Ken Pomeroy Explains KenPom Rankings and Smarter March Madness Bracket Picks

Knowledge at Wharton

Ask about Knowledge at Wharton

Episode Description

ABOUT THE EPISODE Heading into March Madness, Ken Pomeroy, college basketball statistician and founder of KenPom, joins Wharton Moneyball to break down the thinking behind his KenPom rankings – tempo-free efficiency ratings, the four factors (shooting, turnovers, rebounding, and free throws), and how he evaluates prediction accuracy and calibration across a full season. Plus, Eric, Shane, and Adi discuss what caught their eye in sports — from World Baseball Classic odds and preseason workload questions to tennis dominance and what makes today’s stars so statistically extraordinary.

ABOUT THE PODCAST Sports is a game of numbers. Wharton experts Eric Bradlow, Shane Jensen, Cade Massey, and Adi Wyner team up to tackle the world of sports, from current events to longstanding issues such as: What sports streaks are the most impressive? How do you rank the best players? Can athletes be compared across sports? Moneyball Highlights explains how decision-makers in the game can avoid the common mistakes and embrace the data. Episodes are recorded at the Wharton School.

#MarchMadness #CollegeBasketball #NCAATournament #SportsAnalytics #SportsBusiness ----- The Wharton School of the University of Pennsylvania is committed to sharing its intellectual capital through Knowledge at Wharton, the school’s online business analysis journal.

Launched in 1999, Knowledge at Wharton offers free access to:

- Articles, podcasts, and videos highlighting Wharton faculty research and analysis of current business trends - Interviews with book authors from Wharton School Press - In-depth series of curated content like Wharton Executive Education’s Nano Tools for Leaders - A searchable database of more than 10,000 articles covering all aspects of business - A weekly newsletter that delivers Knowledge at Wharton insights directly to your inbox

Visit the Knowledge at Wharton homepage: https://whr.tn/3qUpSBE Subscribe to the Knowledge at Wharton YouTube channel: https://whr.tn/3DEAgAU

