Search found 6278 matches

by RobertJasiek
Sun Apr 12, 2026 11:23 pm
Forum: Forum/Site Suggestions and Bugs
Topic: Database Errors
Replies: 0
Views: 1539

Database Errors

During the previous 24h, L19 was unavailable for several hours due to database errors. An error message appeared with the suggestion to contact an admin but no email address of any admin was shown at the error message so one would not know how to report the occurrence of database errors when they ...
by RobertJasiek
Mon Apr 06, 2026 11:20 pm
Forum: Study Group
Topic: Evaluation of Local Endgames with Long Sequences
Replies: 0
Views: 2164

Evaluation of Local Endgames with Long Sequences

It is insufficient to consider values of followers and individual moves. It is also mandatory to verify whether long alternating sequences of several plays are worth playing successively to know from which followers to derive the initial move value.

I criticise every unverified local endgame ...
by RobertJasiek
Fri Apr 03, 2026 9:05 am
Forum: Computer Go
Topic: Pachi 12 for Windows
Replies: 40
Views: 195878

Re: Pachi 12 for Windows

I do not know how Pachi does it but my study with KataGo has let me redefine joseki as "correct sequence for several standard openings judged by KataGo with sufficiently many playouts".
by RobertJasiek
Tue Mar 31, 2026 10:59 pm
Forum: Announcements
Topic: LifeIn19x19.com v2.0!
Replies: 43
Views: 49015

Re: SEND OUT THOSE EMAILS ALREADY ABOUT L19 RESURRECTION


How are we to get John, Bill, and others


Bill stopped writing, IIRC, in summer 2021 presumably due to personal circumstances, which I suspect might be age- / health-related. It was two months before my publication of very much mathematical endgame theory, in which he would have been very ...
by RobertJasiek
Sun Mar 29, 2026 8:47 am
Forum: Professionals
Topic: "Code Name: BlueSpot" pro vs AI handicap matches
Replies: 4
Views: 1786

Re: "Code Name: BlueSpot" pro vs AI handicap matches

In the pre-AlphaGo-era, some professionals claimed that they would always beat God on H3, or was it H4? Anyway, I expect this prediction is in danger. OTOH, this is the chance for humans to beat AI down to H9 by a) exploiting novel trivial weaknesses of the neural net or b) complex life and death ...
by RobertJasiek
Tue Mar 17, 2026 11:32 pm
Forum: Study Group
Topic: Endgame values with area counting
Replies: 5
Views: 2015

Re: Endgame values with area counting

Here are comments on part 1.

You study area deiri counting. Usually, one would use area miai counting.

For area miai counting, I (and others before me) have stated the principle "Usually, an area move value is 1 larger than a territory move value."

Unsurprisingly, for area deiri counting, you ...
by RobertJasiek
Tue Mar 10, 2026 11:52 pm
Forum: Announcements
Topic: LifeIn19x19.com v2.0!
Replies: 43
Views: 49015

Re: LifeIn19x19.com v2.0!

Currently, when logged in, the Hamburger Menu "Quick links" contains "New Posts". One does not see this option when logged out. When logged in, initially the second row of the menu bar is "Board index . Life In 19x19.com General Topics . Announcements". When logged out, it is just "Board index".

I ...
by RobertJasiek
Tue Mar 10, 2026 9:50 am
Forum: Announcements
Topic: LifeIn19x19.com v2.0!
Replies: 43
Views: 49015

Re: LifeIn19x19.com v2.0!

Many thanks to everybody having brought L19 back to life! It is good that updated code makes hacking less likely.
by RobertJasiek
Mon Jun 02, 2025 7:54 am
Forum: Announcements
Topic: Major new article on go in German
Replies: 7
Views: 26039

Re: Major new article on go in German

if intelligent aliens exist on another planet, the game they would play would be go.

For a game to be reinvented remotely, several game design decisions have to be made. Go (with some reasonably simple ruleset) requires a comparatively small number of such design decisions. Simpler games, such as ...
by RobertJasiek
Sat May 10, 2025 1:47 am
Forum: Computer Go
Topic: Absolute Zero
Replies: 1
Views: 19249

Absolute Zero

In the paper Absolute Zero: Reinforced Self-play Reasoning with Zero Data

https://arxiv.org/pdf/2505.03335

AI learns from scratch on its own to create and prove problems of reasoning incl. deduction, abduction and induction for coding or mathematics in the manner of Alpha(Go)Zero or KataGo ...
by RobertJasiek
Sat Mar 29, 2025 8:44 pm
Forum: Computer Go
Topic: AI Study
Replies: 7
Views: 22132

Re: AI Study

MOST of the time we do not have the situation where A>B at a lower number of playouts but B>A at a very high number of playouts.

It is my observation in Katago study that it changes the relative order of most move candidates frequently for low numbers of playouts. A>B and B>A tend to change ...
by RobertJasiek
Sat Mar 29, 2025 9:31 am
Forum: Computer Go
Topic: AI Study
Replies: 7
Views: 22132

Re: AI Study

MOST of the time we do not have the situation where A>B at a lower number of playouts but B>A at a very high number of playouts.

It is my observation in Katago study that it changes the relative order of most move candidates frequently for low numbers of playouts. A>B and B>A tend to change ...
by RobertJasiek
Fri Mar 28, 2025 7:08 am
Forum: Computer Go
Topic: AI Study
Replies: 7
Views: 22132

Re: AI Study

Neural net computer go is not mathematical theorem proving, that is, we do not know whether a move is perfect play. Some practical search limit is needed, although one can fill one's RAM if a particular query is important. Go tournament players' thinking time are a different consideration and ...
by RobertJasiek
Fri Mar 28, 2025 6:42 am
Forum: Computer Go
Topic: AI Study
Replies: 7
Views: 22132

Re: AI Study

For this purpose, start with one joseki corner and three single stone corners. To limit the amount of study, occupy the diagonally opposite corner with a 4-4 stone. In the two adjacent corners, each choose either a black or a white stone. Either such stone is 3-4 or 4-4 (because mostly others are ...
by RobertJasiek
Fri Mar 28, 2025 1:49 am
Forum: Computer Go
Topic: AI Study
Replies: 7
Views: 22132

AI Study

The basics of study with Katago concern calibration with komi, numbers of playouts, precision of values and study modes.

KOMI

For every ruleset, there is a komi so that initially the winning probability is about 50.0% and the expected score is about 0.0. One can set such a komi or accept other ...