2015 District Point Structure

[Peyton Yeung]

Since matches are no longer announced as win-loss-tie, how will the district points be awarded?

[MikeE]

N pts per win awarded only to the alliance that progresses, is effectively 2N points for each alliance that progresses to the next round.
I suspect the language will be tweaked but it will have the same effect as before.

[dag0620]

http://www.usfirst.org/roboticsprograms/frc/blog-changes-to-district-points-2015

[Hallry]

http://www.chiefdelphi.com/forums/showthread.php?threadid=131965

[Steven Donow]

Here's what I think should be done (haven't given a super detailed amount of thought, so it probably has flaws):

Every score of every match during qualifications at the event is averaged
Every match where you score above the average gives you 2 points, every match below the average gives 0, and getting the average gives you 1.

Unfortunately you won't know where you stand during the event, but this solves the potential problems of basing it on the actual value of points scored:

1. Who knows what an average week 1 match score will be?
2. An average week 4 match score will be higher than an average week 1 score.
3. Taking the above into account, later events would have a HUGE advantage over earlier events.

[nlknauss]

Here's what I think should be done (haven't given a super detailed amount of thought, so it probably has flaws):

Every score of every match during qualifications at the event is averaged
Every match where you score above the average gives you 2 points, every match below the average gives 0, and getting the average gives you 1.

Unfortunately you won't know where you stand during the event, but this solves the potential problems of basing it on the actual value of points scored:

1. Who knows what an average week 1 match score will be?
2. An average week 4 match score will be higher than an average week 1 score.
3. Taking the above into account, later events would have a HUGE advantage over earlier events.

Number 3 is a huge factor here. The ranking method should be balanced from the first week to the last so there is no advantage given to those who compete later. This is a factor to consider while strategizing and I'm hoping clarification on this comes out sometime during week 1. I'm sure they have something up their sleeve.

[scottandme]

They already said it'll be based on the previous point distributions, so it'll probably be a percentile based system based on rank. So 95th percentile scales to 12 "wins", 90th percentile scales to 11 "wins", and on and on. Presumably they're just running the data to find what the average point distribution is and apply that to each event based on it's size.

[MikeE]

They already said it'll be based on the previous point distributions, so it'll probably be a percentile based system based on rank. So 95th percentile scales to 12 "wins", 90th percentile scales to 11 "wins", and on and on. Presumably they're just running the data to find what the average point distribution is and apply that to each event based on it's size.

I agree that's the most likely approach but it does lose one of the features of the previous system that a #1 seed with a 12-0 record has a small points advantage over a #1 seed at a different event with a 10-0 record.
In the big picture this doesn't really affect top teams, but in the middle of the table around the District Championship cut off point 2 points per event does matter. For example in the New England District there were 11 teams within 2 points of the cutoff.