PNW District Point Distribution

[Ian Curtis]

The Pacific Northwest District is interesting this year in that there are the same number of teams and events as last year (although it appears one team registered but did not compete this year). Given this, it seemed reasonable to me that the district point cut-off would be in about the same place as it was last year.

But! It wasn't. This year the last team to qualify on points had 64 points, where last year it was 58 points. This is a non-trivial number of points (and teams!) in the flat spot where the point structure matters most.

Anecdotally, it seemed like the robots were better this year. So, I initially assumed it was mostly a zero sum game... the best robots weren't so much better than the middle tier, and the middle tier ended up with some points that had been previously belonged to the elite.

Looking at the data, this doesn't appear to be the case. The top 25% of the distribution didn't change much (they got slightly better), and the bottom 25% didn't change much (they got a tad worse). However, the middle 50% improved significantly (5-6 points on average).

Overall, the district awarded 351 points more than it did last year, or about a 4% increase and they overwhelmingly went to the middle tier. Where did these points come from? Were there fewer third plays? Does the non W-L-T point structure for qualification rounds award more points than the old system?

[SoftwareBug2.0]

I can think of at least two things that would make a difference in where the bar is:

-At last year’s Oregon State University district 1425 and 2811 were both at their third event and pulled about 70 pts out of the pool each. This year’s Philomath (same general area) event had the winning alliance’s captain’s (1983) points disappearing again but their first pick (955) was not at their third event. 70 extra points is enough to raise 14 teams by 5 points.

-Last year there were only two chairman’s award winners who wouldn’t have moved on based on points. This year there are three, and one of them is not close to the point cutoff. For the purposes of figuring out where the cutoff will be that team their chairman’s award is implicitly worth 30 about points.

[Ian Curtis]

Originally Posted by SoftwareBug2.0

I can think of at least two things that would make a difference in where the bar is:

-At last year’s Oregon State University district 1425 and 2811 were both at their third event and pulled about 70 pts out of the pool each. This year’s Philomath (same general area) event had the winning alliance’s captain’s (1983) points disappearing again but their first pick (955) was not at their third event. 70 extra points is enough to raise 14 teams by 5 points.

-Last year there were only two chairman’s award winners who wouldn’t have moved on based on points. This year there are three, and one of them is not close to the point cutoff. For the purposes of figuring out where the cutoff will be that team their chairman’s award is implicitly worth 30 about points.

70 points per team is presumably concentrated in elimination points. Even with those points re-entering the pool, how do they get divied up into 5 points per team?

The actual normal distribution we are using to award qualification points to teams this year must be more normal than the approximate distribution the real world gives you from W-L-T... I suppose it makes sense that would benefit the middle class more than the tails.

[SoftwareBug2.0]

Originally Posted by Ian Curtis

70 points per team is presumably concentrated in elimination points. Even with those points re-entering the pool, how do they get divied up into 5 points per team?

The actual normal distribution we are using to award qualification points to teams this year must be more normal than the approximate distribution the real world gives you from W-L-T... I suppose it makes sense that would benefit the middle class more than the tails.

Well, 2811's points would have looked something like this:
qualification 16
alliance selection 15
elimination 30
award 10 (?)

The points don't have to be directly awarded to a specific team to make them do something to the rankings. For example, if you took whatever team got the least points and gave them 200 extra points then not only does the first place team now have more points, but the second place spot now has the number of points that the previously 1st place team had. And the third place slot now has the team that used to be in second. So now most of the slots have moved up by 1-2 points each despite only one team's point total changing.