Russell Ries 0:11

Good morning, afternoon, or evening to you, wherever you're listening from. Sunday, October eighteenth. My name is Russell Ries, and you're listening to the Nashville Sports Report.

Matthew Sullivan Introduction

Russell Ries 0:38

Alright, welcome to this week's show. We've got our first returning guest today, Matthew Sullivan, my coworker and a brilliant data analytics and modeling mind, has come on to talk to me about the upcoming NBA season, how he models basketball versus football and or baseball, and a bunch of other really interesting stuff. Matt and I got together earlier this week to talk it all out, and I hope you enjoy my conversation with Matthew Sullivan.

Matthew Sullivan Interview

Russell Ries 1:11

All right. Well, Matt, welcome back to the show. I'm really excited to have you back. I've had a number of people reach out to me and say they enjoyed listening to our first conversation, and I wanted to get you into talk basketball, preview the season, all that. But before that, how are you? Welcome back.

Matthew Sullivan 1:28

I I'm doing really well. Uh it's exciting to be back on the National Sports Report and uh excited to talk about basketball.

Russell Ries 1:36

I'm excited for that too, and I'm excited that you are my first official returning guest. So thank you for that distinction and being that person. I'm not going to give you the same introductory questions I give everybody because you've already answered them all. So let's just dive right in. Last time we talked, you were building out 75teams.com. You had the football simulator underway, and basketball was sort of your main priority. How has the project evolved since we last spoke?

Matthew Sullivan 2:05

So at this point, uh my basketball simulator can run a complete game from end to end. So it picks the two players that are going to jump for the tip. It predicts who wins the tip, who recovers the tip, whether the team with the ball shoots the ball next or turns the ball over, whether there's a foul, it predicts rebounds, it predicts technical fouls, it predicts everything that happens in a game from the beginning of the game to the end of the game at this point.

Russell Ries 2:35

I want to ask a lot of questions, but I want to just start like focusing in on one thing. How do you model or predict who's going to win a tip?

Matthew Sullivan 2:43

You know, when I first started work on this months ago, I can't remember the author's name, but there was a white paper on that very subject. Uh someone has done research on this in the past. Tight has a lot to do with it, but you know, typically someone that wins 15 out of 20, that's a repeatable skill. So that's a rate that stabilizes pretty quickly. So if you see someone win several tips, then chances are that person is going to continue being good at that skill.

Russell Ries 3:18

So do you continue to ingest the data as the season goes along and improve those numbers?

Matthew Sullivan 3:24

I don't actually model again until the season's over. But yes, as the season goes on, I'll ingest the numbers from every game that gets completed. And that will continue to kind of adjust the player's talent in the simulation. So, you know, some of basketball is randomness, but basketball is um it's a lot easier to predict than something like NFL football, where there's just a whole lot of randomness baked into a single game and there aren't that many games in a season. Uh basketball is a whole lot more predictable than that.

Russell Ries 4:00

So is that a quantity thing? You just get so much more data that it makes it easier to predict?

Matthew Sullivan 4:06

Partly. There's 82 games in a regular basketball season, um, only 17 in an NFL season. But there's 22 players on the field in the NFL. There's only five in basketball. And that means that if you have a generational talent like LeBron James or Wimby, that player influences the outcome a lot more than any one player on the football field. You know, obviously having a Tom Brady or Peyton Manning, a star quarterback, can take a mediocre team to a good team. But basketball is where individual talent um has the most impact just because of the small team sizes.

Russell Ries 4:51

It's interesting you say that. I've been sitting here thinking that it would be easier to model something like baseball or football that has sort of like set iterations, like you have a play and then it stops. Whereas basketball, obviously, there's more free flow. Is that an incorrect assumption?

