The Way Too Early KenPom Report

RJKarl said:

Pilch, I'm guessing KU has many more than merely 600% more alumni than Wake, not to mention an entire state.

Click to expand...

/Odom

Interesting, but not surprising, that the "football schools" are all toward the bottom of the list.

Pilchard said:

When you subscribe, you can identify which team you follow most closely.

There is a page on the KP website which ranks the number of followers for each school from #1 (Kansas) to a bunch of schools tied for last. WF is #32 (which is pretty good, but we have more than 6 times fewer followers than Kansas)

Here's the popularity rank for all ACC schools on the KP website:

2. UVA (surprised)
3. Duke
4. UNC
13. L'ville
16. Cuse
20. ND
31. State
32. WF
38. Pitt
42. FSU
47. VT
66. Clemson and GT
72. BC and Miami

Click to expand...

Wake now up to #30. The defense took a large leap to #120.

Here's a little historical look at Wake's kenpom rankings, updated earlier when we were ranked #30, down one spot to #31 now.



Currently, this team's rankings are very similar to the 2002 rankings (when Wake was a 7-seed and lost in the second round to Oregon). This got me thinking about how good this offense is and which Wake teams of the past were similar to how this team plays now. I know these charts are a little large, sorry about that, didn't know how readable they'd be if I shrunk them, but here is a scaled cluster map comparing Wake offense since 2002.



In case it isn't clear, each column is scaled with 1 being the highest, 0 the lowest, and I just kept the kenpom category names. So Dino's teams were the lowest in Assist% and Bz teams wouldn't offensive rebound. And this is compared to Wake's history, not the rest of the country, like we aren't a bad offensive rebounding team this year compared to the nation, but compared to Wake's past this year is on the lower end.

Interestingly, the previous three coaches all had their own distinct cluster. Looks like Dino was the most distinct of our coaches, not necessarily bad as the first chart shows that Bz years were terrible and Dino had reasonable offensive ratings, but he certainly had his own style (no shooting threes, decent offensive rebounding, no passing to teammates).

Manning's first year resembled the last year of Bz, his second year was close to 2007 Prosser (his worst year), and this year has so far stood alone, although closer to Prosser than anyone else. Looking at Manning's numbers though, it appears he's cultivating a higher assist rate than previous Wake teams, along with a lot more 3 point shooting, lower turnover rates, and less offensive rebounding. So it seems like he may be developing a new identity for Wake, but this year the shooting percentages are finally justifying it. Personally, I enjoy watching this style of play, especially when we get to watch a dominant post player to go along with the much improved 3 point shooting. If there are any questions or someone notices something that doesn't make sense just let me know, been doing random Wake data visualization stuff for myself but figured I'd share in case anyone else would find it interesting.

TLDR: Wake's offense is fun to watch this year and it looks like Manning may be developing his own offensive philosophy.

rspDeac said:

Here's a little historical look at Wake's kenpom rankings, updated earlier when we were ranked #30, down one spot to #31 now.

Currently, this team's rankings are very similar to the 2002 rankings (when Wake was a 7-seed and lost in the second round to Oregon). This got me thinking about how good this offense is and which Wake teams of the past were similar to how this team plays now. I know these charts are a little large, sorry about that, didn't know how readable they'd be if I shrunk them, but here is a scaled cluster map comparing Wake offense since 2002.

In case it isn't clear, each column is scaled with 1 being the highest, 0 the lowest, and I just kept the kenpom category names. So Dino's teams were the lowest in Assist% and Bz teams wouldn't offensive rebound. And this is compared to Wake's history, not the rest of the country, like we aren't a bad offensive rebounding team this year compared to the nation, but compared to Wake's past this year is on the lower end.

Interestingly, the previous three coaches all had their own distinct cluster. Looks like Dino was the most distinct of our coaches, not necessarily bad as the first chart shows that Bz years were terrible and Dino had reasonable offensive ratings, but he certainly had his own style (no shooting threes, decent offensive rebounding, no passing to teammates).

Manning's first year resembled the last year of Bz, his second year was close to 2007 Prosser (his worst year), and this year has so far stood alone, although closer to Prosser than anyone else. Looking at Manning's numbers though, it appears he's cultivating a higher assist rate than previous Wake teams, along with a lot more 3 point shooting, lower turnover rates, and less offensive rebounding. So it seems like he may be developing a new identity for Wake, but this year the shooting percentages are finally justifying it. Personally, I enjoy watching this style of play, especially when we get to watch a dominant post player to go along with the much improved 3 point shooting. If there are any questions or someone notices something that doesn't make sense just let me know, been doing random Wake data visualization stuff for myself but figured I'd share in case anyone else would find it interesting.

TLDR: Wake's offense is fun to watch this year and it looks like Manning may be developing his own offensive philosophy.

Click to expand...

Very interesting data. Thanks for sharing.

Manning has a pretty remarkable coaching profile from a stats perspective because he doesn't really have any consistent trends over a 5 year sample size, other than playing fast and his team shooting a high volume of FTs (good). There is really nothing much that his teams do consistently well or consistently poorly, good or bad.

