fanfromthevalley
Well-known member
- Joined
- Jun 23, 2017
- Messages
- 493
- Reaction score
- 368
- Points
- 63
Was channel surfing and found this.....can't root for either coach in this one! Baylor up 34-12 in the second quarter at Baylor.
(Written just before our Rutgers game, so Rutgers results not included)
The silver lining for Marlene's hapless Texas Tech is that their loss to (now 2nd ranked in RPI) Baylor rockets them up about 18 slots in RPI rank, just due to the huge bump in their SoS and in spite of the loss. This puts them at RPI rank #141, 3 slots behind Minnesota at #138. Raw RPIs are .5175, .5230 and .7062 for TT, MN and Baylor, respectively.
Pairs of (raw SoS, SoS rank) for (respectively) TT, MN and Baylor are (.5535, #69), (.4938, #177) and (.6192, #7).
So we see that at this point in the season, MN and TT get almost identical (mid-range) RPI rankings, in spite of MN having a 14-7 won/loss record versus TT's 10-11 record. This is almost entirely due to the fact that Marlene inherited a nice challenging out-of-league schedule from the prior TT coach; whereas Marlene gifted Whalen a very cream-puff schedule, minus two decent teams.
The good news is that (win or lose) the Gophers will get a nice SoS bump (and associated RPI bump) just from playing #14 RPI Rutgers.
To me it shows that TT non-conference schedule was a joke. I looked at it on paper and they only played one power 5 school, Florida, on road and lost. I didn't do dive deep into their schedule, but Houston Baptist, Texas A&M-Corpus Christi, Louisiana Monroe and others are near the bottom of their leagues. They only played two road non-conference games. My guess is their non-conference sos was really low and their RPI before Big XII play was not very good.
You are right in the sense that I guessed wrong on the reason for TT's SoS (.5535, #69) being significantly higher than the Gophers (.4938, #177). TT did also have a cream puff early schedule. And they lost 2 of those games. And they lost all their Big 12 games except one. Whereas the Gophers won all their early games (including higher ranked Syracuse) and (when this was written) had won 3 Big Ten games.
See http://www.realtimerpi.com/rpi_153_Women.html versus http://www.realtimerpi.com/rpi_190_Women.html (which may be updated with the Rutgers win by the time most read this).
So what gives? Gophers results are better than TT results in all aspects, yet before the Rutgers game, both received almost the same RPI rating (and obviously due to TT having an SoS rank of #69 versus Gophers #177), while (as just pointed out) both had early season cream puff schedules, TT perhaps more cream puff than Gophers.
The surprising answer is simply that TT played #2 RPI Baylor twice in its schedule (losing both times, although the earlier loss was not such a blowout). They lucked out in their Big 12 schedule in getting randomly assigned to play the second ranked team twice during their league schedule. The power of playing #2 Baylor twice gives them a huge RPI boost (enough to make them appear the equal to the Gophers).
There is no #2 team in the Big Ten for the Gophers to play (not even once, yet alone twice). [Although we will get a single but lesser bump when we play Maryland. And when updated later today, the Gophers will get a double RPI bump, one from playing Rutgers and one from beating Rutgers.]
The point of my post remains the same. RPI is horribly inept as a measure of the quality of a basketball team. It gives TT a huge RPI boost for getting their butts kicked by #2 Baylor. Twice.
Sadly, the new NET ranking system for the men's game is even worse, and is statistically mal-designed. Such that the Gopher men's team is currently getting unfairly under-rated. Pitino is quoted in a recent Strib article, saying the Gopher MBB NET rating simply makes no sense to him. He is absolutely correct. NET makes even less sense than RPI, but for different reasons that are even more technical than "TT lucks out and not only has the #2 team in their league, but the lucky bums get to be beat by them twice, so that they wrongly get an RPI at the Gopher level when the Gophers are clearly a better team."
Agreed that the metrics seem to really struggle with teams in the middle. It’s obvious who the best and worst are, but the teams on the bubble seem to be the hardest to quantify and differentiate. Missing the tournament can cost a coach their job and the committees have a responsibility to do their best. Hopefully they use many metrics.
Sent from my iPhone using Tapatalk