Why College Football Playoff Predictions Are More Certain Than They Should Be

Knowledge at Wharton

Ask about Knowledge at Wharton

Episode Description

ABOUT THE EPISODE The hosts break down how MLB now mixes analytics into Gold Glove voting and why current defensive metrics still misjudge positioning and value by position. They discuss why NFL power rankings look unusually tight this year, whether league-wide analytics and rule changes are compressing team quality, and why some playoff sims are far too confident. They also dig into the uncertainty in the new 12-team college football playoff, MLB’s recent gambling issues, and the Oilers’ 9–1 loss to Colorado.

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? The Moneyball podcast explains how decision-makers in the game can avoid the common mistakes and embrace the data. Episodes are recorded at the Wharton School.

#SportsAnalytics #MLB #NFL #CollegeFootball #Moneyball ----- 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 to Wharton Moneyball. Welcome to a full hour of sports analytics here on
  2. the Wharton podcast network. Kade Massie hosting with the whole crew. Eric Bradlo
  3. is here. Audi Winer is here. Sane. Jensen is here. Three of the four of us are on campus actually in offices almost
  4. as close as we were last week in studio. But we're back on Riverside coming to you Tuesday afternoon as we typically
  5. do. Show will go up on Wednesday. We are gonna kick things around this week. We are checking in with an old friend,
  6. maybe the original guest. OG has a different meaning in podcast world. The original guest, Neil Payne, might be the
  7. original guest for Wharton Moneyball. Neil is a frequent guest over the last 11 plus years, but he was with us,
  8. probably the only guest we can name who was with us in the basement of the original building before the studio was
  9. built. Neil Payne was a Philadelphiaian at the time. He um has since moved to New York for a while and after that
  10. moved down to Arkansas all the while doing great sports analytics journalism and work and he's
  11. carried a water for a lot of different organizations. He's got a terrific um terrific uh I've just lost my word
  12. >> substack. >> Substack. He's got a terrific Substack almost daily. Neil will keep you posted
  13. in all kinds of interesting ways on sports analytics. And one of the reasons we reached out to Neil this week is
  14. because there are so many different topics we could talk about. It's hard to pick a guest and just choose one of
  15. them. Neil has range. Epstein would love Neil Payne because he's got range. He can talk about any sport. In fact, he
  16. talks about sports we may not be interested in talking about, but we're going to focus on any number of sports.
  17. NASCAR is what I'm referring to. Neil, we'll see if we can get to that. >> Sure.
  18. >> But what why don't we say hello to Neil? Neil, welcome to the show. Thanks for
  19. joining. Hey guys. Hey, it's great to be back. I'm excited to talk to the whole crew
  20. and uh yeah, just love hanging out with you guys. >> Well, we'll take time with you anytime.
  21. I think we had you in person about this time last year. You came through for maybe a conference, some little
  22. football, a little Shabbat. We did a bunch of things in that short little visit. I think this was about a year
  23. ago. So, >> yes, we did. >> We'll take you in any form we can get you. Glad to have a visit. Guys, Neil
  24. has written his Substack on any number of topics over the last couple weeks. We can just kind of pick and go. Why don't
  25. Why don't we pass around choice? Y'all choose which of the topics you want to hear Neil talk about and we'll see if we
  26. can cover four in the in the first half hour here. Shane Jensen, do you have one
  27. that you want to talk about with Neil? Did you have a chance to look at him with his I know Audi looked at a bunch.
