So exactly where does this team stand in Self's tenure of teams?

I am so so so excited. After years and untold hours, I am finally able to run my DPPI for prior years. Well as far back as 1949 at least. But that is so very cool. Getting the data has been an incredible undertaking. I know you are looking at the screen and thinking "this dude seriously needs to get a life". But a hobby is a hobby and this is one of mine. I have been a programming maniac the last 3 months and it is paying off. While I am not so arrogant or deluded to think that my algorithm is some great breakthrough of modern science, I would bet it would more than hold it's own among the various predictive algorithms out there. I do not have the resources of KenPom to get detailed box scores for every game and have multiple programmers on staff, but for the limited data set my budget of time and treasure can afford, I will pit my DPPI against anyone out there, Sagarin, KenPom, NCAA Net or anyone. Okay, enough self adulation.

I have just completed the first round of running the DPPI back through 2004. In case you are interested, that is 18 year (correlates to Self's tenure at Kansas) and exactly 6200 teams on the nose. There were 326 Division 1 teams in 2004. This year there are 357 Div 1 teams but only 347 are playing. The purpose of this post is really to do a comparison of Self's KU teams over the years. But first, a few fun facts and a few notes of explanation. The DPPI is quite complex and has all sorts of intermediate ratings that work to the overall rating. Each intermediate rating sort of stands on it's own and has a purpose of it's own. Here is a brief explanation of the various ratings we will be looking at.

1. Adjusted Offensive Efficiency Rating -- This is essentially the offensive PPP (points per possession) adjusted to the strength of the opponent so all teams can be compared to a set standard. Averages to 1 (or very very close) across all teams.
2. Adjusted defensive efficiency -- This is essentially the defensive PPP (points per possession) adjusted to the strength of the opponent so all teams can be compared to a set standard. Averages to 1 (or very very close) across all teams.
3. Power Rating -- This is a pretty important one. This is the key number used to predict the scores between teams. It is the difference between the adj. off. eff. and the adj. def. eff.
4. Performance Rating -- this is how well a team will be projected to perform against a standardized schedule. This includes the difference between the power ratings of the two teams but also includes the consistency ratings and a few other things that are too complicated to discuss here. Suffice it to say that consistent teams and improving teams will generally rank higher on performance than they do on power. This is where the true predictive aspect of the algorithm comes in and deviates from a more retrodictive algorithm.
5. SOS Rating -- Pretty self-explanatory. A standardized measure of the teams actual schedule of opponents.
6. Total Rating -- this is a magic potion algorithm being my attempt to blend the Power Rating and the Performance rating into a single rating because most people just want a single rating. Where as the Performance rating is pure predictive, the power rating is slightly more retrodictive, the total rating is sort of the all things to all people rating that gives my best attempt to compare to KenPom, Sagarin and others.

Now for a few fun facts.

1. As I mentioned there are 6200 teams in the 18 years.
2. The top team for all 18 years is the 2021 version of Gonzaga in the total rating and performance rating but it is the 2020 Kansas team in Power Rating. Yep this season's Gonzaga is pretty special. But then so was last season's KU team. Those are clearly the top 2 teams out of 6200 teams in the last 18 seasons.
3. The 2008 KU team is 5th overall of all 6200 teams.
4. The 2012 season had the highest overall average at 69.37. The 2018 was lowest at 67.18.
5. The team with the highest 18 year average rating is KANSAS (yea) at 81.09. Duke is 2nd at 81.06, Gonzaga 3rd at 80.95, UNC 4th at 79.77 and Kentucky 5th at 79.26. This is huge, KU really is the gold standard under Coach Self. Soon I will have the data back to 1989 and we can compare Self to Roy.
6. Kansas is 4th in composite 18 year offense but only barely. UNC 1.119, Duke 1.118, Gonzaga 1.117 and KU 1.113. Then it drops off pretty dramatically to Kentucky at 1.089.
6. Surprisingly, Kansas is not 1st in defense. That would be Wisconsin at 0.895. Kansas is, however, 2nd at 0.898. Roy never was a great defensive coach as UNC is 14th.
7. KU, Duke and UNC are all tied for toughest SOS over the 18 years. Then it drops off pretty hard to Kentucky in 4th at 0.566. I do not have it set up at this moment to separate out regular season from NCAA tourney. It you did regular season, KU would undoubtedly be No. 1.
8. I'm not sure if this means much, but the ACC is the toughest conference over the 18 years at 75.29, but the Big 12 is 2nd at 74.91. The Big 10 is 3rd at 74.34.
9. The Big 12 has played the toughest schedule in composite.
10. Kansas has won the most games (518) with Gonzaga 2nd at 516. Duke is 3rd and way back at 505.
11. KU's lowest rank in 18 years is 23rd (2019) although we are flirting with doing worse this season (currently 21st). This is what is amazing though and shows the absolutely incredible consistency of greatness that Bill Self has brought to KU. The team with the 2nd best worst 18 year rating is Gonzaga (under 2 coaches) at 42nd. Other notables are Duke (3rd at 58th), Kentucky (6th at 80), UNC (8th at 89), Michigan St. (11th at 96 and Villanova (14th at 114). This may be the most impressive stat of all. This is mind-bogglingly good. Self has been simply amazing at his ability to bring a decent team year in and year out. Far better than any other coach. Amazing.

Okay, so lets look at how this season's team compares to past KU teams. First, let's look at roster. There is no question, none whatsoever, that this is the least talented team (at least in terms of talent rating coming out of HS) that Self has had. Here is a chart I devised of "roster score" based on the final composite HS rating and the experience level of the players on the roster. This team in every measure is the least talented by KU standards under Coach Self and it is not even close. So it is actually pretty impressive that KU is doing as well as they are given the limited (but not inconsequential) talent).



So lets look at where this team compares in all the ratings. I think it is safe to say that the 2021 version of the Jayhawks is not the worst Jayhawk team under Self. It does rank last in 1 category. It is the worst in adjusted offensive efficiency. I think after watching them this season we all believe that. Again, and I think unsurprisingly, this team is 10th out of 18 Bill Self Jayhawk teams in Adjusted defensive efficiency. Not bad actually. Last season was the gold standard for defensive and offensive efficiency. Looking at last years team numbers, it rekindles the deep regret that we did not get a chance to put that to the test in the NCAA tourney. We really had a great chance to win a National Championship. The 2008 had better raw numbers on offense than the 2020 team, but when adjusted for SOS, last season's team takes the cake. That was a most excellent team. The stinker of all Bill Self teams was my personal least favorite team, the 2019 team. That team lacked chemistry and seemed to be unable to get together. And it showed. That season was full of disruptions and pretty much not much fun.




I hope you all enjoyed this. It is a labor of love that has been many years in the making. Getting my hands on this data has been a great and wonderful thing,

Looks very impressive and just off the top of my head seems pretty accurate. I agree that last years team was going to be hard to beat in the end.

Thank you much for this. This is outstanding work and a great read.
Please pin this to the top section and update as you are able.
Thanks again

Outstanding- Thank you.