Matthew Sullivan 5:08

I don't know that I mean you're great to call that out. Yeah, basketball is more of a continuous kind of game, whereas with football, we have these discrete plays. Same thing with with baseball. It just means that we have to model different things. So when I simulate an NFL game, I have to simulate what personnel package the offense is going to put on the field, and then what formation they're gonna be in, and then how the defense will substitute uh because of who's on the field for the offense and what is the defense gonna blitz or not. So there's all these small little models that concern what's gonna happen on one specific play. But like I was just saying, with basketball, at any given point, there's a number of different things that could happen. You know, the player can shoot, the player can dribble the ball off his foot, somebody can call timeout. There's just all kinds of different things that can happen at any given moment in the basketball game. So it makes the the simulation different, but again, just because of the small team sizes and the fact that coaches don't matter as much in basketball as they do in football, basketball is just overall easier to predict.

Russell Ries 6:17

So it sounds like you're saying individual players are a lot more important in a basketball game than a football game. How large is that variance?

Matthew Sullivan 6:26

I think you would have to bring on one of the people that's created the advanced stats to talk about that. It's a lot, it makes a huge difference. And historically, you can look at teams like uh San Antonio back in the David Robinson years. Robinson was hurt one year. Uh San Antonio finished, I think, with the worst record in the league. They drafted Tim Duncan, and just adding Duncan to the worst team in the league and having Robinson come back, having adding two great players to a terrible team created a dynasty. So just one great player has a tremendous impact.

Russell Ries 7:06

Your website, from what I've seen lately on the basketball side, it has kind of two main sections. You've got the top-level team projections and you've got individual player projections and stats. I'm taking it from what you're saying, those two are really the meat of what's happening behind the scenes.

Matthew Sullivan 7:26

What you're actually seeing are are high-level projections. So those aren't the results of the simulations that take every play into account. Those are the kind of projections that sites like Basketball Monster or ESPN or or Yahoo develop. And they're uh they're normal regression type of projections. It's what analysts have been using for years. It's a it's a great way to do it, and it might even be the best way to do it for projecting an individual player or for projecting a team over the course of a year. What they're just not able to do is to model the interaction between players. And that's that's where the biggest question marks come in predicting the season or even predicting an individual player, is what happens when a player changes teams? You know, great player has a big impact, but what happens when that great player, in the case of LeBron James or or Jalen Brown, who are both moving from different teams to the 76ers this year, there's only one ball. So how does moving from one team to another affect that individual? And how is it going to affect the team? It's relatively easy to predict basketball when you don't have that kind of movement from team to team. Injuries and players moving from one team to another are what makes it difficult. So that's that's what the problem that the simulator will will help to solve.

Russell Ries 8:50

I remember you saying last time we talked that a quarterback's stats are sticky and that you could take them from one place, move them into another, and they're going to get relatively the same results. It sounds like to me you're saying that that's not necessarily the case as much in basketball, or am I over extrapolating?

Matthew Sullivan 9:08

No, that's that's a a great point. So normally on the football field, there's only one quarterback on the field at any given time. So moving teams isn't really going to change that much. He's going to have different receivers to throw to, he's going to have different offensive linemen blocking for him. He may come into a situation where the offensive coordinator draws up plays that he hasn't run before. But by and large, you know, Aaron Rodgers is Aaron Rodgers. Josh Allen is Josh Allen. If Josh Allen moved to a different team next year, he would still be a really good quarterback. The quarterback's role is going to be the same whenever they're on the field. When you take Jalen Brown, who last year was the centerpiece of the Celtics when he was on the floor, and you put him on a team that already has uh two of the best players in the game, and then you also bring in LeBron James, the question of what's he going to do in this new role just creates a lot of uncertainty. Uh he's not going to be able to touch the basketball as frequently as he did when he played for the Celtics.

Russell Ries 10:15

So the modeling is a lot of different hypotheses you make about how to kind of predict things. How do you test which might be the better statistical representation?

