RPI Futures Week 8 (10/21/19)

[bluepenquin]

Missouri
15-5 (9.7%)
Avg 7.8 Range: 3rd to 11th

14-6 (22.3%)
Avg 11.7 Range: 7th to 17th

13-7 (28.6%)
9th: <1%
10th: 1%
11th: 2%
12th: 7%
13th: 12%
14th: 14%
15th: 15%
16th: 16%
17th: 14%
18th: 9%
19th: 5%
20th: 4%
21st: 1%

12-8 (21.1%)
Avg 19.7 Range: 12th to 28th

[bluepenquin]

BYU
17-1 (7.5%)
Avg 12.2 Range: 9th to 16th

16-2 (28.3%)
10th: <1%
11th: 1%
12th: 5%
13th: 7%
14th: 12%
15th: 19%
16th: 19%
17th: 15%
18th: 10%
19th: 6%
20th: 3%
21st: 1%

15-3 (33.4%)
Avg 19.4 Range: 14th to 26th

14-4 (21.7%)
Avg 24.0 Range: 18th to 32nd

[bluepenquin]

Colorado State
18-0 (61.9%)
9th: <1%
10th: <1%
11th: <1%
12th: 1%
13th: 2%
14th: 6%
15th: 7%
16th: 11%
17th: 15%
18th: 16%
19th: 15%
20th: 10%
21st: 10%
22nd: 3%
23rd: 2%
24th: 1%

17-1 (30.6%)
Avg 21.3 Range: 15th to 29th

[bluepenquin]

Oct 22, 2019 10:45:02 GMT -5 trianglevolleyball said:

blue Would you mind doing UNC’s T45 odds if they a) win out (18-10) and b) lose one more (17-11)? Know it’s a stretch but interested if there’s still a marginal chance for a tourney bid.

1000 simulations - 15-3 happened only 6 times.  The RPI rank for those 6 was:  42, 47, 42, 46, 50, 45.  That is a pretty small sample size.  The RPI would likely put them on the bubble - with a couple wins against Florida State being in the top 50.

One more loss and the average was 53.5 RPI.  There were only 5 of 48 times they were under 50.  

[bluepenquin]

Oct 22, 2019 10:12:58 GMT -5 cindra said:

Oct 22, 2019 7:19:51 GMT -5 bluepenquin said:

Pittsburgh:  If 18-0 in Conference (32.2%):

1st: <1%
2nd: 30%
3rd: 39%
4th: 20%
5th: 7%
6th: 3%
7th: <1%

If 17-1 (38.7%)

2nd: 8%
3rd: 18%
4th: 26%
5th: 18%
6th: 17%
7th: 9%
8th: 3%
9th: 1%
10th: <1%


If 16-2 (21.2%)

Average RPI is 6.8 with a range of 3rd to 15th.

Thanks. How do you do all this anyway? Statistician in the real world or just a hobby?

Actually not much of a statistician (real world or hobby).  I do enjoy sports and I enjoy analytics.  No where close to the Statistics knowledge needed to do real analytics work, so I have to rely on just big picture understanding of other people's work and apply to things I am interested in.  RPI Futures is just a lot of manual work and using the built in statistical/analytics work done by Pablo.  I am doing nothing much more than remedial statistical work done in HS.

[bbg95]

Oct 22, 2019 14:27:44 GMT -5 ay2013 said:

Oct 22, 2019 13:45:16 GMT -5 bbg95 said:

With regard to the "false" idea that Nebraska's loss to Northern Illinois hurt them in 2017, this was actually covered in-depth by ESPN at the time in this article. The whole thing is worth reading, but the way the committee justified the discrepancy between the early look and the final bracket was (quoting from the article): "The reveal is not to be taken as set in stone. It's just to stimulate discussion in the sport...In short, it's a suggestion of what the committee is thinking at the time, but not anything that will bind them once they actually start putting together the bracket." They also added that the early reveal doesn't have the final data and is done via conference call rather than in person like the real bracket is. Now, just because the committee says that, you don't necessarily have to accept their explanation. But according to them, the Northern Illinois loss did indeed hurt Nebraska.

Well the committee already knew of the UNI and Oregon loss at the first reveal. If they are hanging their hat on the losses hurting Nebraska in the end and using a conference call v in person as justification, perhaps the reading between the lines conclusion is that there are some influencers on the committee that don’t like Nebraska 🤣

Well, I'm not defending the committee's reasoning or saying they got it right. But if they said the loss mattered to them, I take them at their word.

[bbg95]

Oct 22, 2019 14:42:04 GMT -5 bluepenquin said:

Oct 22, 2019 14:27:44 GMT -5 ay2013 said:

Well the committee already knew of the UNI and Oregon loss at the first reveal. If they are hanging their hat on the losses hurting Nebraska in the end and using a conference call v in person as justification, perhaps the reading between the lines conclusion is that there are some influencers on the committee that don’t like Nebraska 🤣

My take:

1) The early reveal doesn't mean a lot - and you cannot necessarily bridge this to the final. This is a good thing - the Final should be independent.


2) History since regional seeds is too new to make definitive claims. The sample size is too small.


3) What the chairperson says as the reason for making a decision should be taken with a huge grain of salt.  When I think about how the discussion go, there easily could be no consensus of reasons. You have X number of members and each may have their own opinion and different reasons for choosing/voting for one team over another. A decision is finally made - but how the chairperson tells us the reason can be something not completely accurate w/o lying. There may be no way to generalize the deciding factor (out of conference scheduling, common opponents, injury, last 10 games) among so many different opinions. As such, I wouldn't put a lot of stock in those kinds of specifics from the chairperson - and it probably will have no bearing on future selections.

I agree completely with your first two points. On the third, I mostly agree. Every team will have many variables that the committee can consider when deciding how to separate them. If two teams are very close, they have to break the tie somehow. So in that particular case, it was probably some combination of Nebraska's loss to Northern Iowa, their relatively low (for regional hosting purposes) adjusted RPI of No. 6, and a bunch of other factors. I agree that a loss to a weaker opponent wouldn't necessarily disqualify a similar team in a future year, but I do think it mattered in that particular year. The reason I say this is because the chair of the selection committee has to know that they will be asked to justify some of their decisions on the selection show. So when the question about Nebraska comes up, they'll be prepared with some kind of answer. Pointing to the loss to Northern Illinois may have been the easiest way to justify the decision to viewers, as it's easier to understand than talking about the RPI, which tends to make people's eyes glaze over. To take an example from another sport, see how Ohio State has been left out of the college football playoff the last two year largely because of blowout losses to Purdue and Iowa.