  28. How about we get Audi the first choice. Audi, you looked at a bunch. Let's see what he comes up with. Audi's first.
  29. Well, I you know, I was reading all about the Rams and and I definitely was going to cue that one up, but I I
  30. definitely I just heard from Neil that he's got a new one, not even posted yet, on baseball war, and you know, I can't
  31. resist. So, why don't why don't I let Neil tick pick from those two. Yeah, I knew that was going to be catnip for you
  32. guys. So, this is a post that's going to come out uh the same day that the podcast drops, actually. So, it's it
  33. kind of works out that people can go look at it. But I saw Matt Kemp was one of the names on this new uh class of
  34. firsttime Hall of Fame ballot uh potential uh baseball hall of fame members. It's not a great group, you
  35. know, apologies to Cole Hamls. He's probably the best one on there. There's Ryan Brawn on there who has a lot of
  36. baggage. Matt Kemp, Howie Kendrick, you know, guys like that. So, like a lot of Hall of Good type of guys, guys that I
  37. have nostalgia for for that era of like around 2010 type of uh peak of guys, probably not Hall of Famers. Maybe you
  38. guys can argue with me on that. But the one that really >> So, you're saying my sons and I may be
  39. by ourselves in Coopertown. Like, there may be no active player voted on and maybe the veterans committee will put
  40. some people on, but >> Well, I think they'll add the remember these are just the newcomers. So there's
  41. some holdovers from previous years that are uh most likely going to get in. Uh so it's really just the um the
  42. newcomers, but Matt Kemp really jumped out to me because uh as you guys know uh you mentioned I lived in Philly. Uh I
  43. worked for Sports Reference Sean Foreman uh and and all the great folks there um
  44. out in Mount Ary and uh I remember very distinctly Matt Kemp being one of our like early war stars. So we had and I
  45. get into this in the piece about a little about the history of war, how it kind of came out of the a bunch of
  46. different streams crossing with like Bill James and Windshares and Pete Palmer with linear weights and uh
  47. Mitchell Lickman with like the super linear weights where he used uh you know fielding data and base running data and
  48. all that and uh Tango Tiger Tom Tango who now works for MLB. He kind of pulled together all of those and came up with a
  49. framework for war, which then you had a bunch of folks kind of go off and run and calculate it for themselves. You had
  50. Sean Smith who created Rally Monkey War, RAR, it doesn't the R doesn't stand for
  51. reference. It stands for rally monkey. Uh that he was one of the ones. And then Fangraphs had their own version. I think
  52. Dave Cameron was uh involved in David Appleman in making that. And so that's how we ended up with now we have this
  53. like multi-olar war world that we've been living in for a long time with the fan graphs and the baseball reference
  54. the dueling wars. But at the time Sean Smith's war was the one that we had at baseball reference and Matt Kemp had a
  55. 10 war season and I it always like sticks out of my mind like hey man he was a capital G guy that year. He had an
  56. awesome season, but that 10- war season only lasted for about 7 months because the very next spring uh in 2012, we
  57. recalculated war. Uh and uh I worked with Sean Foreman and uh you know, kind of we all consulted together to come up
  58. with just changes in how we would approach the ways that it was calculated, including very importantly
  59. putting defensive runs saved as the fielding metric instead of total zone rating. And that ended up tanking Matt
  60. Kemp's war by a couple of wins uh at that time. Uh and so now he is at eight war. So only if you were around, you
  61. know, the the younger folks just wouldn't understand the reality of Matt Kemp as a 10- war player. But uh if you
  62. were around at the time, you had that picture in your mind. I did at least. And then now you kind of look at it,
  63. you're like, uh, you know, not quite as good. And he wasn't the only one that got dinged. you know, Albert Pooh's had
  64. didn't hadn't had a 10 more season at baseball reference, but he did at that time and uh Adrien Beltree as well. So,
  65. there were some seasons that just kind of uniformly got changed by that. But the only one and and I don't think it
  66. really matters for Matt like Matt Kemp, he wasn't going to make the Hall of Fame
  67. whether he had 10 more that season or eight or whatever. >> Just to point out, just 10 considered
  68. like >> that's like an epic season 10. Well, yeah, 10 war. And I looked at this for
  69. the piece is that everyone who had 10 war as a batter at least in a season except for Al Rosen who had 10.3 war in
  70. 1953 and then he just fell off because of injuries. But every other player in that group is either in the Hall of Fame
  71. right now uh tracking to be in the Hall of Fame Hall of Fame. So we're talking about like Aaron Judge, have to mention
  72. him obligatory every episode. uh Mike Trout, uh Mookie Betts had a 10- war season or they were left out because of
  73. non-performance reasons. So Barry Bonds and uh A-Rod and Sammy Sosa had 10 more seasons. So that's your group. It's like
  74. pretty much there's I think 56 guys and of those 55 are either in the Hall of Fame will be or left out because they
  75. had a scandal. That's a so it is a meaningful barrier. >> Huge. Yes. >> Let me ask let me ask a question. I was
  76. actually I was going to ask the question about how rare a 10 more season is, but
  77. you've already kind of answered that. Let me ask a different question. Imagine I asked you or Audi or anyone else that
  78. thinks about war measures and said, "Look, I don't and it relates to what you just said about Matt Kemp, like I
  79. don't care if you get someone's war that accurately, whatever that means, if they
  80. have a one or two war, who cares? But I really care about getting it accurately,
  81. whatever that means, if they're eight and above, because then we're talking about the right tail, the distribution.
