We’re joined by data scientist and creator of the rapidly growing college basketball analytics platform EvanMiya.com, Evan Miyakawa. Evan’s work has become a trusted resource for coaches at every level—particularly those looking to cut through the noise of early-season statistics and understand which metrics actually matter.
In this conversation, Evan shares how coaches can separate interesting numbers from useful ones, why certain teams have unique statistical fingerprints that predict success, and how lineup data can be leveraged far more effectively than most programs currently use it. We also unpack defensive priority stats (threes, rim, free throws), halftime box score interpretation, and the origins of Evan’s “kill shot” metric—his way of measuring and predicting momentum swings across a season.
Whether you’re analytics-comfortable or analytically curious, Evan offers a clear, practical roadmap for applying data without drowning in it.
What You’ll Learn
1. How to Separate Useful Analytics from Interesting Noise
Why the highest-value decisions come from combining coaching intuition with clean, well-adjusted data—not relying fully on either side. Evan explains how to weigh early-season sample sizes, refine priors, and avoid overreacting to statistical outliers.
2. Identifying the 2–3 Metrics That Predict Your Team’s Success
Teams often win and lose in repeatable ways. Evan shows how programs can isolate their own “keys to victory,” giving coaches clarity on what truly influences outcomes and what should drive practice and game-planning priorities.
3. Making Better Lineup Decisions (and Avoiding the Small-Sample Trap)
Learn why 30 possessions can reveal a spark worth exploring, why 100+ possessions create real confidence, and how properly adjusted lineup data helps identify combinations that elevate—or limit—your team.
4. How to Read a Halftime Box Score with Meaningful Purpose
A practical framework for assessing which stat pairings matter most and how to determine whether discrepancies at halftime signal a real issue or simply reflect pace and game context.
5. Prioritizing Defensive Outcomes: Threes, Rim, Free Throws
A breakdown of how these three defensive outcomes interact, why no team can eliminate all of them every night, and how understanding opponent tendencies can simplify what your defense should prioritize.
6. The “Kill Shot”: Understanding and Predicting Momentum Runs
Evan explains his 10–0 run metric—why preventing opponent runs is more predictive of elite performance than generating them—and why the duration of a run often tells the true story of a game’s momentum.
7. Coaching the Moments Around Runs
Actionable insights on timeout usage, recognizing schematic breakdowns, and managing emotional swings so you can stop a slide early or create momentum without burning resources unnecessarily.
Evan’s work reminds us that analytics aren’t meant to replace coaching instincts—they’re meant to sharpen them. By focusing on the metrics that truly matter, understanding lineup signals, and managing the moments that swing games, coaches can make faster, more confident decisions throughout the season. His insights offer a practical blueprint for turning numbers into competitive edges, one possession at a time.
Transcript
Evan Miyakawa 00:00
The best teams, the teams that end up winning titles or making final fours or being the most successful the most are often really good at preventing 10-0 runs. And really what you’re trying to do is, as you’re trying to assert control over a game, the fewest amount of things that can go out of control, the better.
And so oftentimes some of the best teams are the ones that they are very, very trustworthy with the lead. If you’re up six or if you’re up eight, you’re not really that concerned that a team is gonna quickly claw that back and kind of put you all of a sudden no longer in the driver’s seat.
Dan 01:56
And now, please enjoy our conversation with Evan Miyakawa.
Evan, big fans of all that you do for the game, your site. It’s a wealth of information that I definitely get lost in, in a good way from time to time. Excited to have you on today, so thanks for making the time.
Evan Miyakawa 02:25
Glad to be here, guys. I really appreciate what you do. It’s a really cool show, and I don’t traditionally come from the coaching background, but having something to contribute to this area through data and whatnot is really cool, so I’m glad to be a part of it.
Dan 02:36
Awesome, appreciate it.
Evan Miyakawa 02:36
Thank you,
Dan 02:36
Evan, our first thing that we wanted to dive in with you on is the general theme of what is interesting analytics versus useful analytics for coaches. And I think Pat and I were talking beforehand about, especially kind of early to mid-season right now where you’re starting to get some box scores, you’re starting to get analytics in out of, you know, say your first five to 10, 15 games, and you’re trying to optimize for the right targets for your team. And you see stats out there that are thrown out that are interesting, maybe for someone else’s team or interesting from just the overall perspective, but maybe not useful for your team. And so kind of broadly throwing it back to you on things that are useful versus interesting and how coaches can think about that.
Evan Miyakawa 03:19
My general philosophy when it comes to using data in general is that your most informed position, the position that you can be in to make the best decisions is when you combine your own expert knowledge of the field, if you’re a coach, that means your basketball IQ, your understanding of the game, your knowledge of your roster, that sort of thing with kind of these non-bias objective data-driven tools. And if you combine them together and kind of give them the right proper weight to basically put you in a better spot than you were before without it, that’s kind of where you’re in the best place.
Now, what that doesn’t mean is that you should just totally blindly trust the data, whatever the data is telling you. You kind of forget about any biases that you had, any preconceived notions or thoughts, but also if you just completely ignore the data because you don’t trust it and you don’t learn how to use it well, that’s not going to put you forward in a better spot. So my opinion is that you need to use both.
And so, you know, I’m the one coming from the data side and I am not someone who just looks at the data I have on my website and just blindly trusts it and doesn’t critique it or match it with my own kind of priors, essentially. So when you’re in the beginning of a season, for example, and you don’t have a wealth of information yet, but you’re trying to take advantage of some of the data that’s out there, it kind of for me becomes a target of, okay, what did I think was going to happen or what do I think is the truth? And then how do I start to slowly update or inform my opinion on that based on the data that’s coming in?
And a lot of it really is this battle of not overreacting to the data, but for someone who’s not a data person, how do you know what overreaction is, right? How do you know how much weight should I be putting on this data or is this a data point that I should even be learning anything from rather than just, oh, that’s interesting. So I think just general picture, a lot of my battle when it comes to the metrics that I put out is trying to make it as predictive as possible right away so that when a coach is looking at it or decision makers looking at it, they’re not having to do as much kind of legwork on their end to figure out how much can I learn from this? If anything, I want to try and kind of make the numbers already kind of leaning in that direction of saying like, Hey, this is what I’m actually predicting to do going forward.
Pat 05:31
Evan, looking at that predictive conversation, early season, how should coaches start to then maybe hone in on what stats are really going to be helpful to winning later? So what will carry weight or continue on versus like you said, it’s small sample size, you know, we just shot really well, or we played a team that couldn’t rebound.
So how should coaches start to hone in on the stats that are kind of more specific to their team, whether they thought so or not?
Evan Miyakawa 05:59
When it comes to kind of finding the key sort of overall team stats that are most indicative of your success as a team, this is actually something I’ve put a lot of effort into on my site. I think this sort of came from an idea of there was a school I was working closely with a couple of years ago and part of how they formed their identity as a team was they kind of figured out under their coaching staff over the last five to 10 years, they had kind of identified three common team stats that really spoke to whether they were succeeding or failing.
I’m making this up, but it was something like if we shoot above 36% from three in a game and we out rebound our opponent and we have a turnover rate under 12% or something, then we have found that if these three stats, if we actually look at our record of if we hit all three of these in a game, what’s our record? If we miss all three in a game, what’s our record? Or if we go two and one or one and two, you can look at how the swing from how good to how bad you are based on these specific metrics. You can look at your margin of victory overall. And what they identified was that basically, obviously it’s good to be every team wants to be good at three point shooting. Every team wants to have a low turnover rate. Every team wants to out rebound the other team, but they found for their specific program that it actually was even more important for them than for other teams. I do believe that there are certain stats in terms of just overall team stats you would look at at a box score, also stylistic stats like what’s your three point rate, what percentage of threes you’re taking versus twos, or how often are you getting to the line versus emphasizing other types of shots that wouldn’t get you there as much, or how much are you leaning on starters versus bench scoring, sort of these stylistic things.
