Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
jp
Posts: 1345
Joined: Mon Apr 23, 2018 5:54 am

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by jp » Mon Apr 29, 2019 1:30 pm

Jesse Gersenson wrote:
Mon Apr 29, 2019 11:46 am
M ANSARI wrote:
Mon Apr 29, 2019 4:36 am
... really annoying to see an alien like strong engine play these silly moves!
Silly moves?! Does Lc0 get more points if it wins in less moves?
Asserting Lc0 should play according to your preferred aesthetic is off-topic.
Lc0 turns some endgame wins into draws and draws into losses, so you can't call its endgame problems just aesthetics.

jp
Posts: 1345
Joined: Mon Apr 23, 2018 5:54 am

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by jp » Mon Apr 29, 2019 1:35 pm

MikeGL wrote:
Mon Apr 29, 2019 11:19 am
jp wrote:
Mon Apr 29, 2019 9:24 am
M ANSARI wrote:
Mon Apr 29, 2019 4:36 am
What on earth is going on in the second game that Lc0 won ????
There must have been at least a 50 move shorter win!
After 100... Kxc7, it's mate in 12.
After 129... Kc4, it's mate in 8.
After 154. Qxe1, it's mate in 6.
After 186... Kxf4, it's mate in 7.

So at least a 100-move shorter win.
Maybe 6-men TB positions once reached, should be stopped and auto adjudicated.

I mean no point continuing if its already in tb.

CCC's description sounds like they give 6-man TBs to all engines, but some Lc0 play there looked like it didn't have TBs.

It's better if some tournaments don't give TBs because then it encourages developers to develop engine endgame play rather than covering up weaknesses with TBs.

User avatar
M ANSARI
Posts: 3435
Joined: Thu Mar 16, 2006 6:10 pm

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by M ANSARI » Mon Apr 29, 2019 3:20 pm

Jesse Gersenson wrote:
Mon Apr 29, 2019 11:46 am
M ANSARI wrote:
Mon Apr 29, 2019 4:36 am
What on earth is going on in the second game that Lc0 won ???? I thought trolling play was fixed with the latest net???!!!! There must have been at least a 50 move shorter win! Also this under promoting thing has to get fixed ... really annoying to see an alien like strong engine play these silly moves!
Silly moves?! Does Lc0 get more points if it wins in less moves?

Asserting Lc0 should play according to your preferred aesthetic is off-topic.
Are you serious ???? Aesthetic ???? Playing on for a 100 moves when a mate in 12 is there is not aesthetics! An engine is an analysis tool that is supposed to help us analyze a chess position so that we can train and learn from it. This really makes Lc0 useless to analyze as a stand alone engine and you will always need another good engine for a sanity check. Time to figure out a way to fix this with Lc0 and if takes removing the "zero" then be it. I just can't believe that we are still at this stage.

MikeGL
Posts: 931
Joined: Thu Sep 01, 2011 12:49 pm

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by MikeGL » Mon Apr 29, 2019 4:30 pm

jp wrote:
Mon Apr 29, 2019 1:30 pm
Jesse Gersenson wrote:
Mon Apr 29, 2019 11:46 am
M ANSARI wrote:
Mon Apr 29, 2019 4:36 am
... really annoying to see an alien like strong engine play these silly moves!
Silly moves?! Does Lc0 get more points if it wins in less moves?
Asserting Lc0 should play according to your preferred aesthetic is off-topic.
Lc0 turns some endgame wins into draws and draws into losses, so you can't call its endgame problems just aesthetics.
Do you have sample games, because lately, I have seen the opposite too, where SF was winning but blew it.

Maybe a hybrid where AB is also hardcoded after +3 is shown by NN then NN feeds the winning PV to AB part of the engine.
I told my wife that a husband is like a fine wine; he gets better with age. The next day, she locked me in the cellar.

Modern Times
Posts: 2554
Joined: Thu Jun 07, 2012 9:02 pm

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by Modern Times » Mon Apr 29, 2019 4:55 pm

M ANSARI wrote:
Mon Apr 29, 2019 3:20 pm
Are you serious ???? Aesthetic ???? Playing on for a 100 moves when a mate in 12 is there is not aesthetics! An engine is an analysis tool that is supposed to help us analyze a chess position so that we can train and learn from it. This really makes Lc0 useless to analyze as a stand alone engine and you will always need another good engine for a sanity check. Time to figure out a way to fix this with Lc0 and if takes removing the "zero" then be it. I just can't believe that we are still at this stage.
Totally agree. I think people lose sight of the fact that the primary use of an engine for most people is as an analysis tool. Engine to engine matches and Elo ratings are a fairly standard way of roughly assessing which engines are best up to the task. But for Lc0 that approach clearly does not work. Lc0 is both spectacular, and spectacularly useless, both at the same time. But it will get there. Where it has got to today is incredible.
.

