"My rule was I wouldn't recruit a kid if he had grass in front of his house.
That's not my world. My world was a cracked sidewalk." —Al McGuire

Wednesday, February 26, 2020

We're in Lenten Season, Patience is the Order of the Day.


Don't panic, this blog post shouldn't be as long as the previous one. It'll be ramble-ly to be sure but I think we can get to the point more rapidly. First, the feedback and commentary from the first post was tremendous and I truly appreciate the support because this isn't really my thing. Second, I'd like to highlight a tweet I got from someone on Twitter.

A bit of a side commentary(oh boy, I've really got a case of the rambles), Rubie Q is my favorite Wojo hater on Twitter. He is thoughtful, informed, very snarky, and not a dick. The lesson, push back all you want but don't be a dick.

Any whosie, this tweet hit home because I had this concern when I was doing my analysis. I was leaving out any analysis of whether other not Wojo-led teams are competitive in the general sense not just in the close game sense. Rubie is spot on, one of the "yeah, buts" of the Wojo era has been that he gets blown out in games and I didn't address it. So now I've got to go back to the ole' hypothesis drawing board. The good news, I think the work previously done is still valid but I have an additional hypothesis to test. So what is that hypothesis?

Hypothesis: Wojo-led teams have more non-competitive games lost than the average coach.

OK, big breath.....how the hell do I figure out THAT data set?? Paint Touches rode in with a suggestion (and a very good one) to use the T-Rank website (wonderful website I suggest you visit, it's like KenPom with a two drink minimum and free) which has a +/- metric which basically breaks down as the average margin for each game that the team had. So a -10 in the +/- metric means the team in question was losing on average by 10 through out the game, and when you combine that with a win in that game it would tell you that was a furious come-back win versus something like a +2 result with a win which might tell you it was a dog fight through out. The only flaw in this approach is that the T-Rank data does not cover the full extent of my control group because it doesn't go back that far, plus it would enable/force a different classification system than I originally started out with. I think this approach has a lot of merit going forward but seems like it's an analysis with a slightly different hypothesis to test(definitely something I'll consider as a future article if they don't take the keys to this website away from me). So back to the drawing board, I guess. Now, it's important to remember I'm lazy so I thought about the data I already had and it struck me, if I had every game for the teams I am analyzing I already have the hard analysis done....I know what games were competitive! Ergo, by definition any games that weren't in the list were non-competitive (ie a blow out win or loss). To that end, with a little more information collection I was able to grab the full season results for the Wojo-seasons plus the control group and classify each game in one of four categories: Non-competitive loss, Competitive loss, Competitive win, and Non-competitive win. So with that, we arrive at our first visual.

Lot of numbers, rows and lines....not much to take away from this view alone though it is interesting to note that the control group average for non-competitive losses a season is 3.5. Considering the control group has some very well regarded coaches and a couple of national championship winning teams it seemed a little high to me. So let's look at that information with with some context, I want my loses to be low in number(especially non-competitive) and my wins to be high in number(especially non-competitive) so let's apply some conditional formatting and voila:

As we already knew from the previous article I bored you with, Wojo has been getting steadily better in competitive games but it also appears that he is getting better about not having non-competitive losses. He is certainly not at an elite level in that regard (knock me over with a feather) but he isn't the garbage results he was in the first couple of years at Marquette. We should also acknowledge that he is not good at blowing other teams out so Wojo is generally going to win but not that convincingly. This years team has at least 6 games left to play, I'd expect MU to win the majority left but to do so in a non-blow out fashion, that's a good thing (maybe we aren't collapsing, who knows). However, I wanted to apply another bit of analysis to this....it's easy to not get blown out when you are playing "easy" teams, so let's apply the KenPom rankings to this