Matthew Sullivan 10:27

So we have play-by-play data, very detailed play-by-play data going back to the 1996-97 season. And by very detailed, I mean we know who was on the court for every play. We know if there was a turnover, who turned the ball over, if there's a foul, who committed the foul. We have lots of data about substitutions. So we have pretty good play-by-play data for 30 years. So when we want to use that data to simulate what's going to happen in the future, um, let's say we want to simulate where a player is going to take a shot, just meaning, like if we've already decided that Steph Curry is going to take a shot, you know, we actually have X and Y coordinates. We know historically where Steph Curry shoots the ball when he shoots the ball. So what we'll do is when we want to predict that for the future, is take those stats from the past. We weigh recent stats much more heavily than stats from four or five years ago, and just figure out like how quickly does this stabilize, just meaning that how many shots do we have to look at before they consistently give us a good distribution that shows us where he likes to shoot the ball. And some of these things stabilize really, really quickly. If we're trying to predict, you know, how good a free throw shooter somebody is, it only takes 20 attempts. Like it doesn't take very long to determine whether someone can shoot foul shots or not. But as far as the ability to make three-pointers, we need usually two and a half seasons of data for that. But that's not something that stabilizes until a player's been in the NBA for almost three years.

Russell Ries 12:13

That makes me wonder how do you handle younger players, rookies coming into the league? Not very well.

Matthew Sullivan 12:21

Like rookies are super hard to predict. We take college stats into account, but more than anything, just draft position. You know, I think a lot of us can think of famous busts, you know, players that were drafted first or second overall that didn't pan out. But by and large, scouts know what they're doing, and teams do a pretty good job of identifying talent. But it's really hard to predict how well a player will play in his rookie year. And the other side of that is rookies in general just don't play a lot. You know, stars in the league play 32, 33 minutes a game. Uh, most rookies don't hit 20. So we tend to get excited about players coming in out of college or players coming in from overseas, but you really don't see a lot of impact. And I'm speaking in general here, until the player's been in the league for two or three years.

Russell Ries 13:14

You mentioned something like free throws showing up a lot quicker than three-point shooting percentage. Are there other early indicators that start popping up within those two to three years that kind of point you towards where that rookie might be going?

Matthew Sullivan 13:27

Playing time is the thing that matters the most. You know, just going back to the discussion about why it's hard to predict teams that have new players or players coming back from injury, playing time is always the biggest question mark. We can pretty quickly establish what a player's gonna do when they're on the court. It's it's often hard to predict the the decisions that a coach or a general manager is gonna make about how that playing time gets divided up and who's gonna be on the court at the same time with whom. Yeah, with Cooper Flag last year for the Mavericks, they were a terrible team the year before. We knew Cooper Flag was gonna play a lot, and uh he didn't look great in his first few games, but um he really came on, uh needless to say, uh through the end of the season, and he looks like he's gonna be a perennial all-star. But if Cooper Flag had come on to like the Oklahoma City Thunder, he wouldn't have gotten nearly that same amount of playing time. Uh, Dylan Harper's a good example. There are some people that think the Dylan Harper's talent is on par with with Cooper Flag, but Harper was playing for the Spurs, they went to the finals and he came off the bench and rarely saw more than a couple quarters of play. So a lot just depends on on how many minutes that rookie gets, and that depends on coaching and you know just how well the team is doing. A lot of teams aren't going to devote a lot of minutes to developing a rookie when they're already contending.

Russell Ries 15:00

Are there ways to effectively model a coach's playing style or decision making that you can apply to that?

Matthew Sullivan 15:08

Yes. Yeah, coaches' rotations are very real. So I'm not the first to do this. There are a lot of sites that track the combinations of players that play on the court and uh the plus-minus, just meaning how well do these players do when they're on the court together versus how well does the team do when these players aren't on the court. And yeah, certain combinations of players tend to play a lot together. Coaches know what they're doing. They'll usually want to have someone that can score the ball. You want to have uh playmakers, you want to have rim protection, uh, you want to have good defenders. So just finding that those combinations of players, it is fairly predictable. Coaches tend to settle into rotations by mid-season.

Russell Ries 15:54

I imagine you've come to understand a lot of trends, sort of things like coaches settling in by mid-season. Has building the basketball model taught you anything about the sport that you didn't previously understand?