I will say that eFG% is one of the stats that varies the most for a coach based on personnel. That is basically our strong point this year but Manning has never been any good at having efficient teams before. But that's pretty normal, I picked a coach at random (Mark Turgeon @ Maryland) and his eFG% of his tenure have been 233, 60, 192, 82, 13, 92. And that looks pretty normal. Except for a few programs that shoot a really good %'s consistently, efficiency seems to bob around based on personnel quite a lot.

FckVwls said:

Manning has a pretty remarkable coaching profile from a stats perspective because he doesn't really have any consistent trends over a 5 year sample size, other than playing fast and his team shooting a high volume of FTs (good). There is really nothing much that his teams do consistently well or consistently poorly, good or bad.

Click to expand...

If nothing else, his teams are fun to watch pushing the ball up as fast as they do.

And this teams' FT shooting % is on pace to set a WF all-time record.

Decided to look into kenpom's accuracy for Wake this season, to see how well he is predicting our games as I know one of the common points is he's only miscalled one game's result to date. To start, I haven't been tracking the projected scores all season, so I don't have that data off hand but if anyone does let me know and I can validate what I have (and update any inconsistencies). Without having his projections for the year I am relying on his "Initial State" percentage located on the win probability graphs to get his expected game margin. I went through a bunch of teams and gathered their win% and margin for the remaining games this season to fill in the projected margin for our games this season. These are the projections for Wake by kenpom, vs the actual game margin. Games in the top left or bottom right quandrants are missed in terms of win-loss, so just the one. Here's to hoping we can fill in that bottom right quandrant with a couple upcoming games to make our case for the selection committee. Overall, the average kenpom error for Wake this season is 6.8 points.



The 5 biggest misses this year are:
Texas-El Paso 12 points off (exp10:act22)
College of Charleston 18 points off (exp-1:act17)
Louisiana State 25 points off (exp9:act34)
Miami (FL) 15 points off (exp2:act17)
North Carolina State 20 points off (exp10:act30)

I've also been looking at what about other teams most strongly correlate to our margins, besides the whole we lose to good teams and beat bad teams thing, but I wanted to start with this to figure out how accurate kenpom has been with regards to Wake this year.

Interesting that the five biggest misses have been Wake winning in a blowout.

2011 and 2012 [Redacted] are shockingly bad. Manning's first two years are comparable to [Redacted]'s last two years. The jump that we are experiencing now puts Danny on par with Dino's regime. It will be interesting to see what happens next season. If Collins stays, we should be an NCAA team easily; however, if Collins leaves, I think we will really see how good of a coach Manning really is.

We have struggled heavily with teams that don't turn the ball over much relative to the rest of the country (low TO%) and generally been very successful against teams who do turn the ball over a decent amount. Makes sense given how poor we are as a team at forcing turnovers. Teams that don't turn it over much in general turn it over less against us which makes it difficult for our defense to get stops since we're not forcing many turnovers at all.

RaleighTriad said:

2011 and 2012 [Redacted] are shockingly bad. Manning's first two years are comparable to [Redacted]'s last two years. The jump that we are experiencing now puts Danny on par with Dino's regime. It will be interesting to see what happens next season. If Collins stays, we should be an NCAA team easily; however, if Collins leaves, I think we will really see how good of a coach Manning really is.

Click to expand...

Judging by Kenpom this season would be in the high end of Dino's regime.

I think we remain a top 40 team with or without Collins next year. Top 20 if Collins stays.

Wakeforest22890 said:

We have struggled heavily with teams that don't turn the ball over much relative to the rest of the country (low TO%) and generally been very successful against teams who do turn the ball over a decent amount. Makes sense given how poor we are as a team at forcing turnovers. Teams that don't turn it over much in general turn it over less against us which makes it difficult for our defense to get stops since we're not forcing many turnovers at all.

Click to expand...

I was going to hold off on charts for the rest of the day so as not to inundate this thread with them, but seeing as you hit exactly one of the charts I had, I'll throw it out now. I took our margin of victory and adjusted for location (3.5 points) and correlated that to kenpoms four factors. Here is a table for ACC play:



Offensive and defensive TO% are the most closely tied to our adjusted margin of victory (uh oh Louisville). I kept the correlation matrix to ACC play to see how we were doing lately, on the year these two categories are still the most important, and most categories stay in the same order of importance over the whole year. Top line shows the whole year bottom row ACC play:



I then created a linear model based on the opponents four factors over the course of the year (wanted to use more data than ACC play to create the model, did not include tempo despite it being in the correlation table). This is still based on limited data, but still thought it would be fun to see how it relates to kenpoms projections. Has us losing to Duke by 7, beating Pitt by 17, losing to Louisville by 5, and beating VT by 5. So more bullish on Pitt and VT for sure. Neutral court Louisville is 5 point tougher draw than Duke by this model, kenpom has them closer to 1 point. Yes I know this will get blown apart tomorrow when we get crushed by Duke, but thought it would be fun to take a whirl at anyhow.

Wow that's great stuff.