@helvetica Fall Guys has an interesting version of this—every squad has a score which is the addition of points scored based on where you finished. I've had a lot of situations where the "slower" squad mates actually sway the result because they push the point total over the line, which rewards them finishing and actually makes them feel far more valuable than they might otherwise.
"no individual brings the team down" means that removing any person cannot improve team rank, or adding a person cannot decrease team rank.
"every individual matters" means that adding a person must (potentially) change team rank. since it can't decrease team rank, it must increase it.
so any such scheme that meets those criteria will be biased toward larger teams. but the bias can be made small, eg it might take +100 weak players to match +1 strong player
@helvetica ok, here's a scheme that feels somewhat intuitive to me:
compare two teams by matching the players up in 1v1 battles, strongest first. they're pure hp battles, 100hp defeats 90hp, leaving 10hp left. the winner then heals n% and battles the next in line. adjusting n% adjusts how much size of team overcomes strength of team
@helvetica that feels hard to do. Brainstorming here a little and came up with multiple scores based on different skills. You could be bad at one type of score but better on others, bonus points to make each score orthogonal to another, like rock paper scissors (i.e. being better at one will affect another).
Grab the best of all teams (or median) and have four leaderboards. One per each skill and one for all best scores combined. Multiple awards ensue plus the combined one. A triathlon of sorts
@helvetica hmm probably need to give either handicaps on large teams or advantages on smaller ones similar to the upkeep system of warcraft 3 maybe. Or segment by team size range to minimize bias.
But yeah probably you'd need to iterate and test those. You probably need to see the system in motion or at least simulate it on a excel or something, seems hard to easily test anyway :p. It's an interesting challenge for sure.
@helvetica thinking about it some more, would splitting out small/large teams be acceptable? most of the fancy bottom/top n% just falls apart on less than ten players I think?
@helvetica Maybe take the score of the players in the top half of the team and average them. Whoever is at the bottom doesn’t have to worry about actively bringing their team down. Could add some sort of weighting depending on team size to even out two people being excellent on a team of 3 vs two people being excellent on a team of 1000.
@aaronschendel@helvetica We didn’t want to go with only top half averages because then players would just make small teams of only fast players and their average would easily be better than a large team with a variety of skill levels.
@helvetica Isn't not allowing for a player to feel like they're "bringing the team down" directly at odds with team based leaderboards? Individual achievement of team members is what drives a team up or down the leaderboard.
Add comment