Matthew Sullivan 16:06

I'd say one is that we don't place enough value on team defense. Like the state of the simulator right now is that it can predict uh lots of offensive numbers really well, but it's under-predicting the difference between teams because it's not yet modeling team defense very well. And by team defense, you know, Jalen Brown is a hot topic in the NBA right now. Uh, a lot of people were astonished that the Celtics gave him up to the Sixers for an aging Paul George. But advanced metrics have never really liked Jalen Brown, which is not to say that advanced metrics don't think he's a good player. It's that on a team like the Celtics, you could make an argument that he wasn't even in their top five best players the year they won the championship. And that makes a lot of old school NBA fans angry to say, because he looks like a really good player. His own ball defense is pretty good, but by all accounts, his uh the way that he plays the passing lanes, um, a lot of the the small things on defense, he just doesn't do that very well. So that's kind of the eye-opening thing for me is that when I started to build the simulator, it didn't have any knowledge about that narrative. Like it didn't know that you know Jalen Brown didn't score high in some of these advanced metrics. But in putting together what I've put together so far, it corroborates uh what a lot of the analytics guys have been saying about Jalen Brown for the last few years. So yeah, I I would say that defense just both how important it is and how difficult it is to measure, uh, is the biggest thing that's stood out to me so far.

Russell Ries 18:02

For people listening who may not know, can you talk a little bit more about what you mean when you describe advanced metrics and maybe give an example or two of what those might be?

Matthew Sullivan 18:13

Yeah, so you know, lots of us when we look at a box score, we look at the points the player scored and the rebounds and assists. And um, I mean, those are are pretty easy for us to understand. When we talk about advanced metrics, it's usually things like uh plus-minus scores. There's a metric called box score plus minus, and it basically looks at how does the team do when this player is on the court versus how does the team do when this player is off the court? So it's you know, either a positive or a negative number. It's not perfect, but it's gives you a general idea of you know how much a player contributes to a team's success. So that's one. There's a kind of a trailblazer in in basketball, uh John Hollinger, that's that's written books. He used to to work for ESPN. He had a metric centered around mostly offense called player efficiency rating. So that was an advanced metric that that would look at you know, on offense, yeah, we understand points, but how efficient is this player? You know, Kobe Bryant, I used to love watching Kobe play. His efficiency wasn't close to what you would want in the modern NBA. So I'm not a Shea Gilgis Alexander fan, but I don't like to watch him play, but he's extremely efficient. You know, he's a perimeter player that shoots a high percentage on his shots and gets to the line as as often as anybody. So that's the kind of player that advanced metrics really like. It's not just the volume, but how efficient are they?

Russell Ries 19:56

It seems like there's a real delicate balance between the individual and the five-man unit at any given time. How do you weigh those two against each other? Sometimes it's hard to tell.

Matthew Sullivan 20:10

In the NFL, we know who the good linemen are. There are sites like Pro Football Focus that actually have someone watch every game and give every player on the field grades. But having just a single good left tackle doesn't mean that you're going to have good run blocking or pass blocking. You have to like everybody on the line has to do their job. Basketball, it's it's as we were saying earlier, it's a lot about roles and opportunity. You know, LeBron has scored more points than anybody in the history of the NBA, and he's probably gonna be the fourth or fifth leading scorer on the 76ers this year. And he could still score points if he wanted to, but what he does with that team is and we don't know yet. We we've we've only seen like a pre-season game or two at this point. What role he wants to play is gonna dictate what that team looks like when they're on the court this year. Is he gonna facilitate? Is he gonna rebound? Is is he gonna play defense? We don't know yet. We know that he's a good player, we know that he's one of the best to ever play the game, but as far as how he's gonna impact those five players on the court this year, we we really don't know until we see how they operate with each other.

Russell Ries 21:27

And I guess you must have some sort of understanding of of how a player peaks at a certain age, and then LeBron's obviously he's an ageless marvel, but he is towards the end of his career. Do you have to sort of I don't know the right word, degrade some of his stats in that sense?

Matthew Sullivan 21:45

Yes, uh but that's usually the stats that that we forecast downwards are games played in minutes, which is not to say that he won't score fewer points. I think last year was the the first time uh that he played in. Game didn't score 10 points, happened last year. But we can from year to year see trends in any player, regardless of their age, that are upward or downward. And there is like a typical curve. LeBron obviously is like this huge outlier. I mean, he's he's played through generations of basketball players at this point. There's players that have entered the league, left the league, are commentating or coaching now, and LeBron's still out there playing. But but where you see the biggest differences is player age are just how much they rest. So the biggest unknown for me for the Sixers and for LeBron are how is he going to play with these other four players on the court? You know, who's going to shoot the ball, who's going to defer to the other players. But also just how often are they going to rest him? How many minutes per game is he going to play in the regular season? I imagine it's going to be lower than it was last year for the Lakers.