Read the full transcriptHide the full transcript
  1. Welcome everyone to this week's edition of Wharton Moneyball. I'm Eric Bradlow, professor of marketing statistics and
  2. data science here at the Wharton School. Some combination of myself, my two colleagues are here today, Shane Jensen
  3. and Adi Wyner, both professors of statistics and data science, and Cade Massey are here every week on Wharton
  4. Moneyball here on the Wharton Podcast Network. Well, Adi and Shane, uh nothing changes when I say the best
  5. part of our, I guess I'll call it jobs, non-paying jobs, uh doing what doing Wharton Moneyball is interviewing people
  6. who have made seminal contributions to the applications of statistics and sports. Uh today's guest, Ken Pomeroy,
  7. many-time returning guest, is certainly no exception to that. Um just for our fans that don't know Ken, although I'm
  8. sure most of our fans on Wharton Moneyball do, uh Ken is a college basketball statistician. He's the
  9. founder of the analytics website KenPom. It publishes both tempo-based efficiency
  10. ratings and predictive rankings for every D1 team. So, Ken, on behalf of myself, Adi, and Shane, welcome back to
  11. Wharton Moneyball. Yeah, thanks, Eric. It's uh great to be back on. So, let's just start with the beginning.
  12. I always like to go back in time, and then we'll catch up, if you'd like to, 2026 in this year's NCAA tournament. Um
  13. for those people that don't know, when you first built the KenPom ratings and website, like usually someone does that
  14. because they see either a deficiency in what's being done out there, they see a better way to do something out there, or
  15. they just want to be, you know, build a following and a community. Um why did you build your KenPom system to begin
  16. with? Yeah, it was a a little bit of all of those things. Um when I, you know, when I first built it,
  17. it was the kind of the advent of uh uh you know, kind of random people logging on the internet.
  18. Uh the the was just starting to kind of mature, and uh uh certainly uh the baseball side of
  19. analytics was starting to explode, and there were just, you know, these people online who were getting into advanced
  20. baseball stats, and you know, kind of like outfoxing like traditional media in terms of analysis,
  21. and you know, I like baseball, but I really love basketball, and I really love college basketball, and I I looked and
  22. looked for the the version of that online for for months, and didn't find it. And so, that's kind of
  23. what uh motivated me to to start my site, and I never, you know, imagined it would get to to this point, but I I did
  24. think there would be, you know, a little bit of a following just because there was, you know, you were seeing that
  25. following develop for baseball. Mhm. Um what's What do you think, as you know, it's been uh so, a few seasons
  26. when the internet started to take off, has Kenpom, as you know, a site, as the statistics, has it been around Am I
  27. saying 25 years now? Is that about the right like it? Yeah, you know, when I uh first
  28. uh posted offensive and defensive uh efficiency for all the teams, that was at the end of the
  29. 2004 season. Uh Kenpom in some form, like the website existed before that, and I was doing
  30. college basketball ratings, kind of basic ratings, like you'd see uh you know, Jeff Sagarin do. He was He was
  31. really an inspiration for me uh growing up. But uh yeah, the kind of the the modern version started in 2004, and
  32. I've I've backdated, you know, some seasons before that. But yeah, it's, you know, we're we're definitely over 20
  33. years at this point. So, what would you say is the most, like, you know, if you had to be for a
  34. minute, you tend to be modest, if you were not modest, what would be you would say would be the biggest impact that
  35. your work has had? Like, what's been the biggest misconception the way that you've modeled things has either
  36. corrected, or what do you think you've added to, you know, as we sit here in 2026, what do you think you've added to
  37. the way in which, let's call it player or teams effectiveness or efficiency, or, you know, what are the
  38. sufficient statistics one should measure when one's thinking about team strength?
  39. How do you think about your contribution? I think in a in a really general sense,
  40. it's just about trying to account for the opportunities that teams or players have to do things. Um
  41. so, you know, the the biggest impact on the team level where, you know, we talk about possessions per game or pace or
  42. tempo or whatever you want to call it. Uh you know, in college basketball, there's such a wide range of styles in
  43. that regard that you can't just Traditionally, you know, when I started people just looked at good points scored
  44. per game and points allowed per game and attributed that to offense and defensive
  45. quality, but you really do have to look at points per possession and points allowed per possession uh to give you
  46. the the true picture of of which teams are most effective and um you know, people were thinking about
  47. that before I started my site, but I just to have kind of a an easy reference to see how all the teams did in those
  48. stats and you know, make it easy to compare. That's really I think what what my main contribution has been.
  49. I would think also and Audie I mean just Audie wants to jump in in just a second.
  50. I would also think that number of possessions, you know, and you know, big end teams that get lots of offensive
  51. rebounds, teams that push the ball more, uh teams that get turnovers. I would imagine that, you know, it's some, you
  52. know, as I always say, you know, I'll use my language of business here in the marketing department. What the hell, I'm
  53. sitting in my office in the marketing department. What do retailers care about? They care about making money per
  54. unit and they care about inventory turn. So, what do I care about in basketball?
  55. I care about getting a lot of possessions and being efficient on each of those possessions. Is that too simple
  56. a way to think about it? No, not at all. I uh No, that's it gets back to kind of
  57. the opportunity thing, right? Like that you know, that was one of the early discoveries, too, is you'd see these
  58. teams that you know, functionally you'd watch them and they they'd look really ugly
  59. offensively, you know, uh you wouldn't you wouldn't say they had an effective offense, but uh the trick was, you know,
  60. they were getting like a lot of offensive rebounds or maybe they weren't committing a lot of turnovers, you know,
  61. it's very easy to count things that do happen, but to notice things that don't happen, like not committing turnovers is
  62. something that humans don't necessarily do very well. So, you know, you you'd see like teams
  63. ranked pretty high in my offensive ratings and maybe didn't look pretty, you know, functionally, but it was
  64. because they were you know, just getting a lot more opportunities at shots like you're
  65. talking about. So, you since we're almost talking historically here. So, Dean Oliver's
  66. spoken to uh he's been on the show many times. He's come to speak to our our our
  67. programs here at at Wharton and Penn. And one of the things that he talks about is that when he first started
  68. doing basketball analytics and he's sort of the Bill James of basketball analytics if you will, he said that's
  69. the statistic of possessions didn't even exist. It wasn't even something that was
  70. ever recorded, right? And he almost had to had to take in basketball paper he talks about how he actually had to go
  71. and record and figure out like how you can figure out what the pace of play is cuz it wasn't there. So, uh scroll now
  72. to your work in in college basketball. How did you get to possessions? Is it now a stat or is it did you have to work
  73. that as as hard as he did to create that number? And then of course once you have
  74. possessions you can talk about efficiency. So, how did it come to be for in your in your perspective?
  75. Yeah, well first of all I you know, Dean was a a huge inspiration for for my work and his book, you know,
  76. his original book I think came out in 2002, which so it's not a coincidence that like my site started shortly after
  77. that. Like just the ideas he had really resonated with me. Um yeah, I yeah, so possessions aren't a published stat. Uh
  78. so, you can either count them uh specifically or you can estimate them from box score data and I still estimate
  79. them from box score data for for most things I do. Um that gets you really close enough to the number, but uh
  80. yeah, I mean there's no doubt that uh there were there were some hurdles, you know, early on to kind of getting things
  81. started and that was one of them, but like Dean kind of blazed the trail on that. So, it
  82. actually, you know, made it really easy for me to just, you know, follow follow his work which he helpfully laid out in
  83. his book. Do you do the four factors in college basketball? Is that part of your So, can
  84. you explain to our listeners what the four factors are, please? That'd be great.
  85. Yeah, so the the four factors are, you know, just a basic kind of the basic building blocks of offense or defense.
  86. Uh so, you have shooting, you have turnovers, you have offensive rebounds, and you have free throws. And uh the way
  87. those are measured again is by opportunity. So, you know, field goal percentage is not straight field goal
  88. percentage. It it's called effective field goal percentage. It accounts for the value of a made three-point shot.
  89. So, you know, if I'm shooting >> That's all these favorite math. Three is 50% more than two. Sure [laughter] is.
  90. In basketball. Yeah. Yeah. And you find teams that shoot a lot of three-pointers tend to have lower
