All Hokie, All the Time. Period. Presented by First Bank & Trust Company

N
Joined: 01/01/2005
Posts: 2,340
Likes: 1,995
Subscriber

As a metric person, I would probably just say the RPI has shortcomings

and as opposed to calling it a joke (not a statistical term :-)). Massey is a good one and I am sure yours is too. As a statistician, you know that there is no perfect metric. Each metric has to decide what to include and how much to weight it. There is some subjectivity there. No way around it. You also know that there is a margin of error around any result. What's lousy is that there is a precise cutoff at 16. And our team keeps hovering around that. From what I have seen here, it may not have been the metrics/formulas that got us. People listed categories - wins against top 50, top 10, etc. as other things that were considered. The weighting of those is also subjective. The strength of the SEC conference clearly helps there and hard for us to compete with that in a slate of ACC opponents. That's likely the place where the SEC gets an advantage over every other conference. But, it's not that it's all unearned. They had the best OOC W/L pct of any conference. Pete said that he believed our schedule incorporated a number of strong OOC opponents. And that some of them did not turn out to be as strong as we hoped. But, please keep posting. There's always room for more data-driven commentary.
Posted: 05/11/2026 at 03:28 PM ET
(In response to this post by HokieEngineer22)

Thread Replies (17)

Full Thread →
Professional statistician and occasional poster here.
HokieEngineer22 05/11/26 11:09 AM ET
As a metric person, I would probably just say the RPI has shortcomings
nkhooch 05/11/26 03:28 PM ET
The statistical term re: the RPI would be “extremely underparameterized”
HokieEngineer22 05/11/26 04:15 PM ET
Come on now. The win/loss record "sounds" useful?
nkhooch 05/11/26 04:57 PM ET
That’s not quite how building robust rating systems work.
HokieEngineer22 05/11/26 05:56 PM ET
Best post I have seen on the subject. There is absolutely no logical
Hokie_Forester 05/11/26 02:04 PM ET
Well, how wast Angela Tincher not on the USA Olympic team, yet…
TomA 05/11/26 12:37 PM ET
Can you name the three pitchers that were selected?
MidloHokie 05/11/26 03:46 PM ET
As Ronald Coase said “If You Torture the Data Long Enough, It Will Confess”
Va AandM 05/11/26 12:32 PM ET
Yep, pick the teams you want and then select and weight the stats to
BEST2VT 05/11/26 12:58 PM ET
Ah, Finagle’s Rule: draw curve then plot your points. Works every time **
Va AandM 05/11/26 01:06 PM ET
It feels the same every time this happens. The committees
baumerj 05/11/26 11:35 AM ET
Only comeback is to win. Go Hokies. **
skautdog 05/11/26 11:18 AM ET
You're "RPI is a joke" line, although probably true, was the death ....
hokietony 05/11/26 11:14 AM ET
Then they have contradicted themselves because they didn’t follow RPI…
HokieEngineer22 05/11/26 11:20 AM ET
Committee contradicts self. Story at 11! 😉**
HokieNL 05/11/26 04:27 PM ET
In general they did, only 1 non-top 16 RPI team is a host.**
hokietony 05/11/26 11:22 AM ET
That exception is sort of what I am trying to point out.
HokieEngineer22 05/11/26 11:32 AM ET