Opinions expressed here are my own, and not necessarily those of the CCRL Group.

supersharp77
Posts: 880
Joined: Sat Jul 05, 2014 5:54 am
Location: Southwest USA

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by supersharp77 » Mon Apr 29, 2019 9:38 pm

jp wrote:
Mon Apr 29, 2019 1:30 pm
Jesse Gersenson wrote:
Mon Apr 29, 2019 11:46 am
M ANSARI wrote:
Mon Apr 29, 2019 4:36 am
... really annoying to see an alien like strong engine play these silly moves!
Silly moves?! Does Lc0 get more points if it wins in less moves?
Asserting Lc0 should play according to your preferred aesthetic is off-topic.
Lc0 turns some endgame wins into draws and draws into losses, so you can't call its endgame problems just aesthetics.
Agreed and turns opening wins into Middlegame losses.."" "" etc...etc....Hey where do we get this "New BlueFish Engine"
Or is it just a "Souped Up" Stockfish Dev with lots of Cores? Thx AR :) :wink:

User avatar
M ANSARI
Posts: 3435
Joined: Thu Mar 16, 2006 6:10 pm

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by M ANSARI » Tue Apr 30, 2019 5:31 am

Modern Times wrote:
Mon Apr 29, 2019 4:55 pm
M ANSARI wrote:
Mon Apr 29, 2019 3:20 pm
Are you serious ???? Aesthetic ???? Playing on for a 100 moves when a mate in 12 is there is not aesthetics! An engine is an analysis tool that is supposed to help us analyze a chess position so that we can train and learn from it. This really makes Lc0 useless to analyze as a stand alone engine and you will always need another good engine for a sanity check. Time to figure out a way to fix this with Lc0 and if takes removing the "zero" then be it. I just can't believe that we are still at this stage.
Totally agree. I think people lose sight of the fact that the primary use of an engine for most people is as an analysis tool. Engine to engine matches and Elo ratings are a fairly standard way of roughly assessing which engines are best up to the task. But for Lc0 that approach clearly does not work. Lc0 is both spectacular, and spectacularly useless, both at the same time. But it will get there. Where it has got to today is incredible.
Yes of course ... in no way shape or form do I want to denigrate Lc0 or how it has changed computer chess for the better ... it has. I actually went out and bought a new 2080 Ti card just for it. But it is painful to watch these endgames and it almost feels a bit like Rainman ... where the engine is absolutely brilliant in some things ... but can be totally clueless in what we humans consider simple positions ... AI can suffer from autism?? Take for example this position where I will disregard the many earlier moves that are also delaying the win as you can say that earlier moves might prolong the game but are still strong. But now you have this

[d]4Q3/5K2/3k1P2/2b5/8/8/8/8 w - - 0 104

I mean even a 500 ELO player will think ... hmmm ... Queen check on e7 ... take the bishop and promote another queen and mate with queen. Of course a 3500 ELO engine would not do that and will probably find a quicker mate. But Lc0 here took a further 106 moves to mate !!!! I mean come on ... lets just admit that this is not good and try to fix it! Let's not lose sight of the fact that this is about chess engines helping us play better chess rather than just silicone beasts battling things out. Every move that delays mate is an extra move that can be a blunder or cause a crash. If you are on Chess.com they have this new thing called Puzzle Rush and have to admit I am a little addicted to it. If you go for a mate in 3 rather than a mate in 2 ... it buzzes you as wrong answer. Now that is frustrating and I hated that feature ... but recently I have seen that it has helped me make better moves in normal games as well as I will not play the impulsive move and many times there is a stronger move. So yes Lc0 has been amazing and very exciting to see, but God PLEASE someone fix this trolling thing!!!

jp
Posts: 1345
Joined: Mon Apr 23, 2018 5:54 am

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by jp » Tue Apr 30, 2019 4:22 pm

MikeGL wrote:
Mon Apr 29, 2019 4:30 pm
jp wrote:
Mon Apr 29, 2019 1:30 pm
Lc0 turns some endgame wins into draws and draws into losses, so you can't call its endgame problems just aesthetics.
Do you have sample games, because lately, I have seen the opposite too, where SF was winning but blew it.
I posted last month about the LC0 blog talking about a TCEC 7-man endgame, KNPP vs kbp ( +14 & +153 according to SF), which Lc0 couldn't win. I also remember seeing drawn endgames lost on TCEC or CCC, but didn't note which games. These were not recently.

jp
Posts: 1345
Joined: Mon Apr 23, 2018 5:54 am

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by jp » Tue Apr 30, 2019 4:25 pm

M ANSARI wrote:
Mon Apr 29, 2019 3:20 pm
Time to figure out a way to fix this with Lc0 and if takes removing the "zero" then be it. I just can't believe that we are still at this stage.
It's already non-zero because it has TB rescoring. Making it more non-zero might not solve the problems. I suppose a fork could be tried.

How good is Ender at avoiding the bad endgame play?

carldaman
Posts: 1890
Joined: Sat Jun 02, 2012 12:13 am

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

Post by carldaman » Tue Apr 30, 2019 9:55 pm

Modern Times wrote:
Mon Apr 29, 2019 4:55 pm
M ANSARI wrote:
Mon Apr 29, 2019 3:20 pm
Are you serious ???? Aesthetic ???? Playing on for a 100 moves when a mate in 12 is there is not aesthetics! An engine is an analysis tool that is supposed to help us analyze a chess position so that we can train and learn from it. This really makes Lc0 useless to analyze as a stand alone engine and you will always need another good engine for a sanity check. Time to figure out a way to fix this with Lc0 and if takes removing the "zero" then be it. I just can't believe that we are still at this stage.
Totally agree. I think people lose sight of the fact that the primary use of an engine for most people is as an analysis tool. Engine to engine matches and Elo ratings are a fairly standard way of roughly assessing which engines are best up to the task. But for Lc0 that approach clearly does not work. Lc0 is both spectacular, and spectacularly useless, both at the same time. But it will get there. Where it has got to today is incredible.
+1

What surprises me is how some (smart) folks react when others point out these obvious shortcomings of an otherwise impressive Lc0.

Post Reply