  82. These things has high predictive and importance value. Would people construct a different war measure if what mattered
  83. was getting it right quote unquote in the far right tail of the distribution than if they were trying to get a
  84. measure that was, you know, if you'd like, reflective of everybody along the distribution of war?
  85. >> That's a really great question. And I mean, I think uh uh this gets into also
  86. some of like, you know, true talent versus uh measured performance, like forward-looking talent versus backwards
  87. looking, giving you credit for what you did because like Matt Kemp, he was not a
  88. true talent 10 war player. He's probably not a true talent eight war player that
  89. year, honestly. But the the actions that he did had eight war worth of value as best we could measure it. So, I think
  90. that is maybe the distinction there of trying to kind of pinpoint like true talent versus what you the worth of what
  91. you did. But, it's a point well taken because you would think if we were designing a war that it should apply
  92. equally to an average player versus a superstar player just in the sense of like this is how much these particular
  93. actions are worth or or not worth uh on the field. But at the same time, yeah, there's huge error bars. And I think
  94. that's the main thing that I wanted people to walk away from the story with is, you know, there's so much of a
  95. culture now especially, and you know, I probably contributed to that, a lot of people have contributed to that of
  96. breaking down like, you know, decimal like fractional differences in war or whatever metric you want. Uh, but I
  97. think pro football reference and I think Bill James came up with this in baseball
  98. even the idea of approximate value. We need to keep that in mind for all these metrics because they are still at their
  99. core approximate like it doesn't matter that it has >> Would you be okay if I did the following
  100. and you're right you could translate my question into is the standard error of war constant across the war scale and it
  101. probably is not. Would you be okay if I wanted to build a bootstrap distribution
  102. of performance based on uh and then I could simulate a gazillion seasons and >> No, I would not. That's a different
  103. issue that has to do with predictive accuracy. >> No, I wasn't doing predictive accuracy.
  104. I >> know you're trying to So, the problem with the standard error of war, if if
  105. you want to call it that, right? So, it's not even it's >> I can call it that because it's the
  106. standard error of estimated war. I can call it that. >> No, but you can call it anything you
  107. want. But when you say standard error, we we statisticians think sampling variation is as the root of uncertainty
  108. as the root. Where is that coming from? The uncertainty is not I mean it can be when you do predictive that's that but
  109. these wars that we're talking about the common war Aaron's 10.1 Aaron judges 10.1 Kemp's eight or whatever those are
  110. historical wars that's what he did on the field in the past we have bias these are bias meth problems they're not
  111. uncertainty they're bias problems no my standard error is coming from the fact that Aaron Judge I'll use what Neil said
  112. the true score model I did not observe Aaron Judge or Matt Kemp or Massie for let me finish my for 10,000
  113. observations. I I observed him for some number of finite number of plate appearances. That's a it's a sample of
  114. plate appearances. Any model observed score equals true score plus error is going to have measurement error. And I'm
  115. going to adjust war for that random sampling. That's what I was asking Neil about. Well, and I think that really
  116. gets into when you're talking about the defensive metrics because we can kind of
  117. work out the value of the offense. Although even that has philosophical questions of like we're using sort of an
  118. average weight for a single, an average weight for double or whatever when in fact those had different run values
  119. depending on the context of the situation and then you get into well should we use WPA or you know RE24 or
  120. whatever but then for the even if you're taking the standard sort of like hey we're using generalized values for these
  121. actions the defensive uh measurements have such uh I think error bars around them because They're ultimately trying
  122. to kind of measure a negative, measure the play that wasn't made or the runs that weren't scored. And so there's a
  123. lot of I think uh estimation error around that on defense particularly. So listen, I I got to jump in because Eric