I have found that for a lot of programs, there are certain one or two or three key metrics that if you look at over their most recent season or their most recent four or five seasons, especially if you have the same coach and system, more or less, that they can really kind of point to, hey, for this specific program, or this team or this coach, these are the main things we really need to most games be really trying to hit. And typically, it’s pretty lined up with how will they perform in games. And then similarly, I’ve seen people take that a step further and saying we want to apply this to the opponent. So if we can figure out these metrics for an opponent, we can then figure out how to shut them down. I have a whole set of tools on my site for college basketball teams called keys to victory that basically does a lot of the legwork for you in identifying these metrics or allowing you to punch them in on your own and seeing how they affect your team performance.
Dan 08:26
On your key metrics report and your keys to victory, you’ve got three-point percentage opponent turnovers, offensive rebounds. What else to you goes into the metrics that you work with teams on that are important?
Evan Miyakawa 08:38
I’ve tried to keep it relatively simple in the sense that I’ve not, at least so far, what I have on my site, I don’t drill down to super, super specific things like mid-range shots within the last 10 seconds of the shot clock or something, just because with a lot of these stats, if you get too specific, you can always kind of find some sort of correlation that kind of backs up what you’re hoping to look for. And so I’ve almost intentionally kept it a little bit more high level of just like when you’re looking at a box score after a game, what are the main things that you look at?
You’re typically looking at your two-point percentage, your three-point percentage, your turnovers, rebounds, assists, like all those main box score stats. I’ve mostly kept it to those things because those are well-established stats that we all understand like this is good and this is bad, but when you kind of take them over the totality of a single season or multiple seasons, you can actually start to identify trends between not just like a single metric, but how they play it in tandem. For example, I was looking at some stats recently, I was looking at Duke under John Shire, who’s been there at Duke for a little over three years. And the two main targets that you find kind of for their success or failure for their team over the last three years is if they shoot, it’s two-point percentage and then not fouling a lot. So if they shoot over 56% from two and if they have under 15 fouls a game, now you could change that to foul rate instead of number of fouls, but for the sake of an easier discussion, I just put it as number of fouls. If they shoot over 56% from two, that’s one target and have 15 or fewer fouls, they are 26-0 in the last three seasons, beating teams by an average of 25 points. If they get one out of two, they’re still winning 91% of the time, but if they miss two, that’s really where you see this massive drop in performance. They’re only winning 46% of the time. They have more losses than wins when they’re not shooting well inside the arc and they’re also fouling a lot. That’s a much bigger disparity between how good they are and how bad they are for those two metrics than for most teams. So that’s part of why it identifies those. If you’re a Duke fan, you can probably see like, yeah, I mean, there’s other metrics that matter, like our three-point shooting matters, that’ll be, there’s a lot more fluctuation game to game, but consistently, if we’re really good at the rim and if we’re not getting in foul trouble, we’re not putting the opponent to the line as much, we’re almost always winning those games. So I think there’s things like that for every program that you can typically find.
Dan 10:56
That’s an interesting point you bring up because I think as coaches, you’re hoping for consistency in your analytics. Like you’re trying to find the things that night after night for your team specifically are going to be fairly consistent.
And if that consistency is what’s needed to win or if you need to find other areas to be consistent in, the question is what to do or how to think about variance in your analytics. Like you just mentioned three point shooting. I mean, you can shoot the heck out of the ball one night and lose a game. You cannot shoot it well the next night and win the game. And then you look at your three point percentage and you like, what do you do with that? So I guess the question is the variance in these analytics and how coaches should think about their performance overall from that standpoint.
Evan Miyakawa 11:37
I think a lot of that variance is a word that’s just kind of thrown around that says how inconsistent is it from game to game. I think the question that’s maybe a little bit more purposeful is not variance but control.
How much control do you have over the result that you see from this metric? Because like you could have a team that shoots 45% from 3-1 game and then they shoot 15% the next game but the real question is how much control did they have over that? Were they able to get wide open three-point shots against one team and then they were having a really tough time getting good three-point shots against the next team because that team was really good defensively or their offensive scheme wasn’t working as well. In that case they had a lot of control over that because the whole style of the game was different versus if you’re taking more or less the same shots the same openness from game to game but one of them is 45% one of them is 15% you have relatively little control but you know your process was good right? A lot of it’s about the process whereas like if it gets well established then in general two-point percentage if a guy is supposed to make a 70% chance to lay up at the rim it’s gonna be pretty consistently 70% kind of with that same look no matter what. So a lot of it for me comes down to control and I think that we’re gonna talk about kind of the big conversation about a variance. Yeah three-point shooting is the most hotly talked about main stat that can often influence success playstyle things like that and we’ve seen an evolution of three-point shooting in terms of it being more emphasized at the NBA level coming down to the college level and I think sometimes people give too much stock to the whole three-point variance thing where basically like early in a season especially you’ll be looking at a team and saying like they just blindly look and say they’ve shot 45% from three that is not repeatable so you know we should knock that team down a peg in our minds. Well maybe 45% isn’t repeatable but maybe they’re still gonna be shooting 39% going forward and that’s still gonna be better than most other teams. So just because a three-point percentage is way better than average or way worse than average and it might regress to the mean a little bit doesn’t mean that that’s not still significant that that team is that good in that area whether it comes to their own ability to shoot threes or the ability to stop teams defensively and something I’ve actually done on my site which is kind of a whole other thing is on an individual player level I have projections for how a player is going to perform in all these individual statistical categories three-point shooting,
Evan Miyakawa 13:59
Three-point, two point shooting, assist rate things like that and something I’ve found in general is that three-point shooting if you look at a given season there will be a handful of players in in Division 1 out of say 5,000 there will typically be about 200 who shoot over 40% from three in a given season on a pretty high number of attempts but when you look at most of those players a lot of people will think like oh that player he’ll probably shoot you know if we got him in the transfer portal for example this guy shot 40% from three last year so lock him in he’s gonna shoot 40% this year most of the time almost 90 95% of those times not only do those players shoot worse from three the next year but they shoot under 40% the following year so if that player shot 41% last year and then it dropped down to 37% you might say like we didn’t get what we wanted out of him but the reality is that shooting high 30s consistently is like really really rare in college and so if you’re a career 37 38% shooter that’s still like an AA plus compared to the rest of Division 1 so even if you’re gonna drop a little bit in your three-point shooting the fact that you’re still that much better than average still means that that’s really really valuable and doesn’t necessarily mean that you’ve gotten worse or that you just got lucky early on I mean you can get lucky but still being consistently that better than average is still that’s a skill that’s not just luck
Pat 15:20
You know, talking about control, my mind goes to, you know, one thing coaches can control is who we play and lineups. And as coaches start to look at lineup data, in your opinion, I guess, what is some of the pitfalls of looking at lineup data in terms of what you get again can really take that is useful, what is maybe fool’s gold needs more time.
Evan Miyakawa 15:40
I’ve put a lot of work into lineup data and figuring out what is predictive of future success for a lineup and what is just kind of, you know, it’s just variance, it’s just noise. The reality is, you know, early in a season, if you pull up a list of lineups that a team has played, you’re going to have lineups that have been great and horrible all across the board just because the sample size is so small. And I think that is one of the commonly addressed weaknesses or the lack of trust with lineup data is saying like, okay, well, at what point can actually trust this data?
So just to kind of give some context for this, the first thing that I think is really important to do, no matter what with lineup data, is adjusting for quality of opposition played by that lineup. So if a lineup outscored a team 16 to four, and you put that down into points per possession on offense and defense, but you don’t know anything about like, okay, well, was that a garbage opponent they played? Or was that against a really, really good team in their conference? That makes a significant difference in terms of how much you can learn from that data. And so something I do with pretty much every metric on my site is I always adjust everything for opponent strength. And it’s not just like, oh, here’s the how good the team you played was, therefore, I’m going to adjust the lineup by how good that team was actually look at every single possession, who are the opposing five players on the court? How good are those players and then adjusting each lineup, like at the possession level, because even within the game, that can matter. That’s important.
But even still, that only goes some of the way to kind of removing some of that noise. When you look at a full season’s worth of data, I think most teams, their starting lineup will get somewhere between in a college basketball season, where you have about 35 games, a starting lineup will get somewhere between like 200 and 600 possessions, which is anywhere between like if you’re talking about full games worth of data, that’s three to seven, eight, seven games. So just think about a team that’s played for five games, you can learn a lot about a team in five games, we can have a pretty good understanding of like, hey, is this team a lot better or worse than we thought? And from knowing nothing about the team to watching five games, you can get a pretty informed decision. But when you start to whittle down like, okay, what’s your 10th most used lineup or 15th most used lineup? Most of those lineups are all under a single game’s worth of sample size, 40 to 60 possessions. From a team perspective, if you watch a team play for 30 minutes, that’s a pretty small sample size overall, you’d say I’ve learned some things from this, but how much can I learn from this overall compared to say a whole season’s worth of play? So one of the things that I’ve done with my lineup ratings in general, is if you’re just kind of looking them at face value in terms of what’s the per possession efficiency and offense and defense for this lineup, as a rule of thumb, I start to pay attention to certain trends. Once a lineup is start to hit like 30 possessions, that doesn’t mean I’m overreacting to that data. But what I’m really looking for is, is this lineup really, really good or really, really bad? Because once you start identifying lineups that even on a small sample size are abnormally good or bad, that’s something that you should at least like kind of lodge as a seed in your mind and start to pay more attention to. And maybe if it’s a really, really good lineup, you play it another minute or two per game than you would normally just to kind of, you know, it’s been really, really good so far. So you want to try and get a little bit more data on how good that lineup is because it could end up being a good option for you.
Or if it’s been really, really bad, maybe you play a little bit less. And I’ve seen some cases where there have been extreme examples where a small sample size has still ended up being very indicative of how a lineup was going to do going forward. My best example of this is Kentucky in the 2023 season was a very up and down team. They had Oscar Sheebway, but they couldn’t really figure out what lineup to play around him. This was January and Kentucky fans were kind of going crazy of figuring out like, how do we turn this team around? How do we make sure we get in the tournament? And I did a post kind of analyzing their lineups. And I found a lineup that had only just played 21 possessions. That’s like not even a full half of basketball. But that lineup in those few games that had played in small sense, you added up adjusting for opponent and all that. But it was outscoring the opponent by like three times as many points. They were scoring 150 points per 100 possessions. They were only giving it 50. It was like really, really good. But small sample size, I kind of pointed it out and said, Hey, we’re just trying to try stuff and see what sticks. One wild card play is to just lean into this lineup a little bit more and see what happens. I think it was probably coincidence. But that weekend, John Calipari played that lineup a lot more and they ended up smoking their opponent. And they turned to that lineup pretty much for the rest of the season and it turned their season around and got them comfortably in the tournament.
It ended up playing almost 200 possessions and they were outscoring the opponent by like 66 points per 100 possessions. And for reference, their normal starting lineup was outscoring opponents by just 30. Even in a small sample size, it was so, so good that it would be a little bit like a fool not to kind of pay attention to that data point and try it, especially since things weren’t working. So there have been cases like that. But in general, in terms of like taking conclusive, this is a lineup we should be playing over this one. I typically say only like 100 possessions plus is when I really start to ordering my lineup options and saying like, Hey, we have these two lineups. We’re trying to figure out which ones are starting lineup. They’ve both played 100 possessions. This one is way better than this one. Maybe we should start using this one instead.
Dan 20:58
The intriguing part of that is if and when we do come across something like this on our own teams, what to do about that then as a coach and how to go about playing that lineup more minutes. Because if you just say, well, here’s been the most efficient best lineup, and then you start them, do they become less efficient because now they’re playing against a better quality lineup in the starting lineup for another team versus, hey, we’re going to work this into the middle of the first half more for extended periods because we’re going to be playing against subs of the other team.
I guess those considerations as far as how much more or less do you play lineups or players together based off of information you find, good or bad?
Evan Miyakawa 21:40
As a coach, you’re not just putting players out and saying, okay, the first four minutes want to play this lineup, and then I play this lineup. It’s often more about individual rotations, right? So it’s a bit hard to control that.
To your point, okay, well, if I put this lineup in as a starting lineup, we’ll actually do worse because it’s playing in a harder context, essentially, right? Well, that’s one of the things where adjusting for opponent’s strength down to the possession level really does matter because you’re not just adjusting for the general team you’re playing. You’re actually saying like, hey, when this lineup was on the court in this game, they played against five other really good players. So I’m going to give that lineup more credit than in the previous game when that lineup was in against the other teams, two or three subs were on the floor, for example. And so a lot of what I do is adjusting for that. But even so, like you said, it is a bit difficult. I had an interesting story from last year where another team that was looking at my lineup data pretty closely and I was having conversations with them about it early in the season was Clemson. Clemson had a really good season last year. Their starting lineup had, in terms of just the on-court performance, they had a great starting five. It had just been okay. But they had a lineup where they just switched one player. One bench guy came in and start or if you basically took one guy off the shooting guard and put the other shooting guard in. In like December, this lineup had played. It was probably one of their top four or five most used lineups. It was an obvious substitution to make, but it was not the starting lineup. And it was doing really, really good. By far their best performing lineup with a decent sample size. And so I was talking to their coach about it and he was, okay, I’ve noticed this. This is really cool. I’m hoping we get to use this lineup more, but in their specific case, they didn’t want to upset the starter by saying like, hey, we’re just going to start replacing with this other guy because of lineup data. Like that probably felt a little bit too whimsical. You know what I mean? They ended up having a game against Duke, the best opponent they played all season, where that starter normally replaced. He was sick, so he couldn’t play the game. I think he might have played, but he came off the bench because he was sick, but they were able to start this really efficient lineup. And they ended up beating Duke, who was one of the best teams in college basketball last year. They had a huge victory against them. And then they were like, well, this lineup is so good and we just beat Duke. Now all the public media is going to start saying like, maybe we should start this lineup going forward because we beat Duke. We just happened into it. The reality was they were already identifying this trend on the backend. They just found a very convenient situation in which to start it. They ended up starting it for most of the rest of the year until one of those starters got hurt. And it ended up being a really, really great lineup for them. Led to their best basketball by four of the year. That’s some of the new ones you have to deal with with a coach of trying to get that stuff to work.
Dan 24:11
For sure, when we’re thinking about lineups is how long of a stretch players can play and be highly efficient, and I know there’s different schools of thought, the US, Europe, how they sub, media timeouts, all those kinds of things that are in, but where you see or work with coaches on how long of a stretch a player or lineup plays together.
Evan Miyakawa 24:32
In general, I’m kind of in the school of thought of you want to try and maximize as much as you can your best players playing four minutes. This is kind of the philosophy I share most of the time when talking about players in foul trouble. I would rather have a guy foul out because he maxed out the number of minutes he played, than us lose the game because I sat him for longer than he should have.
