Announcing GoReviewPartner - v0.15 with Pachi and PhoenixGo

[johnsmith]

Thanks for this tool! It works really well on windows.

My 20 cents: maybe incorporate Net Win % alongside the existing ones.

[goame]

Thx for this great tool

Is Gnugo so good at score estimation?
I looked at a game and Gnugo said that black was for some moves more or less +10.5 but Leela said 80% winning chances for white.

It would be good to have some important points:
-2 more buttons to jump to the last or first move.
-play stones more in the middle of the intersections.
-more nice and round stones.
-show white and black moyo.
-show white and black territory.

-click in the right window on variation a, b, c,... and let me see every move with the arrow keys, so it is easier to remember and better to learn.
-use different colors for best, second ... variation.
-take only the 6 moves (variations) with the highest playouts, it doesn't make sense the have one variation with 64400 playouts and another mit 47 playouts.
-display variation 2-6 only if they have at least 25% playouts from the first variation with the most playouts.
-an infinite analysis so I can still analyse the game when looking on it after your first analysis is finished.
-also the number of playouts/simulations of the first analysis for each move should be saved. When I analyse the game later and get higher playouts, the data should change automatically.

There is a new version rn-3.11.1 https://github.com/zakki/Ray/releases see also the issues and ask there if you need help;)

[pnprog]

It is this line in gomill/sgf_grammar.py: (?P<I> [A-Z]{1,8} ) # PropIdent
It says that a property identifier has length 1-8. But MULTIGOGM has length 9.

Excellent!

I will see if I can submit/push a patch to the gomill team. At least I can fix that in my software.

[pnprog]

My 20 cents: maybe incorporate Net Win % alongside the existing ones.

Added to my to do list for v0.5

[pnprog]

Is Gnugo so good at score estimation?
I looked at a game and Gnugo said that black was for some moves more or less +10.5 but Leela said 80% winning chances for white.

Frankly speaking, no, it sucks until the game reaches the end game
But I won't drop Gnugo support for now, I plan to separate the analysis in different tools, with one tool for Gnugo, one for Leela, one for rayon or any other bot that I can work with. You may ask who may need a review by Gnugo, but I believe Gnugo moves are very ok for 15k and above, plus, the analysis time would be ways faster than MCTS bots. Gnugo has plenty of things to offer for beginners.

-2 more buttons to jump to the last or first move.

Added to my todo list for v0.5

-play stones more in the middle of the intersections.

This is the fuzzy stone parameters in settings, change it to 0

-more nice and round stones.

At least, the stones should be round right now For a better looking stone, it's not in my priority list, but it will come one day

-show white and black moyo.
-show white and black territory.

I do not know any method to do that in fact. The closest I got so far is asking Gnugo to quickly play until the end of the game to find out, but it's slow, and does not really work when there are group with unclear life and death status.

-click in the right window on variation a, b, c,... and let me see every move with the arrow keys, so it is easier to remember and better to learn.

Yes, I plan to do that soon. I plan to use the mouse wheel instead of the keyboard, but that's the idea. Yeah, maybe I should use both keyboard and mouse wheel, for laptop users with out mouse wheel.

-use different colors for best, second ... variation.

Currently, the letters indicate the order from best (A) to not so good (B, C, D) and blue/red to differentiate winning rate >50% I do not have much margin to change the colors in fact

-take only the 6 moves (variations) with the highest playouts, it doesn't make sense the have one variation with 64400 playouts and another mit 47 playouts.
-display variation 2-6 only if they have at least 25% playouts from the first variation with the most playouts.
-an infinite analysis so I can still analyse the game when looking on it after your first analysis is finished.
-also the number of playouts/simulations of the first analysis for each move should be saved. When I analyse the game later and get higher playouts, the data should change automatically.

Those are idea that have more to see with Leela internal algorithm. Not something I can change at my level. It will probably evolve as Leela improve in the near future.

There is a new version rn-3.11.1 https://github.com/zakki/Ray/releases see also the issues and ask there if you need help;)

Are you a member of Ray developer team????

[roy7]

-show white and black moyo.
-show white and black territory.

I do not know any method to do that in fact. The closest I got so far is asking Gnugo to quickly play until the end of the game to find out, but it's slow, and does not really work when there are group with unclear life and death status.

Leela has an 'influence' command I think it's called. You could parse that for a bit of info to display.

[pnprog]

Hi!
I finally had a quick try with Ray (rn-3.11.1) under Windows.

As it stands now, Ray seems to not offer that much useful information, see by yourself the result of GTP command "genmove black":

ray_stderr.png (47.82 KiB) Viewed 19831 times

I think the interesting part is the "Best Sequence" that should shows what Ray expects to be the following moves and the different winning percentages (I do not know the meaning).

For comparison, Leela always propose a several sequences with winning percentage, that's much more information useful to game review.

[djhbrown]

In fact, it was more a issue with their declining sight.. And so I implemented this sort of feature in all my Go related softwares.

actually, it isn't. i noticed that behaviour in Japan donkeys' years ago, and realised that it is their sense of the aesthetic - even pros do it in tv games, as do their pro commentators - Shinto is very close to nature, and there are no straight lines in nature (except in Feynman diagrams and cloud chambers tracking subatomic particles). Some players, determined to keep the board beautiful, would even occasionally shift their previous stone a smidgen back offline after i had repositioned it on the spot before making my move.

of course, in the wrong hands, misplacing stones can be used to deceitfully try to throw opp off.

Nevertheless, having the fuzzy placement option in your client makes it all more beautiful.

[djhbrown]

Is Gnugo so good at score estimation?

no, it's not.

[djhbrown]

-show white and black moyo.
-show white and black territory.

I do not know any method to do that

i do
https://papers.ssrn.com/sol3/papers.cfm ... id=2818149

[Javaness2]

You can also read some more respectable papers on score estimation, for example, a bit of a daddy

https://webdocs.cs.ualberta.ca/~mmueller/ps/goeval.pdf

[pnprog]

You can also read some more respectable papers on score estimation, for example, a bit of a daddy

https://webdocs.cs.ualberta.ca/~mmueller/ps/goeval.pdf

Thanks for the link!

I made a bit of research regarding the bot mentioned inside (Explorer). My guess is that those score evaluation methods are similar to what is been used (and probably improved) into Gnugo and Fuego.

Maybe I should have a better look at Fuego for a score estimation.

[roy7]

Once upon a time I think the Leela author told me pacci has the best end game scoring.

[Javaness2]

https://webdocs.cs.ualberta.ca/~mmueller/ps/goeval.pdf is not too bad either

[pnprog]

https://webdocs.cs.ualberta.ca/~mmueller/ps/goeval.pdf is not too bad either

Is that the same document as in your former message?

By the way, I have finished implementing my todo list for v0.5
I will test a bit more, then release an update tomorrow.