Romichess - Homer

[Werner]

Hi,
before the match the situation was not clear:

376 Homer 2.01 2413
379 Romichess P3K 2411
387 Homer 2.01pre3 2387 (only 150 games)

Code: Select all

1   Homer 2.01pre3  2200   01½1101111100½1½11000101½½½1100001½0½11½001½11½110  28.5/50
2   Romichess P3K   2200   10½0010000011½0½00111010½½½0011110½1½00½110½00½001   21.5/50

But the result of this match was clear: +48 elos

regards

[Michael Sherwin]

Not so clear.

Homer did not destroy Romi like Tony said that he would. And what is the error margine for a 50 game match? Way more than 48 elo I would guess.

And if the first 11 and last 8 games are loped off, then Romi outperformed Homer for a span of 31 games.

Yes I believe that Homer is stronger than Romi, but that is just simply not clear!

Edit: I bet that there were alot of wild games though!

[Werner]

Hi Mike,
we will see the places in our list on Sunday - I think the same:
Romi will be very close to Homer (as it is already)!
This was only 1 match and before I do not have 1000 games (+-17) with both engines or much more - the error margin will be too high to say which is really stronger.

But we can say: very close in strength

[Michael Sherwin]

Thanks Werner,

This version of RomiChess is being very hard to nail down. Even for Homer!

I wonder what Tony or Wael think of the result.

[Dr.Wael Deeb]

Thanks Werner,

This version of RomiChess is being very hard to nail down. Even for Homer!

I wonder what Tony or Wael think of the result.

I am welling to agree with Werner here that regardless the result,the two engines are pretty much close in strength....
Of course enabling the learning abilities of Romi could write another scenario though

[Daniel Mehrmann]

Hi,
before the match the situation was not clear:

376 Homer 2.01 2413
379 Romichess P3K 2411
387 Homer 2.01pre3 2387 (only 150 games)

Code: Select all

1   Homer 2.01pre3  2200   01½1101111100½1½11000101½½½1100001½0½11½001½11½110  28.5/50
2   Romichess P3K   2200   10½0010000011½0½00111010½½½0011110½1½00½110½00½001   21.5/50

But the result of this match was clear: +48 elos

regards

btw: Homer 2.01pre3 was compiled 21.3.2007 and is no longer developed. Homer 3.0 is a complete new engine. 3.0 lost his style compared to 2.x. It goes more direction Rybka.

Best,
Daniel

[Ovyron]

btw: Homer 2.01pre3 was compiled 21.3.2007 and is no longer developed. Homer 3.0 is a complete new engine. 3.0 lost his style compared to 2.x. It goes more direction Rybka.

Something like this happened to Glaurung (The rewrite, not the style), and I think some persons stopped testing old Glaurung and moved to the new version. I think that when a new engine appears, it should also have a new name.