  91. field goal percentages than teams that don't. But, obviously, there's a trade-off there. When they make their
  92. shots, they're, you know, getting more points out of it. So, so, there's that. You know, rebounding is offensive
  93. rebounding percentage. So, it's basically the percentage of times that you have a chance to get an offensive
  94. rebound, how often you do so. Turnovers is turnover percent percentage, so the percentage of possessions you get a
  95. turnover on. So, yeah, those are the four factors, and they're they're just great at
  96. you know, basic explanation of what a team does or doesn't do well offensively and defensively.
  97. >> [snorts] >> So, one of the things that I've noticed, just to jump in on the four factors, I
  98. actually teach this in my senior capstone. And one of the things that's remarkable,
  99. I'm not a basketball person by nature. I mean, I've certainly played enough basketball as a kid and in high school.
  100. But, I don't really study it. It's one of my weakest sports, I think, in terms of the majors.
  101. But, I so I've I've come to learn about basketball through statistics, which is kind of odd, right? But, those four
  102. factors are oddly uncorrelated, which is shocking to me. And at least at the professional level.
  103. Is that Is that Can you explain that? And is it true at collegiate level, too? I would imagine the collegiate level's
  104. got to be more correlated, cuz good is good. And when you have tremendous variation, which you do at in the NCAA,
  105. you probably your best teams are probably better in some level. But, I like to >> And just before you answer that, Ken, I
  106. just want to be clear. What Audi may be referring to or not, correct me, correct him.
  107. At the individual player level, this is a classic aggregation issue I would I could imagine.
  108. If I'm a good three-point shooter, I may also be a very good free throw shooter,
  109. but at the aggregate level, at the team level, they may be uncorrelated when I look across teams. So, if you could I'm
  110. just modifying Audi's question. I just want to know if this uncorrelatedness is at the player level, the team level when
  111. we aggregate, or is that not true at all? And then then I'll take Curtis mean. Uh yeah, that's a lot to ponder. I
  112. I'll It's It's more I I would say it's more uncorrelated at the team level than the
  113. player level. Um so, Audi made a good point that, you know, in college, the distribution of talent is much greater
  114. than it is in in the NBA, right? It's the the difference in the best and worst team is enormous in college. And so,
  115. yeah, so you do tend to see that like teams that are good in one thing are going to be good in other things. But,
  116. uh but there's no like there's no reason for you to be good at shooting and also
  117. good at offensive rebounding, right? Like those two things really aren't that related fundamentally skills. And in
  118. fact, you see like almost no teams are great at every single four factor. Like most teams are great at even the great
  119. teams. They might be great at like four of the eight four factors if you consider both sides of the ball. Um and
  120. they might be good at two others, and they might be like below average at two others. Like nobody's great at
  121. everything. So, teams have all sorts of different styles, and they specialize in
  122. different things, and uh and yeah, the fact of the matter is like the way those four factors are defined, they're
  123. basically orthogonal. Like they're not the stats don't bleed into to other four factors. So, you can be great at one
  124. thing and poor at another, and it's not unusual to see very good teams behave that way. Yeah, well, I guess you're uh
  125. just to kind of follow up, you're talking about kind of specialization at the team level. I when when kind of I
  126. thought about what might drive a lack of correlation, uh it would be more specialization at the individual level.
  127. And I know like cuz I I think about baseball. Obviously in baseball it's very there's very much of a you know,
  128. like fielding even like fielding and hitting are not are often negatively correlated because there's a real
  129. specialization there. Basketball obviously there's not as much positional specialization, but is is
  130. that a large part of what's driving maybe a lack of correlation? Yeah, I mean there's there is some
  131. specialization like often times you if you look at a you know, a team so first of all like in basketball we really
  132. struggle to measure measure individual defense, right? We don't have really many good stats for that. You know,
  133. block shots are good for for tall guys, but for small guys you know, they're not going to block
  134. shots and they still can be effective defenders. Um so often times when you look at like uh season stats and you see
  135. like a player who doesn't contribute much offensively like it's almost certain that they're a good defender.
  136. Like that's why they're on the court. So you do see that. I think another like another great example of aggregation is
  137. that when you compare offensive rebounding percentage and defensive rebounding percentage for a team and and
  138. look at all of division one, there really is hardly any correlation between those two. So if I tell you that a team
  139. is a great offensive rebounding team, I really haven't told you anything about whether they're a good defensive
  140. rebounding team. And yet on the player level there obviously is a high correlation there, right? Like a good
  141. offensive rebounder individually is probably a good offensive rebound or a good defensive rebounder as well. So
  142. that's to me that's the best example of that effect you guys are talking about. Let me ask you the following question.
  143. Let's imagine I took just those four factors and I use those as a predictive model to
  144. predict the outcome of winning versus something more sophisticated that I'm sure you done and built over the last 22
  145. 23 plus years. How much do we lose? Like are those sufficient statistics? I don't
  146. mean literally, but quasi sufficient for predicting game outcomes? And if the answer's yes,
  147. great. If not, what else is kind of missing? No, there's not there's not much missing. Like if all you have are are
  148. box score stats, those four factors explain like you know, 99% of offense and defense. The
  149. only thing that's missing are like data errors and uh you know, the possession estimation
  150. and things like that. But yeah, I mean for the most part, you know, yeah, you're going to find you
  151. plug those guys into a regression or whatever and you know, obviously like shooting's going to dominate. It's going
  152. to be the most important thing and then offensive rebounding and turnovers are somewhere like tied for a decent second
  153. and then like free throws are like you know, the weakest thing. But yeah, you you you'd have a really nice model
  154. if you if you use those four things. Can I uh let me just jump in cuz actually I
  155. have this data for the NBA, so not the collegiate. So if you try to predict wins using
  156. point differential, that's about as best as you can get. I mean the game is down
  157. to points, right? So yeah, you can end up have a lot of blowouts and your your point differential might not that be
  158. perfectly accurate with your wins. You can it's it that's you can't expect perfection. But four factors works
  159. almost as good. It's just That's the point in the >> It's it really is amazing how how
  160. I mean there's leftover residual variance because of how you distribute your points. But you're and I'm talking
  161. about season-wide four factors are and I just use four. I'm but I use the differential. So the way I use I can
  162. create a delta four factor for each of them and instead of having eight, you know, four on each side. So
  163. I'm sure I'd have even better if I did if I went down to that. It's just amazing. I'm very excited
  164. about this and I when I teach this I get excited about my students usually look at me and I'm like, why do you care? But
  165. it's because they don't have experience and they don't know how valuable that is.
  166. >> a dream regression scenario you where you've got like four different factors all which are
  167. important but not particularly highly correlated with each other. Yeah. Well, I was just going to say I
  168. mean that's you know, that's why Dean Oliver chose those. That was really intentional the way he
  169. you know, it's pretty obvious like when you think about it, it's obvious like those things should be the four factors.
  170. But just the way they're they're measured and the fact that they're you know, again, they're like the
  171. measurements are completely orthogonal, so they're not you know, directly related to each other in any way.
  172. So let me ask you a follow-up to that. So I in the when I introduced you, I I read and predictive rankings for every
  173. D1 team. So, how do you I mean, obviously you have a ranking vector. Obviously we get
  174. outcomes at the end of the season, which, you know, get the NCAA tournament, but how do you score
  175. yourself? Like, anybody that does predictions always has, let's call it an error metric of some sort and says, "How
  176. well did I do?" And then possibly that leads in good models to model modification if you notice some sort of
  177. systematic errors or stuff. So, I So, how do you score how well you've done in ranking teams? And how does that lead,
  178. well, if you notice something to any type of modification of how you do things? Right. So, the the main way I score
  179. myself is by looking at my game predictions. Um, yeah, so I do look at the overall
  180. ranking as well, but primarily I I use that to like judge my I have a preseason ranking as well, which is