The analysis is simple here, Wojo's Achilles heal is that if he doesn't beat an elite team he gets blown out by them....but that's true of most of the coaches. Put another way, it seems like Wojo is probably a good coach(or above average considering the control group) but hasn't yet figured out how to beat the elite teams consistently and/or be really competitive in those games. However.....here's where I use that pesky trend term again.....Wojo's results have been trending upwards with each season. Wojo is getting more competitive against sterner competition as he has gone on. He's definitely not where we want him to be, but if the trend holds, he'll get there in the next couple of seasons. OK, but wait there's more! I then took all this data I now have and wanted to look at it in a weighted quad fashion (which does a good job of reflecting game location and team rank) to see if there was anything that stood out there. Long story short, each category type got a rating of 1 through 4 (1 being non-competitive loss and 4 being non-competitive win) and I multiplied the number of games in each category based on the weight then categorized them by the quad type and this is the visual you get.

Some housekeeping, the brown toned bars are the control group while the blue toned ones are the MU teams under Wojo. The red bars represent the average value of the control group metric in each quad. One really important thing to keep in mind when looking at this chart.....this years team has at least 6 more games to play this season and they are all going to be Quad 1 or 2 games so the darkest blue bar in the Quad 1 and 2 graph is going to get higher no matter what. Two major take-aways by looking at the data this way: Wojo's team this year is performing about where the non-national championship coaches of the control group performed in their seasons and that pesky trend is aiming positively again.

The conclusion from all of this is, once again, I think Wojo is victim of some narrative forming early in his career at MU that he simply hasn't been able to eliminate and as a consequence the "bad losses" result in more of the "here we go again" response as opposed to being able to reflect on those types of games becoming fewer. In other words, it feels like Wojo is performing as bad as ever, but he's actually getting better. I don't think he is where we want him to be, but this review gives me hope that he can get to where we want him to be. Will he? Very tough to say, but I don't have any evidence that he can't at this point which I will take and run with.

One more piece of context I want to provide as part of this analysis is a bit of commentary on expectations. In a very lucky happenstance, Alan Bykowski (he's the one that does all the good work on Cracked Sidewalks and the lead acolyte in the Howard for NPOY campaign), collected data on all of the coaches that have been to a Final Four in the last 20 years (41 coaches in total). He has data on how long it took the coach to first achieve an NCAA bid, how many seasons it took to reach both a Final Four and(if relevant) a National Championship, as well as how many coaching jobs before they achieved those goals. Below is the list of 41 coaches and the cells colored green are those coaches that have achieve one of the goals faster than Wojo has to date, ie. if a coach got to a Final Four in 5 years or less the cell is green because Wojo definitely hasn't done that (otherwise I wouldn't have to write an article like this)

Puts it in a little perspective, for me any ways. Yes, we are leaving out that Wojo hasn't won an NCAA game nor reached something more achievable like a Sweet Sixteen berth(which are totally reasonable expectations) but there are also a lot of coaches that went on to make a Final Four who didn't even appear in an NCAA game in their third season. Further, the average length in role to make a Final Four is just over 11 years! I'd also like to look at this list in one more way, what does the list look like if we apply these filters: coaches who made an NCAA bid in 3 years or less, made a Final Four, and did all of that in their first head coaching gig:

That's it, that's the list. Three of the five guys on the list (of the last 20 years) went to the Final Four on the backs of other coaches players and never found success anywhere else again. A fourth is struggling in his second gig (#DoneDeal) and the fifth is Brad Stevens who left for the NBA where he is clearly proving to be a very, very good coach. I don't offer this up as some sort of full-throated rebuttle to the concerns over Wojo....it is definitely concerning that he hasn't won an NCAA tournament game and that he hasn't yet found a way to beat elite teams. We also might be in the midst of another late season collapse which would be a huge red flag. What I do offer this up as is food for thought....do we need to be reactionary here or do we play the long game and see where it plays out? I think the expectation should be to win an NCAA tournament this year and earn a Sweet Sixteen berth within three years, and if he doesn't meet those expectations we gotta think about moving on. In the interim, the data tells me #mubb has a coach who is improving and showing little to no signs of having hit his ceiling yet. That's good enough for me anyway.

No comments: