Hosting Scores

[Peter]

(This was copied from another post)

We have added hosting scores to calculate the reliability of a host:

We count all (Tournament) games a player has hosted so far (let's say 10 games).
Then we give each type of game a theoretical number of points depending on the number of human players.

2 humans = 1 point
3 humans = 2 points
4 humans = 3 points

Let's assume the host has played 10 x 2 human player games = 10 theoretical hosting points.

We then count the number of human players who abandoned the game AFTER round 2 (if the game is abandoned earlier, then we don't count any points, in favor of the host), by subtracting 1 point for each human who did not finish round 12.

Let's say from the 10 games with 2 human players, only 3 ended in round 12 with both human players still in game. Then the host would have a hosting success rate of 30%. Of course this does not measure if a non-hosting player left the game voluntarily or due to his conncetion failure. But it still measures how many players were able to make it till the end.

Hosts are then calssified based on the average outcome of all their hosted games:

>80% very good
>60% good
>40% okay
<40% bad

For a player hosting his first 3 human player game and ending up with 1 less human player, it would mean 1 out of 2 points = 50%. Of course the average evens out when more "reliable" games are finished.

We are now evaluating how that system turns out for most of the hosting players, but its likely, it will help better differentiate unreliable hosts (for whatever reasons, intentionally or technical does not matter) and then be able to show their sucess/failure rate in the lobby.

This somewhat turns around the perceived position of strength of a host back to the players not hosting, but then the host still has the inherently "gameplay advantages" of being the host.

[gdgdgdgd]

How do you open a port for hosting games?

[joechompa]

How do I host from a Mac?