  181. completely separate calculation, but they're really like judged on, you know, compared to the final ranking. But,
  182. overall like the algorithm, I yeah, I basically focus on game prediction. So, both uh, you know, predicted uh, point
  183. differential and uh, win probability. And, you know, just a basic How is your model How is your model We always like
  184. to ask this question. Is your model properly calibrated? Which means when you just I want to make sure our
  185. listeners understand, but they do. Um, when the KenPom system says team A has a 70% chance, if you bucket all of those
  186. around 70%, is it near 70%? If you bucket the 20% ones, is it around 20%? So, what do you find there and do you
  187. have the classic problem Shane talks about this all the time? Do you have problems out in the tails, which is
  188. typically where we find, you know, models don't do as well out in the extremes? Yeah, so it is pretty reliable. Um,
  189. yeah, I have like I have a page for those for those stats that updates in in real time basically. And I'm probably
  190. the only probably the only one that looks at it, but you know, I'm always curious. Uh, Uh yeah, it it does really
  191. well. In the extremes, it probably um under predicts the favorite slightly. Like I think this year there were like
  192. 260-ish games where I gave the underdog like less than a 2% chance of winning. And they won zero of those games.
  193. And they should have probably won like three or something, you know, based on the probabilities. So, there's a little
  194. bit of uh overconfidence, I guess, with with extreme underdogs. But outside of that, like for the most
  195. part, it does pretty well. Certainly does pretty well in the probabilities. It has you know, it has another issue
  196. with like scoring margin for extreme underdogs as well. It's constantly like overconfident on these just really super
  197. lopsided matchups. Um but Overconfident in in favor of the underdog. Exactly, yeah. So, like, you
  198. know, when the market says a team's favored by 40, like I have them favored by like 32 or something. And Well,
  199. Oddie's all for that. Oddie's all for regressing back toward some average. Well, I mean, let me let me just point
  200. out a couple things. Um first with response to that, this is something I also do with my class.
  201. There's a sort of the big home dog effect in betting, um which is that when a team is playing
  202. at home and they're the big underdog, they have a tendency historically to cover the spread.
  203. And uh fascinating because that goes back to the '70s. And I I look at this historical data in football in
  204. particular where I have that data. And it just in every 5-year period, it just sticks out, sticks out, sticks out. And
  205. it goes all the way through like 2015 and then it stops. It's just sort of gone. As if And that may be because
  206. people are starting to bet it more more um uh globally. Hard to know exactly. But that's that's
  207. potentially what you're seeing in basketball is that the model, because you're not really talking about betting,
  208. the model says 40 points, but we know that home team's going to outplay there when they're the big dog. Maybe you're
  209. seeing that. But my question is, um Um, way back in the way In fact, you were one of our earliest guests in in in in
  210. what in what Moneyball. Um, we were trying to figure out what some of the secret sauces are in NCAA tournament
  211. prediction. And I have to thank you right away because, you know, I I only play in a small family tournament, but
  212. I've won 40 per 40 four out of five of them. Um, and I just use your your inside information. So, I should send
  213. you send you a check, but if only for you for a couple bucks. Just to get that out there. So, thank you, Ken. But, the
  214. question, um, this is I remember Nate Silver pointing this out. Um, he used to say some of the secret sauce was
  215. shrinkage or regression towards preseason rankings. And that um, you tended to over overvalue the people the
  216. teams that had really good seasons. And um, and one way to to account for that was just regress them slightly to
  217. their preseason expectations. Is that still true? Yeah, it is. You do see that effect. Um,
  218. so, you know, in my system, I you know, my preseason ratings are pretty much washed out by now and like
  219. I really wanted maximum accuracy. They They wouldn't, but there's sort of a perception issue with, you know,
  220. the fact that my ratings are used on the NCAA team sheets for selection and things like that that, you know, you
  221. probably don't want preseason ratings influencing them, but um, that is a factor. I mean, one of the
  222. one of my favorite stats, which may or may not come into play this year, but uh, teams that are teams that
  223. end up being one or two seeds in the NCAA tournament, uh, if they were not ranked in the preseason AP poll,
  224. uh, there's like 38-ish, I think, of these cases over the past 40 years, none of them have made the
  225. final four. Wow. Which you normally expect these are one or two seeds, right? You normally expect
  226. like eight or nine of them to make the final four, but so, it's like definitely a
  227. significant trend and like I said, I don't think it'll happen this year. Nebraska's the only possible case if
  228. they end up as a two seed, but uh, that's, you know, really an example where like, yeah, clearly like teams
  229. have like overshot their preseason expectations and it it does, you know, you do see some regression when it comes
  230. to the NCAA tournament. With strength of schedule kind of is is that part of that story as
  231. well or is it more just um what what do you think is is is kind of driving that? No, I think it's just
  232. what we're we're talking about, right? Like you you know if you had you know, you want to get perfect
  233. information on a team, right? Going into the tournament. And perfect information
  234. is not solely based on what they did that season. Like the NCAA basketball season short, you know, it's only 30 33
  235. games before you get to the tournament. A lot of those games are mismatches, you
  236. know. Yeah, I guess that's what I was going with with kind of very uneven schedule,
  237. I guess, yeah. Yeah, so you know, you maybe have like 20 to 25 games of meaningful information and uh
  238. there's still meaningful information in the preseason, Paul. So yeah, you definitely uh you know, want to include
  239. that when you're making projections. So can I haven't been following it as closely as I normally do. I don't know
  240. why cuz I love college basketball. There is an undefeated team, is there not? Yeah, it's really one of the best
  241. stories of of all time. Tell her this So for those of us that don't follow, like
  242. what the what conference are they in? How likely are they to win their conference? And how far do you project
  243. them to go in the NCAA tournament? I would imagine not very far, but what do you see for Miami of Ohio? Yeah, so
  244. they're in the Mid-American Conference and uh you know, notoriously didn't play a very difficult non-conference
  245. schedule. Like most teams from weaker conferences, they end up playing you know, one of the top teams on the road
  246. just because they actually get, you know, a a nice check from that team for doing so. Um but Miami of Ohio did not
  247. play any of those teams. Debatable whether that was by choice or not. Like they'll tell you they were
  248. they were not able to schedule those teams, you know, under the terms they wanted. But uh anyway, they ended up,
  249. you know, running through their entire schedule undefeated, which uh you you know, people are saying it's
  250. entirely schedule-based, but still really hard to do that even playing a weak schedule.
  251. They're ranked 90th in my predictive rating, so uh Yeah, so you know, if they played in a
  252. power league, they probably would have taken quite a few beatings over the course of the season playing the way
  253. they did. Like you look at teams at the bottom of the Big East, right? And they're ranked in the 90s. Like, you
  254. know, that's about what what they'd equate to. They're oddly like playing getting ready to play their conference
  255. tournament. They're not even favored to win their conference tournament as an undefeated team. Akron lost one game in
  256. conference play to Miami at Miami, but overall has, you know, better point differential metrics and that's kind of
  257. carrying the day in in these predictions. But So, if they go undefeated going into the tournament,
  258. are they like >> a I I don't see how they they have to be a top eight seed, right? If they go
  259. undefeated. I don't even know if they're guaranteed to make the tournament. >> if they win their conference tournament,
  260. by definition, they have to cuz there's an automatic bid. But what Ken's saying is if they lose, let's say in the
  261. semi-finals or something of their they could be whatever 33 and one and be left out. Yeah, so if they they it's very
  262. unlikely they'd be left out based on what we know about how teams are selected and the metrics that are being
  263. used now. Um so they're going to get in. But yeah, seeding-wise, even if they win their
  264. conference tournament, I eight would be pretty optimistic, I feel like. They could be an eight, but yeah,
  265. probably more in that 9 10 11 range. Just because their schedule I mean, they have one of the five weakest schedules
  266. in the country, which just put it this way, like no team that's earned an at-large selection has ever had a
  267. schedule remotely that weak. So, they are they're just like such a wild team. I mean, it's first of all, amazing they
  268. went undefeated. They had so many close calls and every time, you know, they pull it out. Um but they've also like,