Russell Ries 22:59

And you talked about the fact that so far he's only played a couple of preseason games with the Sixers. I know in the NFL, preseason is not at all predictive of who's actually going to play and what and what they'll be doing. Can you glean any information from NBA preseason games that is beneficial or is it sort of the same?

Matthew Sullivan 23:18

If you get on Reddit, people get really excited about individual preseason games. I think we see flashes of potential, especially with young players. But again, the NBA is more predictable than even baseball or the NFL. So if a player wasn't an impact player last year and he's been in the league for a while, then you really shouldn't expect a big leap this year. And so just because someone scores 30 points in the preseason game, uh doesn't mean that they're the next Michael Jordan. And I think we also overreact negatively the other way, too. You know, player has an off-night in the preseason game, people start to panic. You know, in fantasy drafts, he starts to slip if he doesn't play very well in the preseason, but no, it doesn't really mean that much.

Russell Ries 24:07

I want to pivot now to the upcoming season and just get your thoughts on what what the model's telling you, what you're seeing, any surprise teams, risers, fallers, any predictions you care to offer?

Matthew Sullivan 24:22

I'm surprised right now at the projections for the Trailblazers. Uh, that's one of those cases where there's some unknowns, and that you've got Damian Lillard returning after missing a season to injury. And, you know, he's no spring chicken either at this point in his career. So he's difficult to project, but a healthy Damian Lillard is you know one of the best players to ever play the game. And then it's going to be a crowded backcourt because they brought in Ja Morant from the Grizzlies in the offseason. As a Tennessean hated to see Morant go. I think he's an exciting player to watch, but that's a team that's it's hard to project how good they're going to be because it's hard to see how those pieces are going to fit together and how playing time is going to shake out for a team that looks like it has several players that play the same position. So I don't trust what my projections are actually telling me with them right now. Do your projections have them too high? I think so. You know, they were a playoff team last year, and a lot of people were surprised at them making the playoffs. I think they're going to continue to improve. I don't think they're going to make as big a jump as what my projections are telling me right now.

Russell Ries 25:42

As far as jumps being made in your projections, are there any that you're seeing that you do believe in?

Matthew Sullivan 25:48

Not really a jump, but uh my projections have the Celtics on top right now in the East. And uh I believe it. Uh as I was saying earlier, I I've never been a big Jalen Brown guy. They have a ton of talent that aren't household names. They're very deep. I think that they're going to be a tough team to beat in the East this year. And uh, as much as I don't like the way that Oklahoma City plays basketball uh in the West, it's it's hard to argue with them coming out of the West. That series, the playoff series with the Spurs was as invested as I've been in basketball in years, more so than the finals. But um I think that Oklahoma City's window still has a long time before it closes, and that's what I'm most excited about this year is you know the season starts off with the Spurs and the Thunder playing against each other again. I think that's gonna be a fantastic rivalry for years.

Russell Ries 26:51

I want to ask you just sort of a few thoughts on a few more teams, just to get a general sense of what you're thinking. Let's start with the reigning champion, New York Knicks.

Matthew Sullivan 27:02

I love their GM. I love how he put that team together. I didn't expect them to win the league last year, but I wasn't surprised either. You know, when we're talking about there being one ball, that's a that's a great example of putting together a team that uh that doesn't have a bunch of ball-centric guys. So Bridges and Ananobi, they don't need the ball in their hands to be uh successful. Jalen Brunson obviously had a fantastic playoffs. I think Carl Anthony Towns is still underrated. Uh I know that he's having some contract negotiations with the Knicks that aren't going very well right now. I don't think the Knicks win the championship without Cap. Uh I think that he probably matches up better with Wemby than anybody in the NBA at this point, other than maybe Jokic. So I like them. I like their starting five, but Josh Hart is is hurting. I know he's gonna try and play through it, and they're not a particularly deep team. They played the Spurs, and uh the Spurs just weren't ready to win it all yet. But the Knicks lack the depth that teams like the Celtics and the Spurs and the Thunder have, and I think that's the direction the NBA is heading. So I love their starting five. I don't like that Mitchell Robinson left in the offseason. I'd worry about the Knicks if anybody gets hurt.

