Tweet
update on my system regarding replaced players:
my current idea to handle replaced players is this:
1.)
any player is responsible himself to find a substitute when he leaves.
if a player quits without having a substitute who is willing to take over, leaving the game is considered a concession.
2.)
the goal of this system is to keep the overall score 0 and as fairly distributed as possible.
when a player quits and the substitute takes over, the last ranking turn (*9 turn) is looked at.
if the second ranking turn (turn 19) has NOT been reached yet, a quitting player is counted as having been last. this is to prevent players from hopping from one game to another until their land is optimal or just because they feel bored.
if a player actually IS last or it is not clear from the power graph of the most recent *9 turn whether he is last, he is naturally counted as last too.
so...
if a player quits, what happens depends on his current position in the game, or his score gained from the game respectively.
a)
if the player was in the positives, i.e. his score from the game was positive when he left (usually when a player is high on the power graph, the better the overall rating of the player the higher he needs to be), then he receives half that score added to his permanent rating, as a one-time adjustment and final adjustment for that game.
the joining player gets the same amount deducted from his permant rating, also as a one-time adjustment.
usually, that player will still benefit from taking over a high-ranked civ, so that functions somewhat as a compensation.
the reason not the whole score of the leaving player is counted is to disencourage the leading player from 'dropping' the game with benefit this way. cause if you got the whole score, leaving when you're no.1 would be the best to do.
b)
if the leaving player was in the negatives (the special case +/- 0 makes no difference between a) and b)), but is NOT ranked last in the game, he receives half his negative score from the game as 'penalty' to his permanent score.
to even that negative out, the same amount that the player lost is divided evenly on the plus side between the permanent ratings of all remaining players, including the joining player. (or maybe not the joining player ? not sure yet).
c)
if the leaving player was last (or counted last, see above) when he left, his whole (always negative) score is deducted from his permanent rating.
the same positive amount is then, like in b), added evenly to the permanent ratings of all remaining players, plus the joining player.
in all these cases, permanent rating changes add up to 0, so the game can then continue as normal with the game-dependent (=varying) scores always adding up to 0 too.
the only differences are that the new player's rating is now counted towards the 'average player skill' in this game instead of the leaving player's rating, and that the joining player may have his break-even point (where he ends the game with a total +/- 0 points; including the one-time adjustment) a little bit apart from where it would be had he played from the very start.
im aware this probably isnt perfect, but i think it should be quite fair in average.
the main criteria i wanted to or believe i did accomplish with this rule are:
- sticking to the total sum of all scores being 0 (to prevent a possible degenerate effect on the overall ranking table).
- penalizing players for leaving a game early on, possibly even before the first *9 turn.
- rewarding the remaining players for not quitting (done through the potential small score increase when someone else quits)
- treating replaced players that are last in the game the same way as if they actually dropped out (to counter-measure another abusive way to get out of the game cheaper than you should)
- on the other hand not forcing a player to play the game until the bitter end, however long it takes.
my current idea to handle replaced players is this:
1.)
any player is responsible himself to find a substitute when he leaves.
if a player quits without having a substitute who is willing to take over, leaving the game is considered a concession.
2.)
the goal of this system is to keep the overall score 0 and as fairly distributed as possible.
when a player quits and the substitute takes over, the last ranking turn (*9 turn) is looked at.
if the second ranking turn (turn 19) has NOT been reached yet, a quitting player is counted as having been last. this is to prevent players from hopping from one game to another until their land is optimal or just because they feel bored.
if a player actually IS last or it is not clear from the power graph of the most recent *9 turn whether he is last, he is naturally counted as last too.
so...
if a player quits, what happens depends on his current position in the game, or his score gained from the game respectively.
a)
if the player was in the positives, i.e. his score from the game was positive when he left (usually when a player is high on the power graph, the better the overall rating of the player the higher he needs to be), then he receives half that score added to his permanent rating, as a one-time adjustment and final adjustment for that game.
the joining player gets the same amount deducted from his permant rating, also as a one-time adjustment.
usually, that player will still benefit from taking over a high-ranked civ, so that functions somewhat as a compensation.
the reason not the whole score of the leaving player is counted is to disencourage the leading player from 'dropping' the game with benefit this way. cause if you got the whole score, leaving when you're no.1 would be the best to do.
b)
if the leaving player was in the negatives (the special case +/- 0 makes no difference between a) and b)), but is NOT ranked last in the game, he receives half his negative score from the game as 'penalty' to his permanent score.
to even that negative out, the same amount that the player lost is divided evenly on the plus side between the permanent ratings of all remaining players, including the joining player. (or maybe not the joining player ? not sure yet).
c)
if the leaving player was last (or counted last, see above) when he left, his whole (always negative) score is deducted from his permanent rating.
the same positive amount is then, like in b), added evenly to the permanent ratings of all remaining players, plus the joining player.
in all these cases, permanent rating changes add up to 0, so the game can then continue as normal with the game-dependent (=varying) scores always adding up to 0 too.
the only differences are that the new player's rating is now counted towards the 'average player skill' in this game instead of the leaving player's rating, and that the joining player may have his break-even point (where he ends the game with a total +/- 0 points; including the one-time adjustment) a little bit apart from where it would be had he played from the very start.
im aware this probably isnt perfect, but i think it should be quite fair in average.
the main criteria i wanted to or believe i did accomplish with this rule are:
- sticking to the total sum of all scores being 0 (to prevent a possible degenerate effect on the overall ranking table).
- penalizing players for leaving a game early on, possibly even before the first *9 turn.
- rewarding the remaining players for not quitting (done through the potential small score increase when someone else quits)
- treating replaced players that are last in the game the same way as if they actually dropped out (to counter-measure another abusive way to get out of the game cheaper than you should)
- on the other hand not forcing a player to play the game until the bitter end, however long it takes.
Comment