  269. you know, have this kind of Frankenstein-like profile that nobody has ever seen before. It's just a it's
  270. just a wild wild story. So, let me ask you another question about the tournament. One of the like if
  271. you think about the whether it's the four factors and that each of the teams having four factors or maybe it's eight
  272. factors, whatever it is. Um one of the things you hear all the time at the time of the tournament is matchups.
  273. Do matchups matter that much? Like I tend to think like they're talked they make great stories.
  274. And they're over they're talked about a lot. But I you know, whether you want to
  275. call them interaction effects, if you got like the you know, the loyal and maramounts against the slow down teams,
  276. someone's got to win and pose their will on the style of the game. How much do matchups matter or when you're making
  277. predictions, are there matchup variables in there that move the predictions what
  278. we would call a significant effect size? Right. So, there's Yeah, in my predictions there's no matchup effects.
  279. Like historically when I've looked at this or other people have looked at this, it's been extremely challenging to
  280. find any like specific matchup effects where for instance like you know, a good offensive rebounding team is playing a
  281. bad defensive rebounding team. Like might there be some sort of extra advantage there like that's an example.
  282. But the trying to find those like combinations has been pretty much impossible. Like you watch
  283. if you watch the tournament and watch the commentary, like almost everybody will talk about matchups. They'll talk
  284. about that pace effect that you're talking about when you have two contrasting paces and well, it's going
  285. to be really important that one team controls the pace in this game. But man, when you look at it historically,
  286. there's almost nothing to that and sometimes I just feel like you know, analysts feel like they sound
  287. smart talking about matchups and stuff. But in the reality is like it's just it's I'm not saying that those
  288. effects don't exist, but it's been a very challenging to tease that out of the data. By the way, I we as at Wharton
  289. Moneyball, myself, Audi, Shane, Tate is not here today, but he we all appreciate
  290. this. I we all like the way you phrased it. You didn't say these effects don't exist. You just said it's hard to detect
  291. them. It either means a couple things. One is there's not tons and tons of data to do so or the effects are small and
  292. therefore it's hard to detect small effects. And so if they exist, they might exist on a
  293. you know, I always call it a second or third effect size level as opposed to a primary one. If I can interject, it's
  294. like it could be also that these effects are are kind of real at that season that
  295. time, but they're ephemeral. They're not like consistent over year to year that like they'd show up kind of
  296. in a historical analysis. This discussion's kind of making me think about how we retrospectively every year
  297. in baseball we're like, "Oh, well, relief pitching is what wins in playoffs." Because, you know, that's
  298. what you know, won the previous year and stuff like that. I think there's a lot of narrative forming that after
  299. retrospective narrative forming. Because again, as again, as your point yeah, that you know, the people we're
  300. analyzing it, especially analyzing in real time have to talk about something. Uh and I I kind of wonder how much of
  301. these are sort of us we we kind of wrap up particular analysis or narrative into
  302. what's currently happening, how consistent that is you know, year to year, next year, 10
  303. years from now. >> Good point. >> I don't I don't think there's much to that. Yeah, I mean, you like the one thing for
  304. me is, you know, I'll I'll talk to like coaches for like pretty successful coaches and they'll they'll mention, you
  305. know, "Oh, this game's a favorable match-up for us." And they'll explain why and so, you know, you you have to
  306. respect that knowledge. And I think possibly like to the extent match-ups exist, you know, it's it's probably one
  307. of those things that swings like your expected outcome by a point or two or something like that. Like so, that's you
  308. know, that's going to be something that's really hard to detect in in specific situations. But uh
  309. I guess the the key point is you know, match-ups don't turn a 10-point underdog into a favorite. Like I think we can
  310. safely say that. Right. Can you tell us so, before I ask uh a last few questions
  311. about the future and what you're working on now, um what do you see happening in
  312. this year's NCAA tournament? Like are there a couple teams you really like based on your model? Who are those
  313. teams? Who is doing better than let's say either the betting markets or you know, either coaches rankings? Uh how do
  314. you see things going this year? Well, uh the story is really going to be about the the top teams. Like last year
  315. we had a notoriously uh upset-free almost tournament. Uh the top teams were rated exceptionally strong
  316. heading into the tournament and it turned out that it played out that way. Like the top
  317. eight teams in my rating heading into the tournament ended up in the Elite Eight and the top four teams ended up in
  318. the Final Four. So, it was kind of a predictable tournament from that standpoint and
  319. it's a similar structure this year. Like the top four teams are are really strong. You know, Duke, Arizona,
  320. Michigan, and Florida are separating themselves from the rest of the country. I doubt all four of
  321. those teams get to the Final Four. Like I feel like last year was an anomaly on some level, but we are seeing the
  322. structure of college basketball change a bit where there is more separation now between the
  323. teams at the you know, at the top of the rankings and some of the teams that we >> Why do you think that is? Why do you
  324. think there's a more separation? So, really in recent years uh player rules about player movement have
  325. been relaxed uh tremendously and you know, it used to be it was difficult for players to to change programs. Uh often
  326. times they'd have to sit out a year which would discourage them from doing that, but those restrictions have been
  327. removed and so now uh teams can aggregate talent really quickly. You know, if they get a star player on their
  328. roster in the off season, well like that might attract three or four other really
  329. good players from other schools to join them. And so, I think the aggregation of
  330. talent is just more efficient at these top programs and it uh just makes them uh you know, able to produce more
  331. dominant teams than than we used to see. So, a lot of this a lot of obviously the
  332. loosening of the transfer uh rules is a big thing, but also the NIL money. I mean, it's uh the top college players, I
  333. mean, they only a few of them go to the to the NBA and because because there's just aren't
  334. that many openings in the NBA. It it seems that with collegiate money you can make almost a career or at least
  335. you want to stay in the NBA in a way that you never used to. And it just that have any um direct or important impact
  336. on the fact that now the top four, eight, even really are substantially better than the rest of the field
  337. because of that. Is Is there Are you detecting that? Yeah, I think you can say that. Like
  338. certainly still for the top players it makes sense to to go pro. I mean ultimately when you get your second NBA
  339. contract you're going to make more money than you can make in college. So you want to get that
  340. clock started. Yeah, but certainly for like the top, you know, five to 10 players like
  341. they're almost certainly Um but beyond that, you know, like Michigan has Yaxel Landeborg, right? He
  342. could have been a first round draft pick last year but he chose to stay in college for another year and presumably
  343. got, you know, compensated pretty well by Michigan. Moussa Johnson, another player on Michigan who um
  344. you know, kind of falls into that boat. Uh so yeah, so I'd say I'd say for like players outside the top 15 or so, yeah,
  345. they're more likely to to stick around than they used to be and uh there's no doubt college basketball more
  346. talented than it's ever been. Throw in the fact by the way that international players now who have like professional
  347. experience are allowed to play college basketball. G League players who are not who have not signed a contract with an
  348. NBA team are allowed to play college basketball. So uh everybody's looking for a competitive advantage in finding
  349. these guys and it's it's uh you know, made made the college talent pool, you know, better than it's ever been.
  350. So let me just ask you a couple uh complete conclusion questions if you'd like. Um what has changed in your mind?
  351. Like what do you think you've learned about college basketball analytics over the last 20 years? Like what has been
  352. the biggest learning for you? Yeah, that's a good question. I There's obviously a lot of things and uh I'm
  353. blanking on uh on all of them but uh Yeah, I mean, you know, the main thing is like,
  354. you know, there was a time when I started where it was kind of fun to, you know,
  355. mock conventional wisdom and you know, you learn over time that like conventional wisdom is
  356. you know, right more often than not. Like it's not always right. And there are certainly things that I feel like
  357. biases that coaches still have. But over time like as you know, a new generation
  358. of coaches have have come along and kind of grown up with this stuff. You see some of these
