go insei is built as a guide to a deeper understanding of the game of go.

Source: hikago

Go AI

There’s something magical about the game of Go. For thousands of years, it has captured the imagination of those who want to learn what it is to learn, to think about what thinking means.

branton demoss

The recent advent of strong, open source Go AI that can beat top professionals has changed the way we can learn this game. go-insei aims to be a part of this.

The basis of go-insei

KataGo

Very strong open source AlphaGo-style AI created by lightvector.

go-insei utilizes the current strongest networks for its analysis.

WGo.js

An excellent javascript library for purposes of the go game, written by Jan Prokop 6d.

go-insei uses it to display records of professional games, annotated with the strongest variants created using katago.

High-quality SGFs

Game records are selected from the well-maintained database of Go4Go.

AI-commentary on pro games

Each game in the database has been analyzed by the AI of katago with the depth of 10,000 visits per move, resulting in playing strength about two handicap stones (or more) stronger than professionals players.

go-insei displays the moves played by the professional player with A, providing the rank of the move along with any score and winrate loss in the commentary. The top five moves are provided as B to E (or F, if none of the top moves has been played), along with the ensuing principal variation.

After each move, we have added an “empty node” to invite the reader of the game record to guess the next move.

Try yourself

Vivamus integer non suscipit taciti mus etiam at primis tempor sagittis sit, euismod libero facilisi aptent felis blandit cursus gravida sociis erat ante, eleifend lectus nullam dapibus netus.

– Name @company

Vivamus integer non suscipit taciti mus etiam at primis tempor sagittis sit, euismod libero facilisi aptent felis blandit cursus gravida sociis erat ante, eleifend lectus nullam dapibus netus.

– Name @company