to get there tells me that you don't know about it. It's effective at tricking people into losing a rook.

4. "I put it up to chance and hope that nobody has studied Fritz's Gambit yet." ++ Chess is not about hope, or study, chess is about logic and thinking.

My opponent does not know everything. So I am hoping that they don't know how to respond to Fritz's Gambit. My opponent won't always be the same, but I don't know where they are exactly.

Logic:

I play Fritz's Gambit, hoping that Black does not know how to respond.

If Black does not know how to respond and advances the pawn or accepts the gambit, I get a rook.

If Black does know how to respond, then the game goes on.

5. "Why do you need to do that instead of attacking f7 with a queen and another piece early, forcing Black into counterplay or losing by checkmate?" ++ You cannot attack right away out of the opening, that only backfires. You first have to fight for the center. The center is like high ground: it facilitates attack and defense. / "I mean attacking Black's f-file for checkmate or at least four points up in a defended queen attack." ++ That is a naive approach.
You cannot just storm a defense, you first have to occupy high ground, i.e. the center.

Here's an example of a possible game outcome that happens to me a lot: [1. a3 e5 2. c4 Nf6 3. e3 Nc6 4. c5 Bxc5 5. Bc4 Na5 6. Ba2 Nc6 7. Qb3 Na5 8. Qxf7#].

In it, Black maintains at least one layer of control over every spot in the centre after turn 3. I didn't. I waited for Black to kick me into "worse" positions, and then attacked the weakly defended f7. I attacked with my queen and defended my bishop. I don't need the centre to do that. I didn't even occupy the centre once.

Also, I don't need to know that Black could've just played something else. There is always a counter for everything (but that may come at different times). This strategy works most of the time.

6. "are you sure that your best play is the real best play?" ++ Yes, but if you contest that, then use any engine for any time and try to find an improvement for white in any of the 3 losing lines I presented. / "I know where White might be able to improve" ++ Then present your improved sequence where white does not lose. / "the version of Stockfish that I use" ++ Does not matter, you can use any engine, any version, any tuning of it, for any long time per move. Just present an improved sequence of moves.

I said that I can't find a good solution in a practical amount of time, and my Stockfish 11 Lite is outdated. I'd like the rest of your proof that your best play is the real best play, as in the perfect combination of moves against each other for a consistent White loss, every single time, where every move considers every single branch of the move tree. My moves are not perfect, but they're slowly improving. I'm asking you why you think that your moves are perfect. Mine do not matter right now.

7. "do you expect for people to play perfectly?" ++ On lower levels and in fast time controls not. In ICCF correspondence on high level: yes. In classical time control and on high level: almost: average 1.1 error in the Toronto 2024 Candidates' Tournament.

I don't believe that people are going to choose a perfect branch of the move tree, since they can't see the whole thing. None of us has every possible chain of moves memorised. If the people in ICCF correspondence chess in higher levels play perfectly, then they'll have solved chess. Nobody can do that all in their head.

8. "we'd've stopped talking if you've really proved the fact that Grob's Opening is worse and Barnes' Opening is better." ++ I did. / "But you made the statement before you provided definite proof that Grob's Opening is the worst in chess." ++ I made a true statement and I presented evidence.

I'm not convinced. And the truth will always have an infinite root of evidence. So, if it is true that Barnes' Opening is better, you'll need to find that one bit of evidence that I can't respond to. If I run out of evidence, and you don't, you win this argument. But you haven't convinced me yet. If it's true, show me evidence I can't counter.

9. "I can't propose a win." ++ Because there is no forced black win after 1 f3, like there is after 1 g4? / "There is at least one way that Black beats White after 1. f3 when White is a chess engine." ++ There is none if both play perfectly.

I can't propose a win because I can't see every leaf of the move tree. Neither can engines. But there is at least one way. If you're asking me to propose a forced win, you'll just show me a way to counter it, and then I'll have to show you a way to counter that... and if we keep going, we'll solve the whole 1. f3 branch of chess. Not remotely possible within a reasonable amount of time. And I can't show you the counter to Black beating 1. g4 either, or else we'll have to solve that half of the chess move tree too. Also, "perfectly" means down to every single considerable part of the move tree. You say that that means that no perfect game with 1. f3 ends in a win. There is no way for Black to beat 1. g4 either, in that case.

10. "doesn't that mean that it doesn't matter what opening you use, but the rest of your play matters?" ++ Indeed: at lower levels and in fast time controls the opening does not matter. There will be so many mistakes, that you can just deliberately make a mistake 1 g4? on move 1 without severe consequences.

That means that you can also just deliberately make the mistake of 1. f3 without severe consequences. We need to get to a higher level before making more bold statement.

11. "Why would you kingside castle after Grob's Opening?" ++ Stay in the center and white gets checkmated. Queen's side castle after you play c4 is bad as well. The best option is to support g4 with h3 and castle O-O. That is hoiw Basman played it. / "castling would be the bad decision" ++ No 1 g4? is the mistake.

You didn't provide the why of why I would do this. If it's hot inside and snowing cold outside, would I open the window? Yes. Would I go and sit in front of the window and freeze? No. Would I use Grob's Opening? Sure. Would I go ahead and put my king into open territory? No.

12. "Kingside castling will land you in open territory; queenside does not." ++ It does too: you need c4.

That's for Fritz's Gambit. I don't always use Grob's Opening, Black doesn't always respond with d5, I don't always offer Grob's Gambit, Black doesn't always accept Grob's Gambit, and I don't always offer Fritz's Gambit. I don't need to play c4, it's only in the rare chances that Black plays d5 and I choose to play c4.

13. "after sacrificing the d-file pawn" ++ Pawns are valuable, never sacrifice a central pawn for nothing.

This is ignoring the context. The whole sentence is "Sure, it takes longer to set up, but after sacrificing the d-file pawn, White can move their queen somewhere on the d-file and then castle, connecting the rook and the queen, which can become a helpful endgame tool." So, it's not for nothing. It's to open up a path for the connected pieces to attack.

14. "kingside castling requires White to sacrifice the f-file pawn" ++ No.

This also is ignoring the context. With the context, this would say, "Kingside castling requires White to sacrifice the f-file pawn [to connect the rook and the queen with an open attack route]".

15. "I already mentioned making a checkmate cannon with the bishop and the queen, playing h5, and sacrificing the rook." ++ A naive approach that backfires.

It does not always backfire. It depends on my and my opponent's mistakes. It's an effective strategy that I use to force responsive play. If I make a mistake, yes, it backfires, but I just have to not blunder.

"The blunders are all there on the board, waiting to be made." -Savielly Tartakower

"The winner of the game is the player who makes the next-to-last mistake." -Savielly Tartakower

So, what I've got to do is stick to what I know how to do and do it better. There are even some quotes that point toward Tartakower using or approving of this strategy himself:

"The great master places a knight on e5; checkmate follows by itself." -Savielly Tartakower

Tartakower might've been referring to playing [1. e4 e5 2. Nf3 Nf6 3. Nxe5 Nxe4 4. Qf3 d5 5. Qxf7#], or something along similar lines. A knight on e5 attacks the weak f7 position, so lining up a queen with another piece and that position might fool some players into defence or moving away from the kick, both ending in White winning. I do this with a bishop instead, which usually takes six turns but can shift to castled kings a lot easier.

16. "your chess engine is close to perfection" ++ It is if given enough time per move.

Alright, I gave my Stockfish 11 Lite depth 256 and all the memory on my computer to analyse Grob's Opening versus Barnes' Opening on two different windows. And my laptop is starting to become a fryer. I think I might cook some bacon on the back of my laptop while I wait for this to finish.

17. "you don't know where perfection is" ++ I do know. Here for example are 102 perfect games.

So, have you found someone who's solved chess and is willing to teach you everything? Great! Now that you know where perfection is, just do the math, reach it, and consistently beat everyone in chess. Or do you not have the full tree of moves? If you know where the line of perfection is in chess, you'll know how to reach it. If you know the full equation, you'll get the answer.

18. "it doesn't explain how and why White decides to make certain moves" ++ Engine.

No offence intended, but can you explain the engine's reasoning for some of the stranger moves?

19. "Why are draws harder to prove than wins?" ++ A win is straightforward to prove. Generate a sequence. It ends in a win. Now try to improve for the losing side. If all attempts fail, then the win is proven. That is why Losing Chess was proven to be a white win: 'my whole involvement in the project was a sort of “gamble” that White would win in the end, as else there would be little chance of completion'

Alright. I'm trying to beat black using my Stockfish 11 Lite on my fryer ⸺ I mean laptop. It takes, like, thirty minutes per move. Sadly, it's not hot enough to cook bacon just yet.

To wert,

I will argue with an engine if I believe that it's wrong. Engines are here to help us think, not to think for us.

To Ethan_Brollier,

*To Ethan_Brollier,

When I say "best" play, I'm not referring to when you put the game into Chess.com's analysis, it gives everything a "best" rating or more.

Anyways, on turn 32, White for some reason decides to not check Black, allowing Black to move a pawn farther, and closer to becoming a queen, and also leaves White vulnerable to more checks.

On turn 33, White decides to then check, just to get blocked. White used a turn to defend a pawn that could've been sacrificed to give time for White to move a pawn towards promotion.

On turn 34, I can't find the reasoning behind threatening that back pawn. White used a turn to threaten a pawn instead of moving a pawn towards promotion.

Multiple times during the endgame, White makes moves that could've been used to move that pawn on the a-file further towards promotion.

Also, White plays Grob's Gambit and Fritz's Gambit. If these are Chess.com's 3500 ELO bots, then those are probably good moves.

To Quazkie (post #12):

To xiaolizhi24,

1. True. But, it goes back to two points I made earlier: First, people will not play perfectly. Secondly, computers are not entirely perfect either.

This is exactly why I say it is a matter of style at intermediate and lower levels.

2. True. As I've said, I use Grob's Opening often. But, could you offer an explanation on why Grob's Opening is unsound at higher levels?

More experienced players know better what positional weaknesses are created by playing Grob's opening and how to precisely take advantage of said weaknesses. Less experienced players don't see it often and don't prepare for it.

4. Alright. I'll try to find some newer version of Stockfish, or switch. Torch has been recommended to me. But, when I went to the "engine" option, as in the engine that Chess.com is using (Komodo, I believe) which is the maximum difficulty, I tried the Stockfish I have (Stockfish 11 Lite) using Grob's Opening against Komodo... and Stockfish won. I'm a bit confused about that part. (If you want evidence, I unfortunately didn't save the game, so I don't have any. I doubt that Komodo hasn't been upgraded, so I'm not going to get the same results ever again.)

You can change your analysis engine by entering this menu from the analysis page:

Stockfish 16 and Torch (both with the large download and not the "Lite" version) are the strongest. Komodo is fairly old, similar in age to Stockfish 11, so I'm not surprised that Stockfish 11 wins. (Komodo hasn't been updated since 2020. It seems that Chess.com's engine development seems to be focused on Dragon, and more recently, Torch.)

To Ethan_Brollier (post #15):

It's completely possible to hold onto the pawn (with 2. e3). Developing the bishop instead of defending the pawn can be thought of as a sort of gambit, where Black's bishop is misplaced and can be kicked back later with pawn moves.

I'm seeing a lot of posts saying that Barnes' Opening is a forced draw. In actuality, we don't know. Stockfish dev-20240505-070e564c at depth 30 considers the position to be such that in self-play, Black will win 30% of the time and a draw will occur 70% of the time. This doesn't indicate anything about a forced draw; future engines may be more effective at converting advantages, for instance, and we might get more Black wins from the position.

Up to move 10 is my own analysis. It's backed by engines, but I've had all that prepped for a while now as it's pretty forcing and intuitive for White to follow along. Up to move 20 is a mixture of my own analysis and engine-checking for some of the trickier sidelines (as with 17. exf3, which instantly appears better than 17. Bxf3 because of Nd2 Nxf3 where White trades bishop for knight without reason and allows Black's queen entry to the position, but what if 17. Bxf3 Nd2 18. Re1!, after which White appears to hold after 18... Qg5+ 19. Kh1 Nf6 20. Re3!, to which the computer offers 18... Nf6 19. Bg2 Nfe4! and suddenly White's position is very shaky with all the threats which will accompany 20... Qh4), and I'm reasonably sure I could find continuations like the ones I have in a real classical game.

Move 32 countercounterpoint: This logic is principled, if incredibly simple. One thing you learn quickly in chess is to check for blunders. You've probably heard the "checks captures threats" mantra a few times now, but all that's teaching is calculation to beginners, allowing them to visualize the board as it will be after their opponents move, rather than just after theirs.

Notice how you describe this, using language like "for some reason" "allowing Black to move a pawn farther and closer to becoming a queen" "leaving White vulnerable to more checks", which is all vague and very abstract. None of this is concrete, you saw that White could check and likely thought something along the lines of "White can give a check here, that's probably the best move".

Take a peek into my mentality here: "Here White shouldn't give a check because of g6 Kg7 and Black is in time to push the pawn and deliver mate if White tries to match. Instead, Black should either play Kc7 threatening f7 or Kb2 defending d4, both with ideas of Kc2+ because Black's dark squares are really weak around the king and White may have perpetual (or even winning!) chances if Black doesn't respond correctly. However, Black is still far better, as at the end of the day Black is up a pair of connected passed pawns in an otherwise equal endgame"

Your move 33 point is the same as 32 essentially, so I won't bore you with the same thing twice.

Move 34 point: Are you aware of what perpetual is? If so, I apologize in advance. If both sides' best or only moves are to check and block/evade check in inescapable sequence, then the game is declared a draw. White is losing, and so best move is to offer perpetual, forcing Black to avoid it, thereby winning tempo to delay the loss further.

Consider this: bots at 3500 ELO can't find a way for White to not lose in the Grob.