  124. said something that has to get clarified. You're talking about a war that's estimating what what what Neil
  125. calls true talent. True talent is interesting. We can try to figure that out. We can try to measure that on some
  126. scale. But but most war values, particularly the the ones we look at in offense, are not trying to measure true
  127. talent. It's trying to trying to measure what you brought last year in the 650 observations that I saw, not what you
  128. would do next year or what that's not what I'm saying actually. I'm saying something different, which is if you
  129. know, I'll be frequentist for a second, which is hard for me to do, but I'll try. Which is, you know, you observed
  130. this 650, but you could have observed a different 650. and had to observe that which is based obviously I'm going to
  131. condition on the observed data that I had and I could observe a different 650 and that would give me a different
  132. estimate of war. I'm literally talking about the sampling properties of the estimator of war, which is the standard
  133. frequentist way to think about standard errors. And so that's what I'm referring
  134. to. And I would guess that the standard error of war or war hat if you'd like is
  135. not constant along the scale. And that's what I was asking Neil about. But that that first I don't think we know the
  136. answer to that. I would guess that it's probably not. Um but secondly what you're doing is using a frequencies uh
  137. understanding of of standard error which imagines there's some theta an unknown theta and war is not an estimate of an
  138. unknown thing it's unknown to us because we don't know how to define it and and in defensive measure we get all lost
  139. >> it's a different so good by the way by the way I just to our listeners out there I completely agree with your
  140. premise but that's a different issue I would almost use the measurement term of there's a construct validity issue
  141. >> is we don't even know What is that thing that we're measuring? Exactly. >> I was saying if you could define it and
  142. if you could say there's a deterministic mapping between performance and that measure, then you can create a sampling
  143. based estimator of the uncertainty in that because there's a deterministic mapping and all I got to do is reshuffle
  144. the observations in some way. Hey fellas, we I think this is a very reasonable point to wrap on because you
  145. said there's a there's a construct issue here which is obvious to anybody listens
  146. to a single conversation with us about war. So this is one of the clear things we take away made salient from the first
  147. observation that Kim's got knocked from 10 to 8. But Shane hasn't jumped in. Let's give Shane a word and then let's
  148. move to a different topic from Shane. Well, I guess you know since we're talking about things as a construct
  149. anyway, I I this kind of the judge discussion and u got me thinking with war of course historically we've thought
  150. about it like the fact that it's it's pinged or normed relative to replacement. I mean it makes sense for
  151. like roster construction and the building of a baseball team uh to do things that way. Okay. But when you talk
  152. about something like, you know, if you're trying to measure like the deviation of arian judge or something
  153. like that or the MVP conversation, I think it makes like something like more that's like, you know, like a war above
  154. like a average or something like that. I I don't think replacement's the right kind of quantile of the distribution or
  155. whatever to compare something like that to. And I think if we were doing some kind of relative to average baseball
  156. player, not only would that kind of just shift things, but I think it might kind
  157. of change the calculation a little bit, right? Because you'd have a different distribution of fielding and offense
  158. that you'd be comparing like standardizing relative to. >> So that's the question is like what
  159. would that do other than just rescale it? Because the rescaling alone wouldn't be interesting, but I but there may be
  160. other knock-on consequences. So you know for example um different positions might
  161. the gap between players and average might be different that >> or or or I don't know how the fielding
  162. models are like like does it it would change the pool that the field that like like the actual modeling components of
  163. it the pool of players over which you would estimate those models I think would change
  164. >> I want to >> because you're starting to talk about average price you talk about regular
