This is a Fantasy FIRST league operating by the standard rules of SLFF, with one change: All drafts are list-only, and full lists for all events must be submitted all at the same time. This is meant for people or teams who would like to test automatically generated lists, AIs, NNs, Elo systems, whatever.
There will likely be “simple” AIs that also play inside the league, such as highest average points in the last X years, or highest match win rate in the last X years, etc. These will be added at a later time once we see how many people are interested in competing.
The Most Questionable Decisionmaker (Brian Maher, TDav540, saikiranra, Kellen Hill)
Team Lean Cuisine (cobbie, hailey.faiella)
AmonBot (drwhorx)
Eloquence (Caleb Sykes)
The Final Pick (tindleroot)
I had an interesting thought when I saw this thread. Instead of drafting teams, since we are all making full lists, what if everyone was scored over their whole picklist?
For example, say I pick teams in the order a, b, … z, for an event with 50 teams, then my score for the event is 50/50 * s(a) + 49/50 * s(b) + … 1/50 * s(z), so my score will be maximized if I list the teams in descending order of score (i.e. perfect prediction). This would allow us to rate picklists in their entirety and remove any aspect of luck based on picking order.
One other possibility is, if we want to stay true to the spirit of SLFF, is to draft each event with a large number (say, 10000) of random draft orders and take the average score for each player for each event and add them all up.
I like the concept a lot, I was already thinking about using something like this on my own to compare against everyone else.
I think there’s probably be a better scoring function out there than a linear drop from 1 to 0, but I’d have to dive more into the data to find what it would be.
I like this a lot because then we’re not adding an arbitrary scoring metric. Also, this way your list still influences everyone else’s scores so it’s still a direct competition, and not like recycle rush where everyone just does their own thing and is scored independently.