fairplay ?

All the clues you need to find this out are in the paper.

I'll give you another, though- it only posts "GG" or "GG!" after the game. It doesn't always post this, and I suspect it might have been manual (at least on some games).

Up.

bo_sox48 wrote: ↑

Tue Nov 03, 2020 11:49 pm

Some researchers worked with us to develop a bot of their own, (...)

Could you give us the name of that bot on Webdiplomacy so that we can find the link to the games it played ?

This is cool! When was this?

Also:

n anonymous games against humans

Clearly, they don't know if the games were against humans or not, since neither do we.

bo_sox48 wrote: ↑

Tue Nov 03, 2020 11:49 pm

Some researchers worked with us to develop a bot of their own.

Thank you!

Could you point to the games played with the bot ?
How was it named ?

Some researchers worked with us to develop a bot of their own, and after it had played a number of games against humans we tabulated their bot's Ghost Ratings. It ranked 23rd in the gunboat ratings. It would rank lower if inactive players were included, but at the time we chose not to include inactive players because the ratings typically did not include inactive players back then. Even so, it was a very impressive development.

Hello,

I have just found this article on the internet.

The author claims his agent (bot) achieves a rank of 23 / 1128 human players on webdiplomacy.net, ie here.

Is he talking about "our" bots ? Or did he try an improved version ?
I don't remember a challenge took place with them. If not, a hum vs 6 bots is not the same as a bot among numeours humans...