  359. you know, old school fallacies go away. But you know, there's no question like there are things that that coaches do
  360. and and that you know, are on gut instinct or whatever that you know, when you start to analyze them, you might
  361. think they're incorrect. But you know, obviously there's there's some wisdom there that that coaches have
  362. acquired over the years even if it's just ignoring data and and and using their gut instinct.
  363. So, let me ask you um one last question. So, I always talk about the following uh
  364. uh two things. Um it's a kind of a two-part question. Um and Shane and Audie are going to love
  365. this cuz I'm going to use my favorite word is momentum. How much do you think momentum
  366. matters going into the NCAA tournament? Like let's imagine Duke for example, I'll make it up, loses in the second
  367. round of the ACC tournament or Michigan loses in the first or second round of the Big Ten tournament. Does that change
  368. anything for you? Now, of course, we have to It has to do something. In other words, it is recent more than past. We
  369. all Anybody's model counts the current power more than the past. But are you a big believer in momentum or that doesn't
  370. mean that much to you? Oh, I think uh once you uh account for what you're talking about, like
  371. obviously there is more value in more recent data points. Um once you account for that, like
  372. beyond that, I I don't have any particular uh use for momentum. I mean, that's one thing
  373. You've got You're You're joining the Shane and Audie camp. They don't believe in momentum, too. Once you account for
  374. all these other factors, of course. But yes. Yeah, so I mean, there's obviously you
  375. Yeah. Um I mean, there's all sorts of examples, you know, of momentum fooling you. Like momentum
  376. exists until it doesn't, basically. You know, it's it's not very predictable on a a player or team level. Uh you know, I
  377. always try to make the point that like all data matters, okay? Like the games at the beginning of the season matter.
  378. Like there's it's kind of becoming vogue now for people to do analysis and really
  379. like split the season into like small samples and make teams look favorable, you know,
  380. they'll find that that a team played a terrible game like on January 12th and it's like, well, since January 12th,
  381. this has been the best team in the country, you know, and it's like, well, that game on January 12th actually
  382. mattered and you do need to consider >> [laughter] >> you need to consider it. And uh so
  383. so I would be like obviously teams that are playing better now than they played in November, that matters. Like they're
  384. they're a different team in that sense, but still those games in November uh are still data points that uh should be
  385. considered and uh you know, they they are considered in my model. Yeah, Shane and I and probably 99% of
  386. the rest of the planet would rather call it non-stationarity. Maybe they're just
  387. a better team or playing better. Has nothing to do with momentum. This should These are still college kids. I mean,
  388. maybe they're not the college kids that we used to think of in the back in the day because they're so professionalized,
  389. which means that the learning curve has got to be still fairly steep for young folks. So, there's nothing to prevent
  390. that the idea that momentum is really learning and some teams are getting better because they're working to work
  391. together better together better. And I think in basketball in particular, there's a strong sense of of um I don't
  392. want to call it esprit de corps, but I don't know. When two players work together, their team works together,
  393. they get better at anticipating each other. And I can imagine that momentum really appears that way um in so far as
  394. it is non-stationary. Maybe it's definitional, yeah, cuz I like everything you describe, I would just
  395. describe classify as non-stationarity. Yes. And momentum is more like the psychological
  396. >> like like like non-stationarity is like a a a in some latent process that's, you
  397. know, of of how they're playing the game. Momentum is like some extra juice because of the outcome at the outcome
  398. level. They keep winning and therefore I don't know I don't know if that's a way of
  399. deconfounding these two ideas, but >> one more I agree longer term and state-like latent state-like and I
  400. consider the other one localized and based on outcomes. I think that's fair. >> correlation in the recent recent
  401. residual is how you define momentum or something like that. >> I would agree. Well, Ken
  402. on behalf of myself, Adi Wyner, Shane Jensen, as always, we'd like to thank you for joining us here on Wharton
  403. Moneyball. You can catch Ken on his website kenpalm.com. He's again college basketball statistician and the founder
  404. of Analytics website kenpalm.com. Ken, thank you again for joining us this week on Wharton Moneyball.
  405. All right, thanks everybody. Appreciate it. Welcome back to Wharton Moneyball here
  406. on the Wharton Podcast Network. This is Eric Bradlow, Professor of Marketing, Statistics and Data Science and I'm
  407. joined today by my two co-hosts, Adi Wyner and Shane Jensen, both professors of Statistics and Data Science. Some
  408. combination of the three of us and Cade Massey are here every week on Wharton Moneyball. So guys, we just finished
  409. with Ken Pomeroy talking about college basketball. Obviously, there's a lot of else lot of else going on in the
  410. statistics and sports world. So we'll do our normal second half of the show where
  411. we say what caught your eye. Shane, I'll start with you. Anything in particular?
  412. Prob- It's probably going to be about the World Baseball Classic, but either way
  413. >> It is. It is. Yeah, I woke up I woke up a little bit early this morning and tuned tuned into the end of the
  414. Chechnya-Japan match going on this morning. The starting pitcher for Chechnya, this guy Andrei Satoria,
  415. held Japan scoreless for 4.2 innings pitched. This guy was already I I I knew I'd remember this guy. This guy in 2023
  416. the World Baseball Classic struck out like Shohei and a bunch of Japanese pitchers as well. I just think it's kind
  417. of notable cuz he tops out at 80 mph. So, he struck out like he went through the
  418. Team Japan lineup for almost like a five innings. >> he change speeds or does he just have
  419. good location? Changeup and I you know, I yeah, he's got a changeup. He's got trickery. He's
  420. got trickery. He's like I don't know. I don't know what what the analog would be, but it's pretty no I mean
  421. really what's driving this is unfamiliarity, right? This guy pops out of nowhere for every four three years,
  422. but but I think it's just notable. It is pretty fantastic little story that this
  423. guy can kind of has dominated Japan twice now. Um you know, and obviously it's not not not
  424. something where he could go in and pitch in the I don't think he's really a major
  425. league prospect despite uh Well, as you're pointing out and then I want to say one other thing about the baseball
  426. classic before turning it over to Audi. Um you pointed out something. Are there some pitchers that could be effective
  427. once, maybe two starts in the major leagues? Yeah, cuz they're not used to you. They're not familiar with you. But
  428. over a 162 game season, over 30 starts, everything, especially now in the world of AI, everyone's going to track
  429. everything, everything's going to be motion, they're going to know all the pitches you throw, they're going to know
  430. the locations and the zones you throw. You would think that might be done in the world of baseball classic also, but
  431. maybe not as much as it would be over a long season with >> No, yeah, I mean I I mean even before
  432. AI, I think the hitters had RI, real intelligence, where they would just kind of keep like like experience the same
  433. pitcher many many times and kind of I think there's a familiarity that even like pre-analytics
  434. meant that I think every pitcher coming into the major leagues would have that kind of curve of their stuff probably
  435. seeming unfamiliar at first and then hitters adjusting to it and the kind of back and forth. I just When you only
  436. When you top out at 80 miles per hour, the sub straight to kind of have that back and forth and keep the hitters
  437. guessing, I think is probably limited over like a 162 game season. That probably is something you can only pull
  438. off in a in a in a short tournament. But it's still an amazing story. The USA right now has, according to the betting
  439. odds, a 50% chance of winning. I'm ignoring the vig a little bit, but they're about a minus 105. Isn't that
  440. nuts? >> That's nuts. Absolutely nuts. I mean, it's sort of like I mean, you posted
  441. some odds too of like kind of like the um kind of the odds for like a Major League
  442. for the World Series The World Series odds going are also nuts. I think it's like some kind of weird like
  443. unfamiliarity bias or something like that. Like the fact that you know, Japan I I mean, you know, the
  444. US has only won like like one out of four of these things that's even been held. And I don't even know if they're,
  445. you know, Japan may be a bit more kind of well-put-together team. I I'm not even sure like they they they
  446. probably should be favored. But, you know, to kind of give them half the probability is
  447. >> Minus 105. Not that that >> To give them half the probability is disrespectful.
  448. It's disrespectful to Japan and to the Dominican Republic and all the other teams that could kind of in a one-game
  449. playoff >> A one-game? Isn't it one game? Isn't it one Well, yeah, each game each
