Elo difference for a 5 or 6-men syzygy

[brianr]

Tinker is another data point that the weaker the engine eval, the more the benefit from TBs.
I seem to recall Tinker gained something around 75 Elo, which means its eval is quite poor.
That was back long before I had an SSD.

[mwyoung]

I was told earlier that 5-men yield about +5 Elo while 6-men add about 10. 7-men is estimated to be around 15-20 elo. If i recall correctly.

And that is what I achieved also with 7 man TB. I found this to be correct.

In most games it is true the tables are not much help. But when it does help, it either finds a drawing line, or finds a win. And then it works great. So no reason not to use them. If you have a fast NVMe drive. And can probe starting at 1 ply with no slow down.

If your dive is slow. You may lose as much as you gain. Because you are bottlenecking the engine.

So you will need a storage drive fast enough to not bottleneck the CPU.

[Dann Corbit]

I don't play against chess engines nearly as often as I use then to analyze chess positions.
I could care less about 4 or 5 Elo either way. When I analyze I want to get the right answer

Don't get me wrong, I like strong engines.
But lots of shortcuts (e,g, tossing out some under promotions for 3 Elo or tossing out Tablebase access for 1 Elo) is really horrifying.
I would just as soon wait a little bit linger and get the right answer

I actually think this Elo at all costs "chess improvement" model is rather perverse.
I mean, someone sets up a robot and it beats some other engines and has a high Elo.
Where is the pleasure in that? Most of the time they don't even look at the games, for crying out loud.

I want an engine that can unravel the most twisted chess puzzle of all time,
Even more, I want and engine that can teach me things i don't know.

Tablebase files, when used properly, can make an engine play elegant chess.
And especially so with distance to mate, they can make an engine play beautiful chess.
In some ways life is a search for truth and beauty.
Or at least, it ought to be

[mwyoung]

I don't play against chess engines nearly as often as I use then to analyze chess positions.
I could care less about 4 or 5 Elo either way. When I analyze I want to get the right answer

Don't get me wrong, I like strong engines.
But lots of shortcuts (e,g, tossing out some under promotions for 3 Elo or tossing out Tablebase access for 1 Elo) is really horrifying.
I would just as soon wait a little bit linger and get the right answer

I actually think this Elo at all costs "chess improvement" model is rather perverse.
I mean, someone sets up a robot and it beats some other engines and has a high Elo.
Where is the pleasure in that? Most of the time they don't even look at the games, for crying out loud.

I want an engine that can unravel the most twisted chess puzzle of all time,
Even more, I want and engine that can teach me things i don't know.

Tablebase files, when used properly, can make an engine play elegant chess.
And especially so with distance to mate, they can make an engine play beautiful chess.
In some ways life is a search for truth and beauty.
Or at least, it ought to be

I agree in the most part. But people demand maximum performance with engine testing. And I do not test just for myself. And 20 Elo is not meaningless in the age of NNUE and NN engines. 20 Elo is 20 Elo. No matter how you get the 20 Elo.