Russell Ries 28:31

Fair enough. And let's talk a little bit about the Spurs. Obviously, they were the runners up, and Wimby is incredible, but there are a lot more names to that team, as you mentioned.

Matthew Sullivan 28:42

Yeah, brought in Tobias Harris, former UT vol in the offseason, uh, which I think is actually a really good fit for them. They didn't make huge changes other than that. They have maybe the best core three young players in a long time, with Wimby, with Stefan Castle, and with Dylan Harper. So, you know, the with the way that salary cap works in the NBA and the aprons, I'm not sure how long we can keep this team together. But if the Spurs can can keep that core, that's a scary team for years to come. And I personally enjoy watching them play. So while I admire the way that the Thunder put their team together, I hated, I hated watching them on TV last year.

Russell Ries 29:32

I think a lot of people felt that way about their approach to the game. I want to keep it a little closer to Nashville. Memphis Grizzlies are obviously in a bit of a transition right now, getting younger. And what are your thoughts on the Grizz and where they're headed? And do you like what they're doing?

Matthew Sullivan 29:50

I have I wish them the best. I think it's gonna be years before before we see uh a good Grizzlies team. I I don't necessarily think they're gonna be the worst team in the West this year, but they're gonna be pretty close to the bottom. And the big question mark for them is just gonna be how are they gonna get better? I know that a lot of fans wanted to see the NBA address the tanking issues of the past few years. And I appreciate the taking away the incentive for teams to be bad, but I don't see a clear path forward for the Grizzlies at this point.

Russell Ries 30:29

I hate to admit it, but I'm afraid I'm right there with you. All right, a few more rapid fire just because they've changed so much or have been involved in a lot of drama in the offseason. Let's start with the LA Clippers.

Matthew Sullivan 30:42

Kawhi, uh, when he plays as one of the best players in the NBA, even at his age, losing him and really not getting anything in return is is really gonna cripple them. Love Darius Garland, but there's just not a whole lot to get excited about over there right now. That's another team that's probably gonna kind of flounder for the next few years as they figure out who they're gonna build around or just dump the entire roster entirely. Uh, what about their roommates, LA Lakers? I'm not really big on the Lakers. I really like to watch Luca play, Austin Reeves. A lot of fun to watch as well. Uh, I think bringing in Kessler is is a great move for them. Uh, he's an elite rim protector. If he's healthy, he makes them better. But I was talking about the Knicks not having depth. The Knicks have a great starting five. I don't even know that the Lakers have even have a great starting five. It's Luca, it's uh it's Kessler, it's Reeves, and a whole lot of question marks. There's there's not a lot of depth on that team, and they're maybe as top-heavy a team as there is in the league at this point.

Russell Ries 31:56

You mentioned Tobias Harris, former Vol. Is there any future for Dalton Connect, another former Vol who's now with the Lakers?

Matthew Sullivan 32:04

Yeah, they seem to like him a lot, but I'm rooting for him, but I just I don't see it.

Russell Ries 32:10

Okay, well, gun to your head, season's starting soon. Who do you have? Sounds like Oklahoma City and San Antonio as your top two teams in the West?

Matthew Sullivan 32:20

Yes. Yeah, and my head says Oklahoma City, my heart says Spurs, and it really just depends on. I know that the Spurs beat the Thunder last year, but that was as tight a conference finals as I can remember in a long time. Those are really evenly matched teams. I like what Oklahoma City did in the offseason. I like their draft with uh bringing in Mara. I I think they're a toss-up at this point. Um I'm rooting for the Spurs, but I I I worry that the Thunder are gonna stick around.

Russell Ries 32:51

All right, and then the East, who do you like in the conference finals?

