5 years of data

[ctt8410]

To satisfy my own curiosity, I went back and looked at game-by-game KenPom ranks for all 5 years of Haith's tenure at Tulsa so far. Ideally, we'd use the NCAA's new NET metric, but BartTorvik.com has the correlation coefficient as 0.99 between KenPom and NET, so it's a decent proxy.

Cut to the chase, it's not a pretty story. 5 straight years with a final rank worse than preseason rank. 3 straight years without a sniff of the bubble.

 

[ctt8410]

Stating the obvious, but the last 3 years are almost entirely his roster.

 

[ctt8410]

This one's trickier, but I think it's instructive. The NCAA published D1 basketball program expenses in 2015 when Haith was hired. Tulsa ranked #70 on the list. 63 of the top 100 programs on that list kept a coach for at least 4 years in the timeframe since then. Amongst that list, Haith ranks 56th in median KenPom rank vs program expenses rank.

 

[hotterthanyourhusbandTU]

Great data.

If we rewind 5 years, many of us were worried that Haith would have the same results here that he had in his career.

Well...he did. Unreal.

Remove the Mizzou season with Anderson’s players, and most seasons (with his players) are carbon copies.

 

[Original TULSAFAN]

Guy is a bad recruiter
He actually stinks at the most essential duty of the job

 

[aTUfan_rivals]

So for now, our outlook is for another mediocre year.

Yippie!!!!

 

[rusty-c]

...

 

[astonmartin708_rivals]

When Haith first came to TU, I really hoped that his strength would be recruiting since we were in a shiny new conference with bigger name schools. I thought he might have something to really sell kids on. Thus far, he's lagged behind what we really need in terms of talent, while other conference mates have excelled.

No excuses anymore. The buck stops with Haith.

 

[chito_and_leon]

To satisfy my own curiosity, I went back and looked at game-by-game KenPom ranks for all 5 years of Haith's tenure at Tulsa so far. Ideally, we'd use the NCAA's new NET metric, but BartTorvik.com has the correlation coefficient as 0.99 between KenPom and NET, so it's a decent proxy.

Cut to the chase, it's not a pretty story. 5 straight years with a final rank worse than preseason rank. 3 straight years without a sniff of the bubble.

Really interesting. What’s the preseason number based on? Is that mostly expected strength of schedule? I wonder what it means to get worse over the year? We played worse than expected or our opponents did or both?

If you’re not a great recruiter, you need to be a 2+2=7 kind of coach, where you take the pieces and make them into more than the sum of the parts. And if you’re both, you’re a Bill Self. Frank seems to be neither - he’s an ok coach and ok recruiter. Which explains why he’s won 60% of his games over 15 years. Hard to think at this point that he’s got magic somewhere to move that to 80%.

 

[chito_and_leon]

Over franks career he’s won 51% of conference games. Learn to love 9-9...

But if you exclude his first 2 years at each school, so only look at when he has his own guys, he’s won 43% of conference games. So learn to love 8-10.

If you could sum up Frank’s career in one season, this year would be it.

 

[ctt8410]

Really interesting. What’s the preseason number based on? Is that mostly expected strength of schedule? I wonder what it means to get worse over the year? We played worse than expected or our opponents did or both?

The preseason number is based on recent program history, number and quality of returning players, and high-profile additions (I think it's top 50 recruits and D1 transfers). It's a bit of a black box, but they've performed extremely well in the years since Ken has been posting them.

For example, Haith's first team returned 70% of their minutes and the top 4 players from a team that finished the previous year ranked 82nd. Based on that information, the team was projected to be the 48th best team in the country. Based on actual results they played like the 85th best team in the country.

The next year, they returned 80% of their minutes and basically the entire roster. Based on that information, they were projected to be the 47th best team in the country. They finished 65th.

In my opinion, the within-season trends point to a team that didn't meet expectations based on their personnel. New players weren't as good as expected based on previous recruiting and they didn't improve as much as expected from year-to-year. When it happens every year for 5 years, that's pretty damning evidence against a coach imo.

The other thing to remember about KenPom is that it's designed to be purely predictive. One of the greatest benefits of predictive metrics is that they're independent of schedule because all of the context has been adjusted for. It's a difference of team quality vs. team resume.

 

[Babe the Blue Ox_rivals]

We were predicted to finish 9th this season but ended up in 7th. Can he claim that he exceeded expectations?

 

[TU_BLA]

When Haith first came to TU, I really hoped that his strength would be recruiting since we were in a shiny new conference with bigger name schools. I thought he might have something to really sell kids on. Thus far, he's lagged behind what we really need in terms of talent, while other conference mates have excelled.

No excuses anymore. The buck stops with Haith.

And he can’t use the excuse of facilities. The DonRey and the housing our basketball teams get are upper tier in the conference.

 

[TU4ever2]

And he can’t use the excuse of facilities. The DonRey and the housing our basketball teams get are upper tier in the conference.

I don't know about upper tier.

The new Moody, 5/3, Fretitta are pretty nice, as are the FedEx Forum (newer than Reynolds, NBA facility) and Kock which was recently redone. The Youengling center is pretty nice too.

The Reynolds Center is 20 years old.

However, our facilities for basketball are not a negative like they can be in football. They are above average for division 1 easily, we just happen to be in a conference that is pushing for top notch facilities. So he definitely can't use them as an excuse.