"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

Thursday, January 18, 2024

A Good Run of Bad Luck

At the risk of getting an “OK Boomer” from the 30 and under crowd I’m going to a semi-obscure mid-90s reference as the entry point to this blog post. 

The early 90s was the start of the “hey this IP worked 20-30 years ago we should rework it and cash in on those sweet, sweet nostalgia bucks” which is why Richard Donner (Superman, Goonies, Lethal Weapon series) ended up directing a movie called Maverick which was a remake of a late 50s black and white show of the same title. The movie stars Mel Gibson (peak his movie making but pre-his being notably insane), Jodie Foster, James Gardner, Graham Greene, Alfred Molina, James Colburn, and a Danny Glover cameo. It’s silly and not exactly Oscar bait, but it’s fun in a way only the 90s can be. The point of all this is that Maverick is a card shark and a con man who’s trying to scrape together enough money to get into a big poker tournament. Now before you walk away thinking this is a blog about “Phil ‘members things”, the point is that for the majority of the movie you don’t know whether Maverick has a plan, is he falling victim to bad luck, is the con man getting conned, etc. In fact, there’s a song for the movie by Clint Black that is where I got the title for the blog that’s all about what happens when your luck turns bad. And that, my friends, is what brings us to the meat of the article. We find ourselves at the point of the Marquette season where #mubb folks are wondering if the team has been figured out, maybe Shaka is a long con, teams are packing the paint, or we just really miss OMax. All reasonable takes worth examination, but I think the real culprit is….we’ve had a good run of bad luck.

By any measure that counts (wins and losses, shooting percentage), Marquette has struggled since early December. They are 6-4 in that stretch and are shooting 44% overall, 55% from 2 and 30% from 3. Compare that to last season when 95% of this roster shot 46/58/35 respectively. In the last month and a half it's safe to say the team has left points on the court compared to what they are capable of. The question is what is the cause of this? Rationally, we want there to be a tangible reason something isn't working because generally speaking people don't want to believe in things like ghosts (shouts to Lazar Hayward) or bad luck because it's not something you can do anything about. That means there HAS to be something more than Marquette has had bad shooting luck but I think that's exactly what we're looking at. The offensive process is good to great, it's just bad luck/outcomes that have cost us some games of late. Let's dig into the data that I think backs up the conclusion. 

The internet is a wonderful and scary place and nestled within it is a website called Synergy Sports that friend of the blog, Paint Touches, often references. The tl;dr explanation for Synergy is that it’s a website hosting basketball data that uses traditional box scores as well as analysis of game tape (human and automated) to develop advanced analytics giving you deeper insight into the game. One of the key elements that Synergy provides is the classification of shots taken: was that a mid-range or a runner, how many shots did my team take at the rim, long 3 or naw, and was that 3 guarded or unguarded? Why this feature of Synergy is key is that Marquette specifically has a philosophy around what kind of shots the team should be taking. Shaka Smart and Nevada Smith have spoken in the past about MU's offense having shots categorized into three tiers: Tier 1 is shots at the rim and unguarded 3s(take these whenever possible), Tier 2 is guarded 3s and runners(think Oso push shots), and Tier 3 is mid-range jumpers, hook shots, etc. It's not that MU can never take a Tier 3 but it is seriously frowned upon and if you take it I bet you are STRONGLY encouraged to make it. With that background in mind, let’s take a look at a couple of “what is wrong” hypothesis that have been floated

Teams are forcing Marquette to play slower

As the theory goes, Marquette is a fast-paced team and when teams limit possessions overall MU can’t generate enough offense to be successful. Here’s where we’re going to refer to data from Synergy first. I pulled the raw shooting classification numbers, including the total number of attempts for each game of the season:




Keep in mind this is shots taken so it excludes any FTs taken as a result of 7+ team fouls but we’ll use it as a metaphor for the offense as a whole. Marquette has averaged a shade over 62 shot attempts per game. Marquette’s record in games taking 63 shots or better:

        7-2 with the loss coming at Seton Hall and Butler

Record when taking less than 62 shots a game:

        5-3 with the losses being at Providence, at Wisconsin, and Purdue

OK, so nothing conclusive jumps out there. Marquette has lost with and without a lot of shot generation. Some other observations from this data:

1. Marquette generated a TON of shots against Butler (more on this later) 

2. Marquette hit a lull in shot generation in the @Wisconsin through @Providence games by taking less than 60 shots in each of those games which was the start of our ride on the struggle bus. However, Marquette generated it's fewest shots against both UCLA and St Thomas (52) so even if you argue that Marquette can be slowed down they can still be efficient and "Win Anyway". Additionally the volume of shots at the rim was largely unchanged.