  165. players so now you're just kind of you're almost conditioning on kind of starter like you know a different kind
  166. of type of player basically that you'll be kind of calculating these models of. >> Yeah. And the value uh I think first and
  167. foremost the value of sort of like the bulk of playing time would go down because you'd be compared to a zero of
  168. average. The whole thing the in fact in Tango's model the whole thing that uh differentiates wins above average versus
  169. wins above replacement is literally just playing time. it's giving you value for
  170. how much you played because ultimately uh his insight was that there is value to a very slightly below average player
  171. that plays a lot at a level above replacement. That's the entire point of war. But I think you're totally right,
  172. Shane, that for purposes of asking questions of like greatness, the the zero baseline point should not be the
  173. replacement level player because that player, we know they're not great. an average player is a lot better than a
  174. replacement level player, but we're talking wins above average would and they actually have this at baseball
  175. reference if you look at it. Of course, it still has Aaron Judge as the leader by a significant margin over Cal
  176. Raleigh, but uh it is sort of a more reasonable point for those conversations because ultimately wins above average is
  177. what you did to literally help win games for your team. The average team is 500 and whatever you add to that gives your
  178. team a winning record. But you could set that baseline point anywhere. You could
  179. say wins above all-star, wins above Hall of Famer, and most players would be drastically negative under that. But
  180. that could be a cool exercise like build your own replacement level. >> Eric's trying to jump in, Eric.
  181. >> No, I was just going to I I thought you wanted to change the topic. I was just
  182. raising my hand to change. >> I want to let you do that. I'm going to ask one. I'm always trying to get
  183. calibrated myself on these things and this is one of these things I am not calibrated on. You talked about the
  184. threshold of 10 war and we talked about when hopefully we don't talk about judges MVP over Rally, but let's just
  185. touch on that for a moment. Um, what was the war produced by those two players this year? I'm curious how close Raleigh
  186. got to 10. >> Well, a baseball reference judge was almost exactly 10. He was 9.7 and then
  187. Raleigh was 7.4. Now, that could open up. I know you guys have talked about this. You know, I have a whole can of
  188. worms where I think MVP is like a narrative storytelling award and that we should give credence to the the, you
  189. know, unprecedented historical nature of things in in ways that maybe aren't reflected by a stat. You
  190. >> mean Judge the first guy to win the batting title and hit over 50 home runs in 60 years? You are right, Neil.
  191. >> Yeah. No, that's exactly what I was talking about. I wasn't talking about demolishing the catcher record for home
  192. runs by more than a dozen. >> That sucks. or the switch hitter record either. >> Yeah, that too. Yeah.
  193. >> All right, guys. Let's go. Gradle, pick us a new thread from Neil's Substack or
  194. other writings. >> I will. So, I noticed this actually as I was preparing for the show today. It's
  195. in the rundown I put and Neil noticed this as well. So, two things struck me about the NBA right now. I know we're
  196. only roughly 15 games in, but first, OKC is 14-1 with a 15.5 differential. Now, that's, as we know, would break their
  197. last year's record, which broke the record for the highest point differential ever in the history of the
  198. NBA. But that's not what shocked me the most. The thing that shocked me the most
  199. is that there are five teams with a minus 10 plus on the record. And actually, I started to say to myself,
  200. first, you know, OKC is going to be playing those teams. you know, maybe OKC breaks the record this year of 73 and N
  201. on a historically set of bad teams. And two, I've just never seen a distribution
  202. like this. Now, it's early in the season, but I mean, if a bunch of teams end up with a minus 10 differential and
  203. OKC breaks the record by three or four points, I know running up the score on these bad teams. Wow. So, that was the
  204. thing that caught my eye. I know you have looked at this, so I wanted your thoughts on it. Yeah, you said you've
  205. never seen that before. That's because it's like literally never happened before. And that was something that I
  206. wrote about because I noticed the same exact thing where it's like, okay, see, they had this historic year last year.
  207. Uh, but we've been having historic points per game or net rating, whatever you want to measure it in, differential
  208. seasons in the NBA going back almost on like a yearly basis recently. Like the Celtics a couple years ago were in that
  209. conversation. Obviously, we had those uh Steph Curry, Kevin Durant uh Warrior teams before that, but we uh we've kind
  210. of had teams challenging this record. And last year, I noticed this as well because people were asking me, "Hey, do
  211. you think it's more that we have good teams or that we have a bunch of bad teams?" And so, I looked into it and
  212. when I did it last year, I found that yes, there was a slight effect of the really bad teams, but mostly it was just
  213. Washington. They were like the team that was uh substantially worse than the average for their slot. Cuz I looked at
  214. it by like, hey, what's the average differential for like the number one ranked team, the number two all the way
  215. down to number 30? The Wizards were way worse than the average number 30 over the past, I think since the the last
  216. expansion, so 05. But uh the the top teams were better than their averages. So I concluded that it's probably like
  217. the good teams are good with a little bit of bad teams are bad. This year though, it does seem like it's being
  218. driven mainly by the bad teams being really bad. And if you look at the distribution, you have this like
  219. ridiculous historical uh number of teams that are minus 10 or worse, like you mentioned, Eric. And then you have a lot
  220. of teams that are way better, especially the top three, Oklahoma City, Denver, and Houston.
  221. >> They would all break the record. >> They would break the record, right? And this is through uh I didn't compare uh
  222. full season from the past to the first 13 games this year because that would kind of skew things. We were talking
  223. about sampling earlier that you're going to obviously have a wider distribution in a smaller sample of games than an 82
  224. game per team sample. So I just looked at the first 13 games per team of other seasons. And even compared to that, this
  225. is a wildly different uh distribution. Much wider. the highest standard deviation of team points per game
  226. differential to start a season ever, at least since the the merger. I didn't look before the merger, but I think that
  227. that's telling uh that you have also one last thing, you have a weirdly good middle class of teams relative to how
  228. good the middle class usually is, and those teams are feeding as well on these really bad teams. So, it's going to
  229. completely distort what we think of as sort of the benchmarks for a good team. What regular season success means
  230. predictively for the playoffs when you get rid of the Wizards and and the Pacers and the Nets and the Pelicans and
  231. Kings and all these really bad teams. And so, the NBA has had a legitimacy problem for its regular season for a
  232. very long time. And this is, I think, probably shaping up to be the worst that I can remember in that regard where we
  233. know OKC is good, but like how good and relative to other playoff teams that some might be pacing themselves, some
  234. might be load managing, some might be injured, but getting better by the playoffs. We just don't know how good
  235. these teams are relative to each other. And it's like this entire 82 game multimonth sample going into the
  236. playoffs is like I don't want to say completely worthless but like it's not giving us the signal that we're used to.
  237. I was going to ask a question related to this. So as our resident both basian and
  238. probabilist theorist etc. Would you expect the distribution of let's say average point differential per game?