Matthew Sullivan 32:55

I like Boston. I think people are gonna be disappointed in 76ers. I'm not one for hot takes, but I honestly wouldn't be surprised if if uh they're not a play-in team, even. I know Detroit disappointed a lot of people last year, but Cade Cunningham was coming back from a serious injury. I think Boston and Detroit.

Russell Ries 33:20

I just want to touch a little bit more on the Sixers thing. Is it because they have so many mouths to feed? It is, and I I just don't know that they fit well together.

Matthew Sullivan 33:30

So um, you know, the Knicks I was talking about earlier, that's kind of a masterclass in putting together a team as far as as guys that have you know specific roles that that mesh well together on the court. Because think back to just a few years ago when we had the Lakers with LeBron and Anthony Davis, and they brought in Russell Westbrook. Yeah, Russell was on the tail end of his career at that point, but it wasn't just that that he was aging now. It's just he was just a terrible fit for a team that already had LeBron. Like who's gonna handle the ball? Those two players just didn't fit well together. I feel like the 76ers are doing something similar. Yeah, too many mouths to feed, but also I don't think they're gonna be a great defensive team. You know, they despite the fact that Paul George is aging, still a really good defender, uh, replace him with with Jalen Brown and LeBron. Uh, neither neither one of which is gonna bring as much on the defensive end as George did. If Joel L. Embiid is healthy and plays a full season, then you know they're automatically a contender, but uh I wouldn't bet my house on that.

Russell Ries 34:44

Yeah, that's a very big if. All right. So it sounds like Detroit and Boston. If you had to pick between the two, who would you have coming out of the East?

Matthew Sullivan 34:52

I would root for Detroit. I think Boston takes it. Um, and if I could pick three, uh, the Cavaliers, uh really like what Cleveland's doing too.

Russell Ries 35:04

Talk a little bit more about what's happening in Cleveland for folks who don't know.

Matthew Sullivan 35:08

They'll have Harden for a full season. They got Peyton Watson from Denver, who I think will play some serious minutes for them. And uh hard to tell how good he's going to be, but he looks like a promising young player. So, you know, returning Donovan Mitchell, Evan Mobley, Harden, Allen, uh, adding Watson to that mix, that's that's a pretty solid five. And um, my question mark there is still Harden and Mitchell playing together, but Harden has shown uh when he's on a stacked team with players that can score that he's happy to facilitate. For a player that scored as many points in his career as Harden has, I think people forget that he's been a league leader and assist or close to it in several seasons, if if that's the role his team needs him to play.

Russell Ries 35:58

That's a really great point. Yeah, I do think most people envision him shooting from you know half-court line as much as possible, throwing up his heat checks. Any particular rookies you're very excited about, and any you think are a little eh? I like the Michigan guys.

Matthew Sullivan 36:16

I think that Mara will have some opportunity to see the court with the Thunder. I think they drafted him specifically because of the rivalry that they have with the Spurs. He gives them some good length and six extra fouls for Wimby. But I I think that he's gonna develop into a pretty good NBA player. On the Warriors, I like Lindenberg, uh Jaxel Lindenberg. Most rookies don't have a chance to to put in serious minutes, but we've with the Warriors, you've got a case where Porzingis is not going to be playing at the beginning of the season. Who knows how long he's gonna be out? Jimmy Butler is gonna be out until at least 2027. There's a lot of opportunity there for Lindenberg, and uh yeah, he's an old rookie. I think that he's gonna see significant minutes with Golden State, and I think that he's got the talent to play well. I think he'll be a good addition for them while they heal up.

Russell Ries 37:15

All right, two final predictions from you. Sounds like you got Boston and the Thunder meeting in the finals. Who do you like there? And who do you think is gonna be league MVP?

Matthew Sullivan 37:26

If it's the Thunder and the Celtics, I go with the Thunder. I think this is the year that that Wemby wins MVP. You know, we we toss around the term generational talent a lot. I have never seen anything like Victor Wimboyama before. Um he's amazing, and the the idea that he could still get better is uh I think should be frightening for the rest of the league and thrilling for those of us that like to watch the game.