  450. round is one game. Yeah. Oh, come on. Then Right. Right. I mean, you know, it's like I might make an argument to
  451. the following. The Jets could go on a run. I mean, they're out now, but exaggerate this, but to say
  452. I might put the 2025-24 Los Angeles Dodgers in there, and I'm not sure you should take them against
  453. the rest of the field in a one-game playoff at this round. Maybe. Maybe that would be fine to do so. But just think
  454. about if there's, you know, if there's a quarterfinal round, semifinal round, final round, they have to be at over 80%
  455. probability to win each of those games, which maybe they are cuz of, you know, But I don't know. Are the Well, that
  456. makes a the I would have the Japan team and half of the American team, you put them together
  457. and you do have the Dodgers. >> is that's what I was going to say. All right, so never mind. That was my
  458. [laughter] question. I mean, one of the things that makes it tricky is how much of the pitchers are
  459. going to go, right? I mean, schemes >> Yeah. and uh schemes >> pitched like one game against Great
  460. Britain and he's out. Yeah, oh well, that and he's now rethinking it. I mean, the thing is this is all preseason for
  461. the American team and and the Dominican I mean, and it's so it's hard to really know. I'm I
  462. think one of Mine is one of our Maybe the betting odds, but they cannot be the probability. No way. That is sounds like
  463. a betting opportunity in my in my estimate. And I don't usually you know, jump towards identifying
  464. those, but I it really sounds like a betting Let's think about We always talk about some batters, pitchers get off to
  465. slow starts. If we track these players through the first whatever 40, 50 games of the Major League of season this
  466. season. What expectation do you have? Do these people get off to a faster start than normal?
  467. >> That's what I would guess, but do you think that's actually going to be true?
  468. Yeah, but I on the other side of it like you'd also want to track like longevity cuz they're
  469. playing like more intense like, you know, like you know, I mean, Aaron Judge is probably going to have an amazing
  470. season again, but he's he's getting in like very like kind of regular season at-bats right now and that probably that
  471. to me says somebody like him is even more likely to hit the ground running have like an incredible sort of sprint
  472. like May, June. But come September, October like does that, you know, cuz already wear and tear we can see
  473. affecting players when they play up the regular season. Does this have any kind of influence on that?
  474. >> It's always exciting to see all these players come play for different teams, but I just want to point out It's nice
  475. to be able to Aaron Judge cheer for Aaron Judge Yeah, well, Aaron Aaron Judge of course is is the world's
  476. greatest hitter right now, but Bobby Witt Jr. That guy can field. Yeah, and all that and defense. It's [laughter]
  477. incredible. Yeah, yeah. It's like like Judge, you're like he's the best like you have to I mean, he's the best in the
  478. game, but you know, But Bobby Witt, not all that and defense. >> bad for the guy to have to have to be
  479. playing in the same league and the same prime as Aaron Judge and comes from MVP voting, but it just to
  480. you just really see it. I mean, I have to say some of our earliest work Shane my probably my first paper published
  481. paper in baseball was a paper we wrote years ago. The field and one that's still my favorite paper. Yeah. And and
  482. you just sometimes the eye test I mean, I wish we we we knew like the the probability that ball being and he did
  483. two of them being caught, but it's just so it's remarkable how he how he got to the ball then managed to sort of leap up
  484. in a sort of acrobatic motion. He's on his feet and he's throwing 90 mile an hour first base. I mean, WHAT?
  485. >> [laughter] >> JUST JUST it's just it's it's such a the the beauty of the baseball is so much in
  486. the in those kinds of moments. And then of course, the walk-off home runs. When you see these sort of these national
  487. teams have a walk-off and they celebrate the way the way they do is just it's just a pure
  488. joy and it's wonderful to participate in. Just to be clear by the way to what Shane said earlier, I also posted in our
  489. rundown the World Series odds and I agree it's equally shocking in the sense that the next closest We all agree
  490. the Dodgers should be the favorite, of course. Of course. >> But they're five times the odds of every
  491. other team. Like that's the gap. That just can't be it's it's unreasonable. And again, you
  492. you see that who who they faced off against not even in that top seven or eight or six or seven. It's ridiculous.
  493. Those odds. >> Yeah, so that I I just think you know, just some of our fans they're plus 210
  494. at the moment. The Yankees are plus 1,000, the Mariners plus 1,200, the Mets plus 1,300. We're not saying they
  495. shouldn't be the favorite. Maybe even two to one. All right, maybe you could stretch it and say three to one. Well,
  496. that's the thing. I I guess it's almost like a a meta this season question of like what
  497. is the most dominant a baseball team can be? Like, you know, like, you know, I I
  498. I like in in in in in in this day and age. I mean, obviously what the Yankees did back in the '30s
  499. was as dominant as any baseball team could be, but you can no longer be that dominant. Like like in our modern game,
  500. like what's what's the most you'd ever kind of advantage >> look at this, but I should have. If I
  501. went back to 2000 and the Yankees were going for a three-peat back then, they had already won three of four. What were
  502. their preseason betting odds? And it might have been similar. I And let me just be clear. In terms of their
  503. uh gap between them and the second most favorite team. It'll be It would be interesting to see that. I will take a
  504. look and I will post something at W Money Ball. Um so, let me just say what caught my eye. So, obviously guys, you
  505. know I'm a huge tennis fan. And you know, there's definitely the big two. No doubt about it.
  506. I'm starting to think Alcaraz is now one and Sinner is 1A. This Alcaraz is just incredible. I want
  507. to say it a couple of things again. He's 22 years old. Okay? He's got the same number of majors
  508. as John McEnroe won. That's seven. He's got a career Grand Slam. The youngest to do that.
  509. He's currently started the season with 14 straight wins. By the way, the record, if you want to
  510. know what the record is, this is the most impressive thing I've ever heard. It's Djokovic. Take a guess how many
  511. matches to to start the season Djokovic won. In 2011, by the way. 30? 41. He won 41 consecutive matches until
  512. Federer actually beat him at the French. So, he went the entire season until June
  513. without losing a match to start the season. He's won 31 straight matches on outdoor
  514. hard courts. So, the hard court season. Outdoor. Now, Bublik sinner does have an
  515. advantage on inner indoor hard courts. It's unclear why. He's won five straight titles on outdoor hard courts. He has to
  516. be the favorite now of every tournament he goes into. And I just think, you know, this is
  517. something I'm going to talk about as um Audi knows we're doing a if you'd like a
  518. webinar on Friday with a bunch I'm not going to ruin it for our webinar listeners who might be listening to this
  519. as well. But, I've asked Chat GPT for a prediction interval for Carlos Alcaraz and how many majors he's going to end up
  520. with in a 95% interval and I'll tell you how it does it on Friday. And for those
  521. listeners, we're going to post that on the Wharton podcast network as well. Um let me just say it's hard to imagine
  522. it's hard to come up with any reasonable prediction where he's not at least greater than Pete Sampras. In other
  523. words, the I would say, you know, assuming he stays healthy enough the lower bound I have with Sampras won 14
  524. majors. He's 22 with seven majors. It's hard to imagine a scenario where he doesn't I'm not saying he's going to get
  525. to Djokovic 24. I'm not going to say he's going to break his record. That's still 18 more majors. That's a lot of
  526. majors. That's a lot of majors. But, it's hard to imagine a scenario where he doesn't get to 15 which makes him then
  527. the winningest player of all time except for the big three. And I think that's his lower
  528. And that Yeah, well, I mean, it's not really cuz you have to have put cuz you you say it's hard to imagine he could
  529. have a catastrophic injury. And again, I'm not wish I would never wish that on anybody. But,
  530. that's got to be like a small probability P multiplying all this. And if he, you know, I mean, if he basically
  531. he could never win again type of thing. And so, I think I think that type of thing I
  532. mean, we can't we tend not to kind of think about those sort of rare events type things and I I don't I don't think
  533. that will happen obviously. I think it's a very low probability but that's the type of thing worth it. That's the
  534. scenario you where you would imagine him not breaking even So so let me let me respond to that. I think that the the
  535. You have to consider it'll fall off quickly. Not all tennis players have a long careers.
  536. Um some of I mean back in the day they fell off really quickly all the time. >> Audi, he doesn't need one. He doesn't
  537. need a long career. He's won all he's won a half of the last major of the last year. I understand that. eight more
  538. is he like age-wise if you compared him historically would he even be at like kind of the peak? He
  539. wouldn't be at the peak. >> No, 25 is the peak. 25 like 27's the peak for men's tennis. Yeah, but guys