And he could have played four minutes longer, but I was worried about his foul trouble. But then he never picked up a foul the rest of the game, so we could have played him longer. My philosophy with that in general is there is a sort of skill or efficiency kind of curve when it comes to the longer you play or the higher usage you have, your efficiency in different areas drops down, whether that’s at the player level or potentially at the lineup level. Although I think for lineups, unless we’re talking about you’re playing your starting lineup, your main lineup for 10 minutes straight because you just don’t trust anyone else to go on the floor, you know, I think you certainly could see a bit of a tired drop off for a lineup like that. But the most helpful thing I can add to this is I think it is really helpful to experiment with lineups early in the season. Because what you’re trying to do is you have some prior knowledge or prior guesses as to what’s going to be good for your team going into a season, especially in a sport like college basketball, where actually pretty much most sports, the postseason is what really matters, right? Your regular season success matters somewhat. But if you’re great, and then you sort of stumble to the finish line, you’re not good in the postseason, no one’s going to care. So a lot of it is about figuring out what you can do in the early season to figure out how many options you have and optimize those options so that when you get to the late season, you really know what’s working and you’re not having to kind of figure out things as you’re going.
Something that teams can fall into sometimes too early is being too scared too early to try a lot of their different rotations, even in bigger games, because they’re like, I want to win this game. And then by the end of the season, they’ve not really experimented with enough things. And then they don’t have enough really good options to lean on towards the crunch end of the season. Or if a key player gets hurt, they’ve never experimented with not playing with that player for a significant amount of time, rotations, things like that. So in general, I’m a big proponent of, hey, maybe it is slightly suboptimal to be playing this guy 12 minutes instead of five minutes in this particular game in December. But this will pay off when you get to February or March in college basketball, because you’ll have a much better understanding of, hey, can this guy cut it? And we even helped him develop and get comfortable with this role? Or no, this isn’t working, but we have enough data to know that for sure. And so we’re just going to stay away from that particular situation. So that’s one of the things I’m generally a proponent of.
Pat 27:31
And then on this thread, you know, it’s conversation we had before when you’re experimenting even with just like players, you know, how much of a run should you give them to judge if they’ve been impactful in a game? So we look at lineups and he said like maybe at 30 possessions, there’s something there worth experimenting 100.
You can do some things, but if we look at just the game and you’re going to experiment with a lineup earlier, you’re going to give it a chance, how many minutes or possessions do you think coaches should give that lineup just within that game say like, okay, they did something positive. They were, no, it’s not working versus we don’t know because we didn’t give it a fair shake.
Evan Miyakawa 28:07
Like down to a single game level, I do not think that there’s much of a difference between giving a lineup three minutes versus five minutes, because the reality is in a given game, there are so many different combinations of lineups you play. I don’t know the numbers off the top of my head, but most teams will often play at least 10 to 15 lineups in a game, just based on all the different possible combinations of players that you can have.
And over the course of a season, most teams have close to 100 lineup combinations they’ve seen in that season. In a given game, there is so much possession to possession variance. That really is the biggest form of variance in basketball is just most possessions, you either have zero, one, two, or three points, and the best teams get those higher numbers a slightly better percentage of the time than the low numbers. But like when you’re talking about, oh, this lineup played 10 possessions, there is so much variance that can happen and you can just luck into good numbers or bad numbers down into a single game. If I’m a coach and I’m saying, I’m interested in playing this lineup more, and I want to learn something about this lineup in this game today, so I’m going to play this lineup more, and so maybe you play them two minutes more than you normally would, I still would not really take anything from that. Now, it can help inform what you think is happening. If you think this lineup is going to change the game for us and it goes in and it changes the game, then you can be like, great, I have a little bit more confidence to play this lineup going forward, but I wouldn’t just immediately be like, in this spot, I lean on this lineup and it played awesome, therefore, stamp of approval, this lineup is going to be great going forward because you could do the same thing next game and it not be great just because the sample sizes are still so small. So for me, I wouldn’t personally learn too much from any single game when it comes to lineup.
Pat 29:48
When we look at defensive metrics. And if we just look at threes rim or free throws, you know, me and Dan have conversations that how you approach your defense, you can win in different ways by prioritizing the rim trying to take away the threes. But when you look at the stats, what is the interplay between those three metrics? And can teams take away three and rim or you know, if they’re going to prioritize rim is foul free throws going to usually go up? What should coaches maybe be understanding and how these threes rims free throws kind of interweave and work based on what you maybe want to prioritize. I think
Evan Miyakawa 30:22
A lot of it comes down to if you’re able to take away one or two which are the most important ones. Now, I think part of this is one of the challenges that you have with measuring defense at a team level and a player level. I think a theory that a lot of people would probably have is that when you’re talking about the skill of players on the court, the offensive skill or ability does typically trump defensive skill or ability when we’re talking about how much influence they can have on a game. I think that’s probably more so true at the NBA level than the college level.
I think that there are things, the college level players are not astronomically good such that a single extremely gifted player can just like no matter what a defense throws at them, it doesn’t really matter. I think in most cases, defense and defensive intensity and scheme and defensive ability does have a pretty big impact in college. But still, I would say that you typically give if you have an equally talented offensive player and a typically equally talented defensive player, offense is typically going to win more often than not in terms of which one really matters more. I’ve actually seen this with all of my player ratings, I have an offensive and a defensive rating for every player in terms of the value that they bring to the court. And if I’m looking at last year, for example, the best offensive player in the country had a rating of plus seven points per 100 possessions, that’s how much value that player was adding offensively. And you look at the best defensive players, and it’s more like a plus five. And so if you’re saying I’d rather have the which one would rather have the best offensive or best defensive player, typically, the offense is the one that the defense still really matters. But the offense gets a little bit more weight in terms of like the best skilled players will have a bigger difference than the best defense, if that makes sense.
So in general, my thought kind of back to what you said is that you can do a lot to control things defensively, especially at the college level, I don’t know, maybe it’s more so at the NBA level, actually, but I don’t know if you can always just say consistently not in and out our defense is going to take away all three of these areas. It’s often more about like, consistently, this is the area that we are able to our skill set really works or scheme really works to do a really good job of taking this away. And we’re okay with the others, or for even an opponent saying like, hey, looking at their keys to success going back to those key team stats, right? If you’re going into a game and saying which one should we prioritize? Looking at the history under that team or that coach and saying, how have they fared if they shot the three ball really well or really poorly? Or how have they fared if they got to the rim a lot or got the free throw line a lot? And a lot of times that can actually be the most helpful thing to look at as saying, hey, this particular team, when they get to the rim a lot, they’re awesome.
Evan Miyakawa 32:56
And when they don’t, they really struggle. So let’s really lean in on on limiting that and that sort of being the key to that game.
Dan 33:04
Evan, this has been awesome so far. Appreciate your thoughts on all that stuff.
We want to transition now to a segment on the show we call Start, Sub, or Sit. We’re gonna give you three options around a topic, ask you to start one of those options, Sub 1 and then Sit 1, speaking of lineup data, and then we will discuss from there. So, Evan, if you’re ready, we’ll dive in this first one. Sounds great. All right, first one has to do with halftime box scores and whether or not you’re MBA all the way down in the high school below, you might not have access to high-level analytics in the game, things like that, but people are getting box scores at halftime. And we’re gonna ask you to start, Sub, Sit. What’s the most impactful combination of two stats that we’ll give you here that coaches can look at at halftime to actually make hopefully correct adjustments or whatnot going into that second half and so we’re gonna kind of work out the four factors here and we’re gonna take into account, we’re gonna give you one for each one of these, which is the field goal percentage. So field goal percentage is gonna be in all of these but field goal percentage plus another stat, which one of those combinations is the most impactful for winning would be your start.
Evan Miyakawa 34:19
Just to clarify, I could go two different ways with this. One would be, this is the most predictive for who’s gonna win the game, versus this is a thing that I can address at halftime and try and change to alter our chance at winning the game.
Dan 34:32
Yeah, that one, because I think that one’s going to be more coach-driven, that coaches are discussing that. Got it.
Start, subset, most impactful coupling of stats on the box score. Option one is field goal percentage plus your rebounds. So looking at field goals and then the rebound margin. Option two is field goal percentage and looking at turnovers, the turnover margin. Option three is field goal percentage plus free throws, the free throw margin. So your start here would be the combo that’s most impactful.
Evan Miyakawa 35:03
So we have rebounds, turnovers, and free throws. I think I’m gonna go start with turnovers. For me, the impact on who’s gonna win the game most, I think it’s probably pretty close between turnovers and rebounds, but I think there is more that you can do schematically to change the turnovers, both in terms of changing what you’re doing offensively to not put yourself in as many weak spots or making as many risky plays. And defensively, there’s a lot you can do defensively to try and change how much pressure a team is getting that you’re able to force them into making mistakes. I would sub the rebounds and I would bench free throws.
Dan 35:43
I’m gonna start with the turnovers, and for coaches, it’s like you go in at halftime, and part of what Pat and I were talking about beforehand, and this is actually a question we got from a listener of the show. Basically, you get in there at halftime, and you’re looking at the box score.
How do you, within those few minutes you meet as a staff before you meet with your team, make correct decisions on where to go in the game? And there’s all these things you could look at, but what do you just need to narrow in on that’s most impactful? So, obviously, field goal percentage being probably the number one. If you’re not shooting it well, it doesn’t matter what you do, it’s gonna be a tough night. But with the turnovers, finding the optimal amount of aggressiveness, pace of play, possessions, versus the limiting turnovers, and what coaches maybe can do or what to think about in the turnover situation.
Evan Miyakawa 36:33
A lot of it is the context of, you know, if you identify a problem at halftime, you know, hey, you know, we’re getting grossly out rebounded, and we thought that coming into this game, we identified this as an area that we need to focus on and we think we can be successful, then oftentimes, you probably already kind of know what you need to try and do better with rebounds. You’re not necessarily having to reinvent the wheel completely at halftime, whereas if it’s an area that you hadn’t identified at all as something that you need to focus on, and then you’re getting crushed in that area, then you’re kind of having to make some more adjustments on the fly that you hadn’t really planned for.
And sometimes that can be hit or miss as to whether that actually worked. So I think a lot of it comes within the context of have you already prepped for this? And, you know, when it comes back to those keys to victory, we talked about earlier, if you already has a staff are going in saying like, hey, these two metrics, we really need to be good at, say, our two point shooting, our percentage at the rim and our turnovers. And you know kind of what you need to do in those areas and then you get to halftime and those are both not doing well, then it’s like we identified this as an area that we need to do well in and we’re not doing well. So we really need to try and figure out what we can do to get this back on track because we know that this is going to be important in this game and it’s proving to be important because we’re getting killed and we’re not doing well in these areas.
Pat 37:47
Just looking at the stat sheet, we isolate turnovers, rebounds, free throws. And since everything is so context driven, what should coaches be aware of when you’re looking at those three things, maybe looking at other parts of the stat sheet in terms of, okay, well, turnovers are high, but it’s a fast pace, like there’s been X amount of possessions.
So it’s just like more possessions, more turnovers, or our free throws are really down or we’re getting crushed. But you know, we had a great shooting out shot 50%. How should coaches then maybe understand if they dig deeper on the stat sheet, if they think this is really high, this is really low, where the correlations may come that can also explain what they’re seeing.
Evan Miyakawa 38:24
I think maybe my answer to this comes in, when I typically look at a box score in terms of the team stats, I’m immediately looking for discrepancies between both teams and kind of looking at the give and take. Okay, if one team had 50 rebounds and the other team had 30, okay, that team was way better at rebounds, that’s great. I’m immediately looking for and saying, did they give something else up somewhere else? If I then go down to the turnover and say, well, this team out rebounded the other team by 30, but they had 10 more turnovers that they lost, actually that may not be worth the win in that area because you took a bigger loss somewhere else.
Or I might be looking at the number of free throws. If you out read bounded your opponent by 15, but then you had 25 fewer free throws in the other team, that may not be worth that discrepancy. So often I’m looking for, hey, what are the areas in which we did worse than the other team? And we want to correct that, but if we’re getting way more advantage in other areas, that’s probably fine, because you ultimately want to be having an advantage in more areas than you’re having a disadvantage. So I’m looking at not just individual ones, but all of them together and saying, where are the biggest discrepancies? And typically if I see one big discrepancy, there will be an opposite one in a different category if a game’s pretty even. And if I’m willing to live with that, giving up in one area to be much better than the other, that can be fine in some cases. So that’s often kind of the context I’m looking at as saying all of these together, where are the big differences and which team has more of them essentially.
Dan 39:52
Yeah, it makes sense. Offensive rebounding is a concept. Subject comes up all the time on the podcast and trying to get extra possessions and how many players to send to the glass and what an offensive rebounding percentage, what it looks like for great offensive rebounding teams and how to like try to get to that percentage. Any thoughts there on just offensive rebounding percentage and where the great teams are above or at?
Evan Miyakawa 40:16
I have on my lineup page, I have something that actually predicts what each individual lineup’s predicted offensive rebounding rate is going forward. So I’m curious to see what the top ones in Division 1 are.
So the best lineup in Division 1 right now has a predicted offensive rebounding rate of 48%. That means that they would be getting almost half of the possible rebounds on the offensive end. Now, that is, for a given lineup, that is extremely, extremely high. If I’m actually looking at this team, which is probably the best offensive rebounding team in the country, looks like their average is around 40%. 40% in college basketball is like elite, elite for a team of if you’re getting 40% in a game, that is an absolutely huge deal. And I think people notice that when they’re watching, oh my word, this team keeps getting offensive rebound off offensive rebound. I don’t think most teams can match 40%, but that is the upper tier. If you’re hitting that, man, you’re doing really well on that end.
Pat 42:05
All right, Evan, our last start subset for you, looking at your site and talking to you before, you talked about the kill shot. And it’s the ability of the team to either go on a run or not give up a run and the predictive that they will have success or win that game.
So I’d like to give you three things about this kill shot going on a run. And your start would be which one you think would be the most important for a coach, how they coach the game in that moment, strategize in that moment, to make this run, stop a run and increase their chance of winning the game. So option one, would it be for a coach to make a 10-0 run? So in that moment, do what it takes to get a 10-0 run. Would it be option two, prevent a 10-0 run? Or would it be option three, understanding when these possible runs are or are not occurring? So start of the game, middle third of the game, late third.
Evan Miyakawa 43:05
Good question. So, start would be preventing a 10-0 run, sub would be going on a 10-0 run, and then bench would be understanding when these happen, but obviously they’re all important. Yeah, the kill shot is a metric that I basically created a couple years ago as I was watching basketball trying to figure out, in games we are very aware of momentum, we’re very aware of, hey, this game seemingly changed in a matter of minutes because this team suddenly started playing great and this team played poorly, whether that’s them scoring three buckets in a row or going on a 10-0 run or 15-2 run or something like that. And I wanted to actually figure out a way of measuring which teams are really good at going on these scoring spurts across a whole season, and not just even measuring like, hey, these teams are really good at going on these runs, but these teams are often in these games that are back and forth and back and forth, and they’re really just kind of topsy-turvy because you have all these big swings, and that’s just kind of how this team is. And so basically I’ve been measuring this over the course of seasons for the last several years of every time a team goes on a 10-0 run, it’s a very cut and dry way of measuring it. There are obviously other runs that are important, like a 12-2 run, but for the sake of just interpretability and tracking it live, 10-0 runs are really easy to track, and often if you’re watching a TV broadcast, they’ll even put up a little graphic that says, 10-0 run in the last 134 or something like that. So as an audience, you can often be pretty aware of this. And so what I’ve looked at is seeing which teams go on the most of these 10-0 runs, which teams also give up the fewest runs in terms of they don’t give up scoring momentum against them. And how predictive is that? And then is that important for having a good team? And what I basically found is that in most games, there more often than not is some sort of 10-0 run. It happens pretty often, pretty often as in you’ll see them at least once in a game fairly often. So a team with at least one 10-0 run in a game wins over 70% of the time. If you have more 10-0 runs or what I call kill shots in a game, you win over 80% of the time. And if you’re getting two or more than the longer your run or the more of them you have, your probability of winning keeps going up. And what I’ve found kind of going back to your question of the start subset is that the best teams, the teams that end up winning titles or making final fours or being the most successful than most are often really good at preventing 10-0 runs. When you look at the number of runs that they give up on a season, it’s typically really low. And the reason why I think that is more predictive of success is because oftentimes these teams, if you’re going to be a national title caliber team, typically most games you are going to be in control and you’re going to win most games more often than not.
