A design of PoE ( Proof of Excellence) token #1
Why PoE token matters? Please read https://medium.com/@RagnarDoge/history-of-proof-of-excellence-and-new-possibility-decentralized-game-a7bb75652fc6
The key of PoE is to run a tournament. The game for the tournament should be: Verifiable and Unsolved.
Verifiable
For winners to get the winning rewards — tokens, the result should be verifiable by code on the chain (smart contracts).
Unsolved
A game with all predictable outcomes is called a solved game, which is meaningless for PoE, because computers could easily reach the TOP level of excellence, and the “game” is over forever.
The level of excellence should not be capped. Or, precisely to say, the game should leave “infinite” space for players to explore, even the most powerful AI players, in the far future. Somehow, PoW (proof of work) is a PoE of a hash-computing game.
Randomness
Randomness is not a must, but if randomness is involved, it should only make the results match the excellence-ness of the players.
AI players
AI players should not be worried but welcomed. The incentives given by the PoE token could be a great driving force for the development of AI.
We can imagine, in the future, for small AI teams, PoE tokens could be the direct income for their AI development. With proper design, PoE tokens will have a great positive impact in AI industry as Bitcoin has done in ASIC / chip industry.
Go Game is the one
Go Game is a good unsolved game with simple rules. Two players play stones on the board by turn, try to occupy as much area as possible, and stones could not breath (without Qi) are captured by the opposite players.
Quote from https://en.wikipedia.org/wiki/Go_(game)
Despite its relatively simple rules, Go is extremely complex. Compared to chess, Go has both a larger board with more scope for play and longer games and, on average, many more alternatives to consider per move. The number of legal board positions in Go has been calculated to be approximately 2.1×10¹⁷⁰, which is vastly greater than the number of atoms in the observable universe, estimated to be of the order of 10⁸⁰.
Go Game is a game that the possible outcomes could never be listed out by our computers. Thus, it’s a perfect game for any game AI to improve itself.
The verifiability of Go Game
How to tell who wins in a go game?
Scoring could be confusing for humans in the real game situation. But actually, it’s not very complicated, and verifiable by computer program. There are basically two ways of scoring a go game. (check https://senseis.xmp.net/?Scoring)
Area scoring: Surround points + Stones on the board
Territory scoring: Surround points - Captured stones ± Seki…
The two ways generally yield the same results. But area scoring is easy to understand and does not need special handling for seki (mutual life) situation.
How to score go with Area scoring?
There are black stones, white stones, and empty points on the board.
We count black stones as points for black side, white stones for white side.
For empty points, there’s three situations:
- empty points surrounded by black stones and border, are points for black;
- empty points surrounded by black stones and border, are points for black;
- empty points touched both with black and white stones, are unoccupied, does not belong to black neither white.
Go game on blockchain
Can people enjoy playing go game on blockchain now? Na, not yet.
People has tried before, but not quite successful. Some relevant pages here.
https://www.reddit.com/r/ethereum/comments/85mnxv/ethernalgo_is_a_real_go_game_running_100_on/
https://etherscan.io/address/0x34bd7814bfef669936f05e355abd882acfc569b2#code
https://cointelegraph.com/news/play_go_game_with_bitcoin
The good news and the bad news
Good news: Millions of people are playing go game
Quote from https://en.wikipedia.org/wiki/Go_(game)
A 2016 survey by the International Go Federation’s 75 member nations found that there are over 46 million people worldwide who know how to play Go and over 20 million current players, the majority of whom live in East Asia.
Some are actually playing online https://online-go.com/
Good news: Big companies are playing go game
AlphaGo from Google: https://deepmind.com/research/case-studies/alphago-the-story-so-far
OpenGo from Meta: https://ai.facebook.com/tools/elf-opengo/
FineArtGo and PhoenixGo from Tencent:
https://en.wikipedia.org/wiki/Fine_Art_(software)
https://github.com/Tencent/PhoenixGo
Good news: AI startups and open source teams are playing go game
https://github.com/leela-zero/leela-zero
https://github.com/lightvector/KataGo
Bad news: It’s pretty hard to play go game on the chain
There’s no fully decentralized go game platform for people to play, yet.
In the next article, I will propose a design of on chain go game with details.