  540. like Borg and McEnroe they were kind of done by 27. They were done by Actually, Borg actually retired at 26. McEnroe
  541. never won a major past the age of 25. Right, so those those career trajectories are possible. Um injury is
  542. possible although I don't know how common that is in in tennis career >> Very injuries are are not that rare but like
  543. career-ending injuries are very rare. I guess probably injury slash mileage is really
  544. what I'm kind of talking about. I think really what's more more of a the real wild card is who develops
  545. to offer him incredible That's a good question. That's really I mean so the the big three had the big three. They
  546. had to go against each other. They were slightly shifted but Number one, you pointed out Audi why a lot of people
  547. right now right now By the way, you guys remember this maybe don't remember cuz I've been following it. There was the
  548. big two. As a matter of fact, the record for the most consecutive majors between
  549. two players is Federer and Nadal, 11 straight. Sinner and Alcaraz are in now nine
  550. straight major finals just the two of them. That was until Djokovic came along. So look, if it stays the big two
  551. he's got a very strong chance of winning the most majors. >> Yes. Cuz if it comes to big three, that
  552. changes or four, that changes the math entirely. Mhm. That's all. That's all I was going to
  553. say, you know. >> think that's the biggest wild card in terms of prediction in the future. Among
  554. them all the wild cards we've talked about, that's the biggest uncertainty, whether he'll have to deal with two,
  555. one, or no can can challenge. >> fascinating sport because I mean, Eric, for what how old were you
  556. like what what how how much of your life have you not convinced yourselves that you're
  557. watching the greatest tennis player of all time? Like like for almost the entire adult life
  558. you probably were watching somebody and you're like, "That's the greatest tennis
  559. player of all time." >> Federer to like Well, I'll even go before >> Nadal to Djokovic, maybe somebody even
  560. before Federer. Yeah, it was never in the Borg, McEnroe, Connors era, Lendl because they were all I mean, Borg was
  561. better but they all beat each other enough that it it's clear Borg was slightly the
  562. best, but they each won seven, nine, 10, 11 majors. So, there was a there was a great strength there. Yeah, so I guess
  563. it's since Federer basically, right? Well, then Sampras came along and Sampras was just better. I mean, Agassi
  564. was a close second and they they you know, Agassi and and Sampras didn't have that unbalanced a record against each
  565. other, but Sampras was just the best. He was the best of his generation. And then
  566. of course there was the big three and that's been the rest of my adult life. >> know, it's just fascinating that it's
  567. been like, you know, that's like a 30-year period or something like that where at the time you thought you were
  568. watching the greatest tennis player of all time. Yeah. No, I agree. No, no, I'm I agree with
  569. Audi though. I think if he doesn't get to, let's say, better than Sampras, 15 plus, the most likely reason is two or
  570. three other players come up and then he wins his share but he wins one of four, one of five. So maybe of the next 30
  571. majors he wins five or six which is no slouch. That's nothing to be ashamed of but he's 30 and he wakes up and he's got
  572. 12 majors not 18 to 20 majors and that could absolutely happen. So guys maybe one last thing I wanted to
  573. talk about um something's happened in golf the last couple weeks and I just want to put it
  574. into comparison. So Shane Lowry and I know his last name is Berger maybe Justin Berger you know they've lost the
  575. last two tournaments in golf leading three strokes with three to play. Now I the reason I like to point this out
  576. besides that's hard to do as a pro. Tiger Woods is 51 and two leading not with three to play
  577. the whole fourth round to play. And so these seem like extraordinarily rare events that we've seen the last two
  578. weeks where essentially it's been a collapse by the leader going into the last couple holes and
  579. literally it wasn't that the other guys eagled and birdied. Both these guys made
  580. double bogeys, bogey like they literally played the last three holes in plus three and the other person played it in
  581. minus one where you know it's the old expression you know Ben Hogan said yeah all I got to do is win the Masters is
  582. par the last 11 holes. Let's see you par the last 11 holes on the Masters. It's not that simple but this wasn't that
  583. hard. So I would just want to point out Yeah. that I think we're seeing something in golf where some players
  584. just you know let's not take for granted players that just win at this incredible
  585. pace leading into the final round. It's not that simple. >> Oh no I I feel like that's one of the
  586. most unusual ways and I mean Tiger was unusual in a lot of different ways but I think that was one of his most unusual
  587. things is that kind of like lack of like, you know, not even not even a random amount of collapses or whatever
  588. in the last round. I think that's really kind of what you're talking about. I don't know if they do any kind of kind
  589. of collapse metric where it's like cuz you got you know you can imagine like some stati- like
  590. kind of like, you know, like how often is how often do how often does it happen that a a golfer drops three shots in
  591. three holes straight? I mean, that probably the you know, all happens kind of throughout the week like
  592. >> think I think what what what Shane is asking for is a baseline. Yeah, yeah, like like what you know, how
  593. Like obviously we treat ones at the end of the last round kind of specially because they usually have more
  594. consequence or at least we notice if they have more consequence. But like I kind of wonder what the baseline rate of
  595. that kind And I guess the the other question is not only the baseline, but what you point out, Eric, is it's
  596. happening because the the the uh the player in the lead blows it. Uh I wonder whether how often it it happens because
  597. of or some combination thereof. They blow it and they do really well. Those are great questions. Um but that just
  598. does lead to my question, my I guess my final thought here, um how um in which sports does psychology
  599. really play an important role? Um cracking under pressure. We know we see it in penalty kicks in soccer in soccer.
  600. And we I've seen it at second base in Yankee Stadium enough times. >> [laughter]
  601. >> You know, where they get the yips, right? And they can't throw anymore. Uh we've seen that happen. But how and Not
  602. to be up there. Golf golf I think what I'm thinking about sports that have a lot of it, it's kind of ones
  603. that almost do have that pause and action where you can kind of have you know, like baseball I think is
  604. you know, the ideal substrate for cuz you have very high leverage leverage events and there's also like the
  605. you like you really get in your own head because you know, there's like this long
  606. stop and action between kind of Yeah, the only thing opportunities. I agree, Shane. The only thing I would add to
  607. what Audi said is that let's remember though the two players I'm talking about were It's Daniel Berger, by the way. The
  608. The two players I'm talking about were the best players for 69 holes. So, that's the other thing you have to
  609. condition on. You can't just look at any three-hole stretch and say, "Well, the guy's plus three." That happens all the
  610. time in golf. Yeah, but these were the best players for 69 holes. And then all of a sudden, the last three? So, that's
  611. my point. >> No, and I I I would I would the base rate would be to sort of see if there's sort of like,
  612. you know, if you kind of had the base rate of like any three-hole window, and then you start looking at, "Oh, well,
  613. what happens now if we condition on Is there a different kind of rate for that last day or that last afternoon?" That I
  614. think That would be the kind of interesting comparison. Well, guys, it's been college basketball to in World
  615. Baseball Classic. We talked some tennis. We talked some golf. But, this is what we do on Wharton Moneyball. So, on
  616. behalf of myself, my colleague and friend Adi Wyner, my colleague and friend Shane Jensen, some combination of
  617. the three of us and Tae Massie here every week on Wharton Moneyball. On behalf of our sound engineer and
  618. producer today, Aaron Tran, on behalf of Dee Patel and Marissa Reno, we'd like to
  619. thank you for joining us here on the Wharton Podcast Network. Between now and next week, enjoy your sports, enjoy your
  620. statistics. We'll see you next week here on Wharton Moneyball.

Brief Summary

Ken Pomeroy discusses basketball analytics, team efficiency, and NCAA tournament predictions on Wharton Moneyball.

Key points

  • Ken Pomeroy Returns. Ken Pomeroy, the founder of KenPom, discusses his journey in basketball analytics.
  • The Four Factors Explained. Ken Pomeroy breaks down the four key factors that determine basketball success.
  • Miami of Ohio's Undefeated Run. Miami of Ohio went undefeated but faced criticism for a weak schedule. It's an impressive feat nonetheless!
  • The Impact of Matchups. Matchups are often discussed in tournaments, but their actual impact on outcomes is questionable.
  • Ken Pomeroy on College Basketball Analytics. Ken Pomeroy shares his insights on the evolution of analytics in college basketball.
  • The Impact of Momentum in College Basketball. Discussing whether momentum matters in NCAA tournaments, with insights from analytics.

Episode highlights

Tension

40/ 100Low tension

Related episodes