Evan Miyakawa 45:38
And really what you’re trying to do is as you’re trying to assert control over a game, the fewest amount of things that can go out of control, the better. And so oftentimes some of the best teams are the ones that they are very, very trustworthy with the lead. If you’re up six or if you’re up eight, you’re not really that concerned that a team is going to quickly claw that back and put you all of a sudden no longer in the driver’s seat. And so that has been a little bit more predictive of team success in terms of teams that don’t give up 10-0 runs rather than teams that go on lots of them. Because a lot of times those teams often are more prone to giving them up two sometimes. There are sometimes teams that they have tons of 10-0 runs in a season, but they also give up a decent amount too. And so that means that when push comes to shove in the postseason, there’s still a little bit of anxiety if they could have a big run, but they also could give it up in return so they’re maybe not as secure. And if things just don’t work out in the direction and all of a sudden they’re in a game where they give up a run or two, then you’re kind of toast. So I have found that the best teams like Houston, for example, has been a team that consistently over the years has not given up many 10-0 runs. And additionally, a lot of the 10-0 runs that Houston has is not because they’re offensively electric, it’s because they’re defensively stifling. And so I’ve also seen the length of time that it takes for a team to go on a 10-0 run is pretty important too. Because a lot of times we think about 10-0 runs as, oh my word, they just scored like three threes and a two, and they got to the line. And it all happened in the course of five possessions, and it was just really quick. But oftentimes the runs that matter more are the ones where you have a 10-0 run over the course of four or five minutes, where your offensive production is whatever, but defensively the other team is just you’re completely stifling them and they can’t score at all. That often is way more powerful because you’re just sucking the life out of the other team. And so it’s your defense causing that 10-0 run to happen that ends up being way more powerful. And so I’ve often found that to be the case as well.
Pat 47:36
I like what you said, the length of the run. That was also where my mind was going. But I’d like to follow up on your sit. Why is the when of this run or not run, I guess, preventing a run. I mean, I know it has some value, but not as important to you in your mind or what the stats are saying.
Evan Miyakawa 47:53
I mean, you forced me in a tough position there. I had to sit one and I didn’t want to so I’m having to still say It’s very important I think one of the things that does come into mind is that when I’m tracking these kill shots over the course of a season There are some kill shots that don’t matter in the sense of you might be already up 20 and then you put on a 10-0 run You’ve already won the game at that point or similarly there are some kill shots where a team is up by like 32 points and then the other team scores like 12 points unanswered and Then your lead is down to 20 and it’s like, okay We might need to like lock in a little bit here But this lead is still comfortable that can happen Especially in blowouts if you’re up by 40 you can give up a kill shot and it’s kind of meaningless so that is just something to keep in mind of like not every 10-0 run is created equal a Decent number of them happen in games or situations where the result is kind of already pretty much decided But in a closely contested game a kill shot at the wrong time you can make or break the game I think what’s really interesting though is like in a lot of high-level games that I watch Oftentimes what I’ll see is that it’s typically about who can have the more of those Scoring runs or those periods of time when they’re really dominating over a team because oftentimes in a high-level game You’ll have both teams have a period like that if it’s a tightly contested game and then your team goes on a run You’re up eight Typically you expect okay Well at some point this other team is gonna put on a run and they’re gonna come back and no matter how hard you fight it that might just happen and It’s really just about not letting that snowball you might give up one run back But you don’t want to give up two or three or have that run be extended across 10-15 minutes where all of a sudden they’ve outscored you 40 to 15.
It’s sometimes hard to Completely eliminate runs against you if you’re playing a really good opponent But it’s really more about the number of runs or the length of time in which they’re clearly have the advantage That can often kind of be the case more than just oh, we can’t have any run happen against us whatsoever It’s more just okay. We need to limit when this happens and maybe go on another one of our own
Pat 49:56
In that vein, the ability of a coach to control a run or prevent a run, but with timeouts being a limited source, how much a course coach should be paying attention to this and understanding they’ll know when these runs are having, versus selling out to like, hey, let’s take the time out. Let’s make sure we get a bucket here to go on the 10-0 run or calling it early versus maybe they’ve gone 6-0 or now it’s 8-0, now we should call it time out. How much should coaches be selling out when they can have a chance to control these moments of runs or preventing runs?
Evan Miyakawa 50:29
I don’t have any research to back this up as much in terms of the timing of timeouts and how they contribute to these runs. My best guess at this would be that oftentimes I think some of the best timeouts are when you as a coach either identify one of two things. Either there’s something schematically that I’ve identified as going completely wrong that we can easily adjust and I just want to take a timeout to be able to adjust this that this doesn’t keep happening. Or two, my players are not putting in enough effort and I need to shake them and get them kind of locked back in. And a lot of times, either one of those two things, if you call a timeout at the right time and prevent that stretch from going further of, my players don’t seem dialed in or schematically this thing is wrong and we can address this really quickly. If you’re able to call that as early as possible once you’ve recognized that, then that can potentially eliminate a really bad stretch that you’re giving up. But if you think the process is right and the opponent has just hit two or three step back threes in a row and your process has been good and you’ve just gone unlucky and you only have a limited number of timeouts, maybe you just trust the process and you use that timeout for when one of those other situations happens later. So that would be my initial kind of take on that. But if you identify something that you can change, the early you call it to maybe you stop it at a 6-0 run instead of a 12-0 run against you, that can be really big.
Dan 51:51
Evan, you’re off the start, sub, or sit hot seat. Thanks for going through all those options with us. That was a ton of fun. And we’ve got a final question to close here, but before we do, thanks again for your time, your thoughts. This was a really interesting and helpful conversation. So thanks for coming on today.
Evan Miyakawa 52:05
Yeah, really glad this was awesome.
Dan 52:06
That’s awesome. Thank you, appreciate that. Thank you, Evan. Evan, our last question that we ask all the guests is what’s the best investment that you’ve made in your career?
Evan Miyakawa 52:15
I’ve got two answers for this. One easy one that I’ll briefly mention. The easy one is I have my PhD in statistics and I think that getting that education really, really helped me in my career, not just in sports, but just in general. I’ve always been a big stats guy. I’ve always been a huge sports guy. And so being able to do sports analytics is something I’ve always dreamed about doing. And part of how I built my platform is just by starting to share what I was doing on social media and publicly and getting some feedback on it. But I think a lot of even just the credibility that comes with having a graduate degree certainly gives some level of, hey, this guy probably knows what he’s talking about. But at the end of the day, it’s not just a title. It does mean something, but I do think that the skills that I gained investing in that education were really valuable.