3. The lull coincides with the first in a string of injuries for the team (Stevie Mitchell) so there may be more correlation between pace of play and player availability than there is other teams deliberately slowing the pace.

Packing the Paint

Now let's look at it as a percentage for each shot classification for the season (# of shots/# of total attempts):




What we’re looking for here is are there any significant changes in the amount (%) of shots being taken at the rim, significant move from guarded to unguarded 3s, and/or increase in number of runners taken. What jumps out to me here is that in the last 10 games there aren't hugely obvious variances from what we've seen all season in terms of what looks the offense is generating/taking. 12-19% of their shots are unguarded 3s and 40+% of their shots are at the rim. Two games do stand out: Providence had a season low in unguarded 3s by a long shot and Bucky gave us a large percentage of unguarded 3s and took away at the rim shots. More on Providence later, but that Bucky result stands out as an indicator that Gard said "hey shoot 3s with our cold gym and weird ball, we're not letting you go to the rim" and it paid off. But if teams as a whole are packing the paint in the last 10 games it doesn't seem to bear out in the shots that Marquette is generating. Plus some of the games where MU had the fewest at the rim shots (Texas, Kansas, UCLA, Creighton) Marquette is winning anyway.

If teams are packing the paint as a general rule, they aren’t very good at it.

“Figured Out”

 If we group shots based on tier and look at it for all of last season, the beginning of this season, and the last 10 games this is what that looks like:



First of all, holy cats did the offense know what the heck it was doing to start the season. Yes there's buy games in there but the schedule strength wasn't that appreciably different as the average opponent KenPom ranking in the first 7 was 93 and the last 10 was 73 (thanks Mick Cronin for dragging it down for everyone, you really are the worst). Second of all, yes from a shot selection standpoint the team has come back to earth somewhat but they are still at the same pace as last year for great shots and have converted a fair number of their bad shots into good ones (all without OMax I might add).

Let’s look at the game-by-game shot “quality”:






What stands out here is that while there is variation in Tier 1 shots (great) coming during loses and close wins, those shots are by and large being converted into Tier 2 shots (good). In fact, Texas was the game where our shot selection was the "worst" on the season whereas our shot selection was solid against Purdue, Butler, and Seton Hall and resulted in a loss. The point being, if teams have figured out Marquette's offense it’s not consistently figured out and really only at the margins in that we take a few less great shots and a few more just good shots. Look at those tier 3 numbers, even in the last 10 games the team has significantly improved on taking good shots instead of bad ones.


Wrong shooters shooting

The next argument is "well Phil, you silly, teams are sagging off bad shooters", ok well, I have a chart for that too:




This chart breaks down who is taking 3s regardless of "type". Prior to December Tyler, Kam, and Jopwagon(David) were taking 52.3% of the 3s but in the last 10 games they are taking 63.4%! And that's with Ben Gold more than doubling the 3s he’s taken this season in the last 10 games (33 out of 51 taken). So in the first 7 games Marquette was taking more tier 1 shots but they were generally coming from our less capable shooters and in the last 10 games when we've thought the team isn't playing well they are not only outstripping last seasons shot selection but are also significantly increasing the amount of 3s that our best shooters are taking....that should be an ideal outcome but it clearly isn't!

 All of this is really just a long throat clearing to say, I don't think the offense has been figured out as they are still taking the shots they want and with who they want.


OK Smart Guy, What Is The Problem


Let's get to the part where I Perry Mason a point that blows your mind......remember all those unguarded 3s and shots at the rim they are generating as part of tier 1 shots? Whadya think Marquette is shooting on those on a game by game basis?






Stare at that in all of it's HP Lovecraftian glory. On 2s, Marquette has not been shooting well during key games in the last 10 like Providence and Butler but generally as well. Last season they shot 55% from 2 and during this stretch, 4 out of the 10 games they shot under that mark (and went 1-3 in those games). They generated shots at the rim (3 out of 4 were at or above season average) but they simply did not make them at a high enough rate.

When it comes to unguarded 3s however, its not a case of Marquette not shooting well during the swoon, they just flat out can't shoot unguarded 3s except when they are playing Texas! For a little context, we go back to Synergy where they conveniently allow me to search team's shooting percentage based on unguarded shots. The median team (181 out of 362) in NCAA Division 1 basketball this season is shooting unguarded 3s at a 37.5% clip. Marquette hit or exceeded that rate just 4 times all season. Marquette shot 3s last year at a 35.3% clip and they hit or exceeded that rate just 5 times all season. Overall, Marquette is shooting a paltry 33% on unguarded 3s and that's held aloft by Shaka's blood feud against Texas, without that game it drops to 31%! I'm presuming you want to know where Marquette ranks this season out of 362 Division 1 teams since you've made it this far in my rambling.......


I put that in large font so it would really pop. Marquette has been inexplicably bad at shooting unguarded 3s this season! 3-point defense is often thought of as containing significant luck because so much can depend on the shot, rims, the ball, etc. Now the whole point of unguarded shots is that there is no defense; it's no different than putting shots up in the gym after practice, so a make/miss becomes almost entirely skill/luck. Even Steph Curry has some shooting variance and gets slumpy at times but 31% on unguarded 3s is just cray cray. There is no way teams have figured out how to guard us into ignominy on unguarded 3s, it’s just not a thing



To start to bring this to a close, let’s revisit a couple of the "more on that later” from earlier cause it is now very much later. First, that Butler game should go directly into the Basketball Infamy Hall of Fame. Given what Marquette did on both ends of the floor (season best 70% of tier 1 shots, most TOs on D, 24% of shots are unguarded 3s) they should have won that game by a million. But instead they shot 50% from 2 and 17% from 3…..like, sometimes basketball just sucks ya know. Second, the Providence game stands out as maybe a game where the opponent outclasses Marquette as opposed to it shooting itself in the foot. In that game Providence forced MU to take a season low amount of tier 1 shots and outside of Texas (LULZ) forced Marquette into the highest volume of bad shots. Some of that might have been the physicality and poor officiating (Win Anyway) but there is a non-zero chance Kim English has Shaka figured out which would have been an impossible sentence to imagine coming into the season. Hopefully it's non-repeatable with Bryce Hopkins out for the season, but it's something to keep an eye on.

 OK, I know this blog has taken forever to get here but let's get to the conclusion phase. I think we've walked through the process part of the Marquette experience, what they are trying to do, and validating the process is working. There’s a game or two where Marquette got away from it but it’s not wholesale or some sort of coaching blind spot being exposed. If the offense is doing what it wants and we’re getting the ball in the hands of the folks we want, the only thing left is luck. Just to show you the impact of unguarded 3s of which variance is almost exclusively luck, let’s look at every game this season where Marquette shot less than 35.3% on unguarded 3s (their overall 3pt% last season) and assume a world where they actually got to that rate. This is very conservative since the median team 3pt% on unguarded is 37.5%s. If we adjust the score by "giving" Marquette extra based only on the unguarded 3 shooting result, this is what the season would look like:








The big takeaway is Marquette would win the Butler and @Seton Hall games and get in spitting distance of Wisconsin (which would strengthen my argument that had Jeff Anderson not pulled his pants down on national TV for calling a foul on Jopwagon when the Badger he was guarding forgot how to walk for a couple of seconds, MU wins that game but I digress) and we'd all feel better with 14 wins now right? What if MU got to 40% on unguarded 3s (top 100 unguarded rate):





So we beat Bucky as well, cool.


Bottom line, if we are at all less unlucky our record is even better than it is now and the only losses where Marquette was clearly outplayed were Purdue, @Wisconsin and @Providence, I can definitely live with that. No guarantee our luck turns or we "positively regress" to the mean, but it also means we haven’t been figured out which would be way worse IMO. You know what else is unlucky? We've had significant injuries to 3 key players in the rotation. So yes, they've struggled some in the last 10 games but between shooting and injury luck is that really a surprise?

 I get it, as my father would say(who speaks 95% of the time in colloquialisms) "if wishes were horses, beggars would ride", you might say this doesn't matter because we can't change anything. That's of course correct, but also the whole point is that Marquette, no matter how frustrating or how much you might be changing expectations, is very much the result of misfortune as opposed to anything “wrong”. It’s a keep the faith sort of situation. Likely the team will see if the law of probability bends back in our favor. As Clint Black sings:

 "Til it's time for a windfall and not a single minute too soon
I've been too long overdue, now I'm gonna shoot the moon
I'd bet it all on a good run of bad luck"


We just gotta keep betting on that luck to turn.



No comments: