Why I stood up for Allie is why I stand up for FF2

[gonzochess75]

So, if they had just called out clearly that it was a neural network instead of an engine, it would be ok? Too easy. From the ChessBase announcement: “First implemented in chess in Stockfish 12, now it powers the new Fat Fritz 2, a neural network twice its size, which takes it to the next level – vast chess knowledge paired with lightning speed.” So Fat Fritz 2, according to ChessBase, is a neural net.

Do people in this forum actually read or think for themselves? Or if Andy says it’s a clone it’s a clone. If Andy says that Albert claimed it was an engine, then it must be so.

[dkappe]

Gonzo, you are perpetually on tilt. Read the ChessBase release again.

[Guenther]

And dkappe is now defending the Stockfish clone Fire 8, which has already been determined by cucumber, Alayan, Andrew Grant, et cetera to be a Stockfish clone a few months ago, using the source code of Fire 8 beta that its author Norman Schmidt has given them. He claims that making judgments based on Fire 8's source code (and of Fire 4/Seagull's source code) is of shoddy quality.

Yes? So? Are The White Rose, Night Nurse, Dark Horse or Toga net all blatant clones of Stockfish? How come they don’t play like one another or SF? (See the SIMEX of a few months ago.)

White Rose, Night Nurse, Dark Horse, and so on are neural nets, not engines. The nets can inserted into Stockfish and the engine is of course Stockfish. If they are inserted into another NNUE player like Igel or Rubichess, then the engine isn't Stockfish; it is Igel or Rubichess, et cetera. However, the likes of White Rose, Night Nurse, etc cannot play chess on their own independent of an engine. It's like calling Stein or Bad Gyal an engine and a Leela clone; no they aren't; if Stein is inserted into Leela then the engine is Leela, if it is inserted into Allie then the engine is Allie, etc.

You need to be able to distinguish between a neural network and an engine. Night Nurse is a neural net for the NNUE architecture. Stein is a neural net for the Ankan/Leela architecture. Allie is a mostly unique engine that only shares the Ankan backends; it could run Ankan/Leela style nets like Stein and J92. Igel is a mostly unique engine that only shares the NNUE backends, it could run Night Nurse style nets. Fat Fritz 2 claims to be its own engine by Albert Silver but but is really a clone of Stockfish 13 dev with only a different neural network. And Fire 8 of course is a straight up Stockfish 8 clone with minimal changes to Stockfish 8.

None of this controversey over Fat Fritz 2 would have happened imo if Albert Silver stated from the get go that he was selling a different neural network for the Stockfish engine for 100 euros, but because he tried passing Fat Fritz 2 as a unique engine instead of a neural network for Stockfish, the comnunity is thus right in calling the engine Fat Fritz 2 a Stockfish clone.

So, if they had just called out clearly that it was a neural network instead of an engine, it would be ok? Too easy. From the ChessBase announcement: “First implemented in chess in Stockfish 12, now it powers the new Fat Fritz 2, a neural network twice its size, which takes it to the next level – vast chess knowledge paired with lightning speed.” So Fat Fritz 2, according to ChessBase, is a neural net.

Do people in this forum actually read or think for themselves? Or if Andy says it’s a clone it’s a clone. If Andy says that Albert claimed it was an engine, then it must be so.

Bad for you Kappe that the WaybackArchives are not yet following your Orwellean agenda of permanently rewritting history.
(and of course you know that CB meanwhile has rewritten the header text (and all others too) around a dozen of times)

This are the header lines from 2021-02-09
(only in the later additional notes it is said to be based on SF12, which is also misleading as it simply is SF12,
nothing about this in the header lines, except crap - highlighting most funny crap by me)
https://web.archive.org/web/20210210020 ... sbase.com/

Fat Fritz 2.0 - The new number 1
2/9/2021 – Fat Fritz 2.0 is the successor to the revolutionary Fat Fritz, which was based on the famous AlphaZero algorithms. Using a new Japanese AI technology that achieves optimal performance on regular computerprocessors (CPUs – no expensive graphic card required) it combines the best of both worlds: a massive new neural network, trained by Albert Silver with the original Fat Fritz, while learning from the surgical precision of Stockfish’s legendary search. This new version takes chess analysis to the next level and is a must for players of all skill levels, whether beginner or professional, who won’t accept any compromises. Its understanding is unparalleled, as it learned by literally studying billions of chess positions. Fat Fritz 2 comes with the latest version of the most advanced chess playing interface in the world, which provides coach functions, automatic game analysis, a database with 1 million games as well as a six-month subscription to ChessBase Premium Account.

[gonzochess75]

Gonzo, you are perpetually on tilt. Read the ChessBase release again.

The release means nothing. That is what the actual binary that ChessBase released says. Albert Silver modified Stockfish sources to rename the engine itself to FatFritz 2 and add himself as an author. Try looking at this screenshot again.



Why don't you go ahead and open up FF2 on a command line and verify for yourself ^^^^

[Sopel]

Gonzo, you are perpetually on tilt. Read the ChessBase release again.

Are you trying to say that chessbase knows better what a product is than the product itself?

[AndrewGrant]

When you show respect, it never lessens your position.
In fact, it always strengthens it.
On the other hand, when you are disrespectful, it is a demonstration that you cannot imagine a suitable yet civil argument.

Have some respect to this forum and delete this thread.

Agreed. This thread is nothing but dkpappe's passive aggressive attempt to concerned troll Allie. He's no friend of Allie and has been banned from the Allie discord forum.

Yo, any chance you can unban him? Get the Leela folks to do the same? The SF folks too? Thats why hes on talkchess so much now, hes got too much time on his hands. Maybe its time to make a new alias? Alberto Plata is taken, but you could probably come up with something classy.

[dkappe]

Bad for you Kappe that the WaybackArchives are not yet following your Orwellean agenda of permanently rewritting history.
(and of course you know that CB meanwhile has rewritten the header text (and all others too) around a dozen of times)

This are the header lines from 2021-02-09
(only in the later additional notes it is said to be based on SF12, which is also misleading as it simply is SF12,
nothing about this in the header lines, except crap - highlighting most funny crap by me)
https://web.archive.org/web/20210210020 ... sbase.com/

Fat Fritz 2.0 - The new number 1
2/9/2021 – Fat Fritz 2.0 is the successor to the revolutionary Fat Fritz, which was based on the famous AlphaZero algorithms. Using a new Japanese AI technology that achieves optimal performance on regular computerprocessors (CPUs – no expensive graphic card required) it combines the best of both worlds: a massive new neural network, trained by Albert Silver with the original Fat Fritz, while learning from the surgical precision of Stockfish’s legendary search. This new version takes chess analysis to the next level and is a must for players of all skill levels, whether beginner or professional, who won’t accept any compromises. Its understanding is unparalleled, as it learned by literally studying billions of chess positions. Fat Fritz 2 comes with the latest version of the most advanced chess playing interface in the world, which provides coach functions, automatic game analysis, a database with 1 million games as well as a six-month subscription to ChessBase Premium Account.

Ooh, I have an Orwellian agenda? Double plus ungood! I think we can agree that the original announcement is gibberish. Learning from the stockfish search? But it is again identified as a neural network.

Reading comprehension is so fundamental.

[gonzochess75]

Ooh, I have an Orwellian agenda? Double plus ungood! I think we can agree that the original announcement is gibberish. Learning from the stockfish search? But it is again identified as a neural network.

Reading comprehension is so fundamental.

If it is identified as a neural network, then why does the executable call the engine "Fat Fritz 2" <- Are you able to read and comprehend this?

[connor_mcmonigle]

And dkappe is now defending the Stockfish clone Fire 8, which has already been determined by cucumber, Alayan, Andrew Grant, et cetera to be a Stockfish clone a few months ago, using the source code of Fire 8 beta that its author Norman Schmidt has given them. He claims that making judgments based on Fire 8's source code (and of Fire 4/Seagull's source code) is of shoddy quality.

Yes? So? Are The White Rose, Night Nurse, Dark Horse or Toga net all blatant clones of Stockfish? How come they don’t play like one another or SF? (See the SIMEX of a few months ago.)

White Rose, Night Nurse, Dark Horse, and so on are neural nets, not engines. The nets can inserted into Stockfish and the engine is of course Stockfish. If they are inserted into another NNUE player like Igel or Rubichess, then the engine isn't Stockfish; it is Igel or Rubichess, et cetera. However, the likes of White Rose, Night Nurse, etc cannot play chess on their own independent of an engine. It's like calling Stein or Bad Gyal an engine and a Leela clone; no they aren't; if Stein is inserted into Leela then the engine is Leela, if it is inserted into Allie then the engine is Allie, etc.

You need to be able to distinguish between a neural network and an engine. Night Nurse is a neural net for the NNUE architecture. Stein is a neural net for the Ankan/Leela architecture. Allie is a mostly unique engine that only shares the Ankan backends; it could run Ankan/Leela style nets like Stein and J92. Igel is a mostly unique engine that only shares the NNUE backends, it could run Night Nurse style nets. Fat Fritz 2 claims to be its own engine by Albert Silver but but is really a clone of Stockfish 13 dev with only a different neural network. And Fire 8 of course is a straight up Stockfish 8 clone with minimal changes to Stockfish 8.

None of this controversey over Fat Fritz 2 would have happened imo if Albert Silver stated from the get go that he was selling a different neural network for the Stockfish engine for 100 euros, but because he tried passing Fat Fritz 2 as a unique engine instead of a neural network for Stockfish, the comnunity is thus right in calling the engine Fat Fritz 2 a Stockfish clone.

So, if they had just called out clearly that it was a neural network instead of an engine, it would be ok? Too easy. From the ChessBase announcement: “First implemented in chess in Stockfish 12, now it powers the new Fat Fritz 2, a neural network twice its size, which takes it to the next level – vast chess knowledge paired with lightning speed.” So Fat Fritz 2, according to ChessBase, is a neural net.

Do people in this forum actually read or think for themselves? Or if Andy says it’s a clone it’s a clone. If Andy says that Albert claimed it was an engine, then it must be so.

It's bad faith arguments, such as the above, that are angering so many people, Dietrich. As Dann suggests, some individuals may have been disrespectful in their responses within this thread. However, given such posts, it's clear to me that the disrespectful responses are certainly induced, if not deserved, by you. Can we really blame the bull for the charging the matador?

You know, as well as I do, that ChessBase overwhelmingly marketed the FF2 network, bundled with SF, as a new engine - the new number one engine in fact. Whether the marketing materials contain some double speak is irrelevant.

[dkappe]

When you show respect, it never lessens your position.
In fact, it always strengthens it.
On the other hand, when you are disrespectful, it is a demonstration that you cannot imagine a suitable yet civil argument.

Have some respect to this forum and delete this thread.

Agreed. This thread is nothing but dkpappe's passive aggressive attempt to concerned troll Allie. He's no friend of Allie and has been banned from the Allie discord forum.

I will miss the two posts a month, mostly about jjosh’s nets.