But I think the more interesting one that I want to talk about that’s approachable for everybody is learning and practicing communication. So something that I’ve done from a very early time with my website is just trying to put as much of my work out there in the public as possible to figure out what is the best way I can communicate these ideas in a way that is tangible for people who are not data people like me, who are fans, coaches, et cetera, who want to find this valuable. And so practicing, how do I write things that communicate that? How do I speak in a way that communicates that? And especially from the speaking part, I try and have as many conversations with people like you guys as possible because most of my work, a lot of my day consists of programming and coding and math and these very technical things. But at the end of the day, it’s only as useful as how I’m able to communicate these ideas and make them tangible and usable for the end users. I’ve practiced a lot and try and practice as much as possible. How do I take these complex ideas and translating them into a way that each person who’s learning something or listening can be like, ah, that’s what I wanted to latch onto. I’ve learned something, I can use something and boiling it down to the really simple things. And so practicing communicating, practicing talking about the things that you’re thinking about and passionate about and communicating them to as many different kinds of audiences as possible. People who are just like you and I really understand it or people who don’t understand it much at all, but you know that you have something valuable you can give them. That’s been really, really important and I think is a big part of my platform is I don’t just have all these nerdy spreadsheets. I actually try and make them, whether it’s someone who’s visiting my site and trying to figure out how to use it, or whether it’s someone seeing posts and I’m doing social media or someone listening to me on an interview, trying to make it as digestible as possible. And I think that’s been really, really valuable.
Dan 55:08
All right, Pat. Hey, that was awesome.
Let’s hop into this recap. Anybody that knows Evan’s work knows how high level his stuff is. And another interesting conversation today on things that can help us as coaches from the analytical side. Well, let’s go ahead and dive in and for our first takeaway, I’ll throw it to you.
Pat 55:25
Yeah, so I’ll start with the first bucket. And I’m the chance to talk with Evan before and you know, in our prep, I enjoyed where we took it because I think these stats are obviously very, very interesting can be extremely beneficial to your team and what it’s telling you about your team. But it can also be very overwhelming as a coach, I know, we ourselves get lost into it. And so kind of approaching it with Evan today, like, how can we quickly get to what is useful for our team? And what is just kind of weighing us down?
Or what is distracting us from what should be? Our attention. Evan spoke tremendously on it, obviously, I mean, it’s what he does and how he tries to make his stats as predictive as possible. So it can be as actionable as possible. And when looking kind of like, yeah, these early season indicators. And it goes back to, I think, which was nice to hear from obviously know what we hear from coaches, but also from his side as the analytical side, it comes down to some combination of as a coach, you got to marry your gut to what the stats are saying. And like what you’re seeing or feeling or what you thought your team was going to be or is showing then just kind of looking at through the stats is not an all or nothing. But okay, is it supporting the data or where do we need to get better at? And it was refreshing hearing from a stat guy like we’re not completely out of the game yet. You know, it’s not all AI.
Dan 56:41
Yeah, just set up a computer on the first bench and do the subs and the timeouts.
Pat 56:48
Yeah. So I quickly, I think at the end, when we got into the lineup stuff, that was a conversation I was itching to have something we haven’t talked about before. And I was just giving my two big takeaways there. And I gave a lot of good kind of the stats, what the coaches should be understanding some good examples. But to me, it was early season coaches should be willing to try different lineups and then also avoid the pitfall of overreacting off of one game.
If you want to try one lineup and the one game, it goes bad. That’s still not enough to say it’s never going to work or if it goes great, it’s like, it’s always going to work. You know, I thought he gave a great thing talking, you know, 30 possessions. If you can look at 30 possessions, maybe there’s something there to continue to pick at or expound upon a hundred possessions is maybe when you start to get like concrete answers. But early on as coaches, it’s beneficial if we take the risk, because as we all know, it’s long seasons, injuries are going to happen. People are going to get sick, whatnot. And you need to then have some trust and some lineups and be able to steer the ship somewhere.
Dan 57:47
Yeah. To your point, I agree. It’s also interesting of what kind of team you have, what kind of coaching situation you’re in, what the rest of your schedule looks like, how you’re going to potentially do in your conference, because taking a lot of chances with lineups and players early, if you take losses, there’s the human side of this that can be detrimental to, well, hey, coach played all these different lineups and we ended up dropping a game and now we’re under 500. And then the stuff in the locker room takes place versus you got a mature team, say top seven or eight, pretty solid, but you’re able to maybe play some periphery guys on the back end that you can still win the game and you’re still confident. This is why it’s the art of coaching. And so like, yes, you trend on playing some lineups and seeing different things early because you need that. What was interesting is it got later, sometimes you get these little coaching gifts early in the season where it feels maybe bad in the moment, but it actually kind of helps you with this stuff, which is, I don’t know, early game foul trouble for a key player where you’re forced to go to another lineup or another guy where it doesn’t like hurt the flow or the locker room because it’s like you needed to do that. Or the case with Clemson, where you talked about you find a more efficient lineup, but you can’t just snap your fingers and change a starting lineup and think it’s not going to affect everything in the locker room, but player was sick, then they played better. And then like, as a coach, like, okay, this makes sense. Or even we’ve heard coaches on here talk about sometimes a minor, we hate to say an injury is good, but like someone tweaks an ankle and has to sit out a game, but they’re going to be back the next game. Well, that game forces you to then play these lineups. So I guess what I’m getting at is this is like, he talked about the human side of things. How can you see different lineups while keeping your team together and hopefully win games?
Pat 59:32
Like you said, that’s the art. And when you’re gifted these moments, maybe try out small lineups, of course, every arc of a team is different. If you’re a veteran team, and you have a strong feeling, you’re going to be good. And you know, late season, like, hey, we’re going to need to be able to play small, we’re going to need to be able to switch early on, experimenting lineups that way, you know, because it’s like, all right, I’m confident we’re going to win games.
Now, of course, yeah, if you’re a young team, a young coach, or there’s pressure on the line. Yeah, then of course, like it gets harder, just I want to see a switchable lineup tonight, because you know, you’re thinking March, we’re going to win was what we need to win right now.
Dan 01:00:09
We need to win on Tuesday, so our practice on Wednesday is not terrible.
Pat 01:00:13
Yeah, I get that. Like it’s definitely a fine line between like, well, as a machine, it was said to like, even with lines you can look like the fine line between things, stubborn and innovative. It can be very blurry at times.
Dan 01:00:26
Yeah, and I think like this where Evan was great of giving you, here’s about the number of possessions that make sense. Here’s like the things to really look at because I think what we wanted to have in this first conversation is how you can take early season data, whatever it is, and optimize for the right things. And so if you do have those two or three young guys that you want to get in games and get them some minutes, how many possessions then gives you an approximate starting point for whether or not it’s working or not. I think Evan was great on that today.
Whatever your philosophy, whatever your coaching situation is as far as getting a different look at all these different lineups, like he said, like three to five minutes in a game, like he said, between three minutes and five minutes. If you’re just looking at a one game thing, there’s not a lot there, but if you’re able to get three, four minutes a game for these other substitution patterns, that should be enough maybe over the course of four or five games to give you what you should do next.
Pat 01:01:18
Dan, let’s keep it rolling here. I’ll throw it to you for the second takeaway of our conversation.
Dan 01:01:23
Yeah, I’ll go to my start subset, which was the halftime box scores. And so I think, you know, you and I were thinking, we had some coaches follow up after, we had a great conversation with Ken Pomeroy and Jordan Sperber on analytics and interesting stuff here. And then we had, you know, coaches reach out, follow-ups and more. I think one of the things was the halftime box score and the kind of four factors or things that coaches can really look at.
And then like make a approximate guess as to what to do without overreacting to other things. So I think we took into account like, yeah, your field goal percentage is probably the number one driver. I think analytically that backs up to like, if you don’t shoot well from the field or you’re giving up high percentage from the field, you’re in for a tough night. So take that into account with the rebounds, turnovers and free throws. I also just took away that again, field goal percentage on defense. I think it might’ve been your follow-up or I forgot where it was in the conversation, but just how incredibly important that is to try to optimize for your field goal percentage on defense is what I’m talking about.
