Lizard 11.2

[Homayoun]

lizard 11.2
https://github.com/liamt19/Lizard/releases/tag/v11.2

[Sedat Canbaz]

lizard 11.2
https://github.com/liamt19/Lizard/releases/tag/v11.2

Thanks Homayoun, plus to Lizard's author as well

Btw, right now I am running a new strength test,
Which are close to 3700 Elo, and I wonder much
Lizard 11.2 will be close or not... On other hand,
Later I will check also not only the strength, but
Overall draw percentage too...you may know, for
Book tours, as far as possible I pick ones, which
Are capable to produce lower draw values...
Because both (fun + strength) means GREAT !)

Greetings

[Homayoun]

lizard 11.2
https://github.com/liamt19/Lizard/releases/tag/v11.2

Thanks Homayoun, plus to Lizard's author as well

Btw, right now I am running a new strength test,
Which are close to 3700 Elo, and I wonder much
Lizard 11.2 will be close or not... On other hand,
Later I will check also not only the strength, but
Overall draw percentage too...you may know, for
Book tours, as far as possible I pick ones, which
Are capable to produce lower draw values...
Because both (fun + strength) means GREAT !)

Greetings

Hi Sedat
Thanks too, for your interest. As you see I have shared Horsie engine too. Horsie is a port of lizard engine from C# to C++. According to author’s personal tests Horsie 1.0 is almost 30 Elo stronger than lizard 11.2 under short time controls. I will be so thankful if both engines would participate in your tests.
Best regards

[ZamChess]

Thank you !

[Homayoun]

Thank you !

Thanks too, ZamChess.

[Sedat Canbaz]

lizard 11.2
https://github.com/liamt19/Lizard/releases/tag/v11.2

Thanks Homayoun, plus to Lizard's author as well

Btw, right now I am running a new strength test,
Which are close to 3700 Elo, and I wonder much
Lizard 11.2 will be close or not... On other hand,
Later I will check also not only the strength, but
Overall draw percentage too...you may know, for
Book tours, as far as possible I pick ones, which
Are capable to produce lower draw values...
Because both (fun + strength) means GREAT !)

Greetings

Hi Sedat
Thanks too, for your interest. As you see I have shared Horsie engine too. Horsie is a port of lizard engine from C# to C++. According to author’s personal tests Horsie 1.0 is almost 30 Elo stronger than lizard 11.2 under short time controls. I will be so thankful if both engines would participate in your tests.
Best regards

Not at all... really? I missed it..ok, I will add Horsie as player too!

Btw, so far Lizard 11.2 is close to 3700, but for better stats:
More games are needed to be played out...

Later I will inform...

Greetings

[Sedat Canbaz]

UPDATE

Here are 1st gauntlet results, played by Horsie 1.0
Sure we need more games, but anyhow it seems
That latest new Horsie 1.0 is close to 3700+ Elo!

And why these kind of strength tests are important?
E.g as without running them...we can not be sure...
Moreover, a lot of draws may appear...e.g please
Check Horsie 1.0 vs Lizard 11.2's draw ratio = 96%

Btw, some may say more games needed..right.. but if
Checking right now my tournament PC, after 93 games:
Both (Horsie 1.0 vs Lizard 11.2) has draw ratio = 94%

What does it mean ? I think that
Both (Horsie 1.0 vs Lizard 11.2) are trained each other a lot...so leading
To more equal (draw) points, actually same issue happens via books too e.g
If they're trained much each other..later their lines become more drawish...
Of course..via updates ) otherwise forget about it )

On other hand,
There are several exciting techniques to reduce the draws..e.g testing in
Handicapped mode too..then no doubt we will see a lot of wins for sure,
At least for nowadays...since almost all are not trained in specific cond.

Code: Select all

Rank Name                          Elo     +/-   Games   Score    Draw 
   0 Horsie 1.0                      5      14     416   50.7%   81.3% 
   1 Berserk 080923                 33      46      53   54.7%   75.5% 
   2 Raubfisch X46b c40             27      32      51   53.9%   88.2% 
   3 Obsidian 11                     0      49      52   50.0%   73.1% 
   4 K.Dragon 2.6                    0      26      52   50.0%   92.3% 
   5 Lizard 11.2                   -14      19      51   48.0%   96.1% 
   6 Cfish 250621 c40              -26      40      54   46.3%   81.5% 
   7 Spider 080621 c40             -27      57      51   46.1%   64.7% 
   8 PlentyChess 1.0               -34      43      52   45.2%   78.8% 

416 of 800 games finished.

Conditions:
2x Epyc 7B12, CuteChess120, 1 Core, Ponder OFF, 30s+0.6s, Balsa/Unique, 64 Hash

[Homayoun]

Many thanks Sedat. I think in near future Horsie with be a replacement for lizard engine.
Best regards

[Sedat Canbaz]

Many thanks Sedat. I think in near future Horsie with be a replacement for lizard engine.
Best regards

You are welcome Homayoun )

Nice..yes it seems also that
The new Horsie is more powerful than Lizard..

But my main goal, to pick the ones which,
Are less drawish each other...the target is
The draw ratio to be minus plus 75% - 80%
And according to latest tests, to be close
To 3700+ Elo plus to be with these low draws,
Not many can manage to do that.. but I will
Still keep testing other engines too...sooner
Or later..the picture will be clear )

Greetings

[Sedat Canbaz]

@ to Mr. Homayoun

Just I'd like to inform you also that
After running / testing many games vs diff. various Top engines
Both (Lizard 11.2 + Horsie 1.0 ) performed close in performance,
And I can clearly confirm that they are over than 3700 + Elo points

In other words, the above notes can be counted as good news.. )

The bad news or sad is that, unfortunately I could not find any
3700+ engine (playing itself) to produce less than 80% draws
K.Dragon 2.5, Obsidian 11 seems to be less drawish close: 85%
But mostly of the rest Top engines are produced over than 90%

After all what does it mean?)
Raubfish X48 SC is still best..e.g close to 3750 Elo and
In same time keeps the draws, usually lower than 85%

Of course here I refer in case of using strong openings,
Such as if playing via Balsa + Unique opening suites etc.
And be aware too, only for the current tests (completed
In 2 days) more than 20.000 + new bullet games played...
I mean the data is serious regarding all over !)

That's all for now ) see you later my friend )

Greetings