Russell Ries 37:55

Yeah, I couldn't agree more. Monster who can handle and shoot and do everything. He's I I do think that's a fair use of the term generational talent for sure. All right, Matt. Well, a year from now, when I talk to you and we're previewing the season next year, what additions or changes to the model do you hope to have implemented? What what's your vision for where you want to take the model?

Matthew Sullivan 38:19

So the biggest challenges right now are cost. And without getting like too deep into the tech industry, I host all of my cloud infrastructure on Azure, and I can't actually even pay for GPUs to do modeling for me. You know, I've had a request in to Azure for three months now to get even one machine with a GPU to help me model, and can't even get that. And that's just because of the big tech companies gobbling up GPUs to build LLMs. So, operations-wise, for me, that means moving back most of the calculations to physical machines here at my house, which is not something I thought that I would be doing in the year 2027. I'm excited to see what the simulations are doing on the NFL side. I've already integrated them with my NFL fantasy teams, and that's that's where I want to go with things that I'm building. Is it's cool to project what's going to happen, but what makes it really valuable is the ability to you know take the simulations and pick your starting lineup or draft a team based on the projections. And then going forward, I want to do intergenerational fantasy games. So uh I've already laid the groundwork to eventually be able to play games and or allow people to form teams that have, you know, Michael Jordan and Shaq playing alongside Steph Curry and LeBron uh for a basketball geek like me. That's uh super exciting.

Russell Ries 40:00

Right on. I really appreciate you coming back on to talk about the NBA season and and how the model works. I think it's really fascinating to hear and kind of get a peek under the hood. So thank you. Yeah, thank you. Thanks for having me, Russell. Anytime.

Weekly Recap

Russell Ries 40:16

And that was the conversation with Matt Sullivan. My thanks again to him for coming on. I feel like I always learn something when I talk to Matt. Now in this week's news. Last week the Titans lost 26-3 at home against the Houston Texans. That's all we're gonna say about that. And just this afternoon, they picked up their first win on the road at the Indianapolis Colts. The Titans ended up taking it 26-20 after being down 13-0 at the half. They are now 1-5 on the season, and next week, welcome the Cleveland Browns to town at noon. Hopefully, they can build on this momentum and pick up another win. In college football, Vanderbilt lost 31-21 to Arkansas yesterday. They are now 3-4 overall and 0-4 in the SEC. Next week, they are on the road at a suddenly feisty-looking Kentucky team. Tennessee State beat Morgan State 25-16, bringing their record to 2-5 on the year. Next week, they welcomed Charleston Southern. And Middle Tennessee State University beat Delaware 33-30 on Tuesday, bringing their record to 3-3 overall and 1-1 in Conference USA. They were off yesterday. And they are next at Florida International University on Tuesday evening at 6 p.m. The Predators had a busy homestand this week. Tuesday, they lost 5-2 to the Vegas Golden Knights. Then Thursday, they lost 3-2 to the Sharks in a shootout, so still picked up one point there. And finally, on Saturday, they won 5-2 over the St. Louis Blues. This coming week, the Brads are on the road at Boston on Thursday night at 6 p.m. And then at Pittsburgh on Saturday night at 6 p.m. as well, before welcoming the Colorado Avalanche into town next Sunday at 7 p.m. Vanderbilt Women's Soccer continued their excellent season, defeating Florida 3-0 on Thursday. They are now 13-0-1 and 6-0-1 in SEC play. They are currently playing Alabama, probably as this episode comes out. Finally, Nashville SC won against Visiting Sporting Kansas City on Wednesday night, 4-0, and then headed on the road on Saturday night, ending up with a 1-1 draw at San Jose. The draw clinches first place in the Eastern Conference for Nashville SC, and a win next Saturday at CF Montreal would officially claim the supporters shield for Nashville

Outro

Russell Ries 42:59

SC. Alright, that is our show for this week. Thank you so much for listening. My thanks again to Matt Sullivan. I really enjoy talking to him any chance I get. As always, you can reach out to us on social media. We're on Twitter and Blue Sky at NSH Sports Report. There's also a link in the description below that will allow you to email us directly, or you can find that same link on our website, www.natural sportsreport.com. My name is Russell Ries. I appreciate you taking the time to listen, and we'll talk to you again next week. Thanks.