Pat 01:02:27
Yeah, you know, again, we kind of came back with our conversation referencing Ken Pomeroy and Jordan sperm, or just how much influence we actually have over turnovers and looking at that at the halftime on our offensive end, whatever it is, you’re reinforcing the decisions your players are making, who you’re choosing to play as a Jeff and Gundy likes to say, you know, dumb gets you beat. So, but yeah, whether it’s, you know, getting back to playing off a two or just emphasizing brim decisions, but then also like defensively, if like turnovers is, you know, we got to turn them over more.
And, you know, I think we’ve been looking at a lot through SGTV, but of course it’s like increasing pressure, but even as coaches, we start to build out our base defense, just ways we can think about implementing techniques to just try to generate some more returners without having to like completely sell out. I’m like, yeah, okay, we need to become full court pressure teams. You know, we looked at an interesting technique from coach Patino last year at New Mexico, now with Xavier about just jumping on ball screen and coverages on pickups and just that increasing deflections, these opportunistic times that we as coaches actually do have some control over. And then that can really pay dividends when you look at your stat sheets, because I think Evan raised a really good point. It wasn’t this starts upset. It’s all starting to blur together now, but offensive skill is going to win out over defensive skill historically. And, you know, of course in the NBA, but it trickles down to every level, but good offense beats good defense. And so how as coaches can we be thinking just to try to change the margins a little bit and, you know, I think really for me, these last couple of conversations, analytics got me thinking a lot about turnovers and it doesn’t have to be, like I said, changing like overhaul of how you pressure.
Dan 01:04:08
Yeah, I also wrote that down, the offense versus defensive players. And I think he said the best offensive player last year had a differential offensively, a plus seven, the best defensive player had a plus five.
And so I think we heard that too, from Paul Maroy and Sperber about more efficient offense beats more efficient defense and go back to our conversation with Utah Jazz head coach, Will Hardy. We talked about a kind of late game with him and what you’re optimizing for from a lineup perspective. And I remember him talking about you got to score the ball for your team to have a chance to win and like, yeah, you might be able to throw out a great defensive lineup, but if that team can’t score, it’s obviously a problem to close out and win games. And so I remember him talking about obviously every situation is different. You can make subs on free throws and timeouts for defense, but like overall, you’re trying to optimize to be able to score. And I think there’s also, it’s not an analytic, but there’s a mental factor to your team when you can’t score the basketball and how that affects your defense. I know we say you got to defend on every possession and all that, but if you go empty trips, two, three, four, five times in a row, it’s hard to play that tough defense every single time as much as we want them to. And you say it just human nature, it’s deflating. If you can’t put the ball in the basket for your defense too. I think that’s a nuance that doesn’t show up on the stats, but when you go more offense and you can make a three or get to the free throw line, you can set your defense. I think that has a big impact on your team overall.
Pat 01:05:38
Yeah, I mean, honestly, I’ll just take it from here because I think that bleeds really nicely. And our last takeaway talking about this kill shot.
And I know we’ve talked outside of the pod a lot about making a 10 Oh run and the predictiveness and having winning that game. But what was fun to hear Evan talk about that maybe more so preventing 10 Oh runs has a higher impact on your ability to win that game. And I think it goes back to everything we were just saying, which, yeah, offensive skill is so high, like if your defense can hold teams to never make 10 Oh runs, I think that’s saying something a lot about your defense. And I like this point, the length of time, being a factor. So maybe you guys aren’t scoring, but it’s taking them four or five minutes, we’re making even an eight Oh run, you’re still in the fight, you know, you’re still in the game. So I really like that point. And I thought my follow up on then like, okay, how does that influence timeout decision making, like being aware, like, hey, it’s eight Oh run, balls out of bounds. Should we call a timeout, set something up or let it flow or for a run against a six Oh run, should we do it now versus eight Oh run, like, you know, we definitely got to do it. He raised good points, you know, just approaching it, I think as we all do, is there systematically something just wrong that we can fix is effort bad, that we can’t substitute anymore or need to, you know, call a timeout to substitute or versus Yeah, it’s just like, hey, again, like the process is all right. We’re unlucky so far. But let’s trust that, you know, it’s when they get to that eight Oh, 10 Oh run, like the process is going to win out and we’ll be able to stop them there and get a bucket ourselves. So I like that advice, just being actionable with coaches and not so by the numbers that like, hey, immediately, like you got to take it because of the its ability to impact your winning or losing that game.
Dan 01:07:20
Yeah, another area where I think obviously, as Evan mentioned, just the coaching feel and the art of this thing comes into play. But something like this, the knowledge of preventing a 10-0 run is a little bit more predictable of winning than going on a 10-0 run and when it is and all that. I think that working knowledge as a coach in your mind during a game, there’s not a lot we can overly control all the time when a game’s going on. And players got to make shots, they got to make plays.
Of course, we have control over subs and all those kinds of things. But if you know, are you on the road where it’s probably more unlikely for you to go on a 10-0 run in someone else’s gym based off offensively a little harder to score. But you can at least prevent the 10-0 run to keep a seat at the table and win in the game kind of thing. Having that knowledge to, like you mentioned, okay, they go on a 6-7-0 run, we’re calling a timeout on the road especially, so we make sure we don’t give up that 10-0 run. Yeah, just like I said, if you’re having a call bunch of timeouts to stop a lot of runs, well, you might just warm up the bus. But at least, in the back of your mind, knowing this stuff I think is really good. And then I did think when it occurs is interesting too. We’ve all been in games where, I don’t know, team does go on an early run, you get down 11-2 or 12-2 and there’s so much time to climb back in versus you’re down four with six minutes to go and they do a kill shot then, well, that’s going to be a little tougher I think that takes into account as well.
Pat 01:08:47
Yeah. And us prepping for this question, you know, we all say, well, you know, whatever you come out, you immediately give a 10 Oh run, but there’s so much time for the other team to make their runs. And it’s not like whoever makes the first 10 Oh run wins, you know, over the next than 30 minutes, they continue to go on three separate 10 Oh ones. But your first minute 10 Oh run doesn’t weigh as heavy.
Dan 01:09:07
Yeah, we’ve all been in games where let’s say it’s early to mid second half and a team is like on the fritz You’re up three or four and if you go on a 10-0 run then And put them down 14-15 In the second half just mentally certain teams will I almost like give up but you can kind of put them away from a mental Perspective in that second half with that kill shot versus a first half. Maybe coach calls the timeout lights them up a little bit You know, they come out play harder. It’s like, you know, they got more time to come back So, of course, that’s another area where it does matter But I just think you know his work on the preventing I thought was an interesting thing today that I definitely took away from the whole show
Pat 01:09:45
Yeah, absolutely.
Dan 01:09:46
Pat, anything else you just thought how we have more time or if it moved in that direction would have been interesting.
Pat 01:09:52
As we were talking about the kill shot, one thing that popped in my mind, so maybe just following up on the, I mean, I know in one of our early, early conversations, we had it on the hot hand, you know, if we’re looking at like runs, well, what about a guy makes two, three shots in a row? What’s the correlation there between him making his next one or versus keep going to him, feed him or, you know, I really believe if you see a couple of guys make shots, like is it shooting contagious? Just, you know, and I guess what the working knowledge or stats are that nowadays.
Dan 01:10:21
Yeah, you’re going back to Ben Cohen. I think that was episode like six, seven, I mean. Yeah, early days. Early days, yeah, good one on that with Ben Cohen on his book, The Hot Hand. So yeah, definitely would have been interesting as well. It’d be interesting to hear Evan’s thoughts on those kinds of things. But once again, fantastic conversation. For those that haven’t checked out, Evan’s site, evanmiya.com, I know he’s got all this stuff on social media too. Just to add, he does communicate at a high, high level, all this stuff, which was part of his best investment. I mean, he really does. So make sure you check it out. Thank you everybody for listening and we’ll see you next time.