  239. Could I make an argument that the central limit theorem would apply because it's shots and then of course
  240. it's games and it's averages of games and the question reason I'm asking is if I don't know if Neil looked at this but
  241. if I looked at the histogram of point differentials would it be approximately normally distributed or should it
  242. conceptually therefore okay so if it is can we significance tests against that to see I
  243. mean listen even one last part of the sentence could we score seasons on how non-normal they are and maybe this is
  244. what Neil did and you know I I'm just asking you a question from a good idea actually
  245. >> I'll just throw in I would guess that the true talent whatever your true differential is on a neutral field or
  246. whatever you want to measure that is probably also normal doesn't have to be but I would imagine it probably is it
  247. might have a spike in slab or something like that you know some tail marginal tail probability giving you
  248. >> true talent in in in terms of the team quality or individual >> team quality yeah team quality Okay,
  249. >> team quality. I mean, remember the teams are a bunch of averages to start with,
  250. right? Um, but they don't have to. And then, of course, the actual what actually transpires over the course of a
  251. season is probably normally distributed conditional on your true talent. That's probably true as well. So, you're going
  252. to you're going to get something that's approximately, but I think you're you're
  253. your your your inquisitive look there, Kate, suggests that there's no reason why true talent should be normally
  254. distributed, right? >> I just I don't think the individual talent is normally distributed in the
  255. professional league. So that that's that >> no no no not at all but but team if you
  256. look at sort of the distribution of the winning percentages historically. >> So what you're saying a just for all our
  257. listeners out there is that you're basically going to get a Gaussian a normal mixture of normals which will be
  258. normal. But of course if if what Kate is suggesting by his expression is that the
  259. distribution of team talent was skewed or multimodal then you're going to get a non-normal mixture of normal. They
  260. provide a different distribution. That's right. But either way, I I I like this idea of like let's do a significance
  261. test on the distribution of point differentials and maybe we could score seasons by how non-normal they are or
  262. this we could look at the variance across seasons, look for patterns and stuff. >> Great, great, great. But we are being
  263. geeky analysts here instead of asking the practical question which is why why would it be that we have these tails
  264. this year? Well, I think one of the reasons why is that you have like your usual tanking
  265. teams uh that are just like perma bad and have been bad. I mentioned the Wizards. I mean, the Wizards just like,
  266. you know, they they have not really been relevant in a while. You also have teams
  267. that I think, and I had an interesting conversation with someone, one of my commenters about this, where they were
  268. like, well, did the Pelicans really like try to tank by getting worse? And I looked at it and I was like, well, you
  269. know, they did bring in Jordan P, who might be like a one-man tank uh operation himself. But I think they had
  270. this premise of, well, we won 49 games two years ago, won 21 last year, but Zion Williamson was injured and like all
  271. of our guys kind of underperformed at the same time. So, we're just going to kind of cross our fingers. We'll make a
  272. few uh superficial changes but then cross our fingers and hope that on the other side we can kind of regress back
  273. to that 49 win mean. Now they just fired their coach Willie Green which indicates
  274. that and and Zion has missed like half the season which indicates that they did not regret the mean they regressed
  275. toward was the 21 win version not the 49 win version. Uh so you have a few teams
  276. like that. Sacramento also in that group where it's almost like passive tanking like they just made some bad moves and
  277. they've they had some success a few years ago which was rare for them and then they misread that as being capable
  278. of being carried forward but they didn't do enough to keep the momentum going and
  279. so instead they've just kind of like slid into oblivion. Then I think the Indiana Pacers are really interesting
  280. because I think they're doing a version of if you remember the 1997 9697 season in the NBA. David Robinson was hurt
  281. really early in the season for the San Antonio Spurs. And so they just kind of did like an impromptu tank. This was a
  282. team that was like a perennial 50- win conference finalist type of team for years, but they opportunistically
  283. decided like, hey, the Admiral's out. We're not going to be uh worth anything anyway. I think it was Greg Papovich's
  284. first year as coach as well. So, they were just like, "Let's do a situational tank. We'll try to get as good of a
  285. draft pick as possible and then we'll reload next year when he's healthy and we can kind of uh, you know, move
  286. forward that way." >> You're going to tell us this got him Tim Duncan >> and it ended up getting them Tim Duncan.
  287. They got very lucky in the draft as Celtics fans. Maybe you Shane could remember this. Uh but uh they they got
  288. lucky in the lottery to be able to get Tim Duncan, but it was also very shrewd, which I think is the right way to do
  289. tanking of not this like deliberate multi-year kill the fan base type of like we're just this malaise sets in,
  290. but situationally if you're Indiana Tyres >> call that they call that the process.
  291. >> The process. Yes, that's what we call it now. We've developed that for it. Uh
  292. which, you know, that's a whole other thing we could open up. But I I like this Indiana approach of if you're going
  293. to tank, Tyrus Hallebertton is out for the year and you're coming off this finals appearance, but it was a little
  294. bit of a fluky finals appearance. They would be one of the weirder teams to have won even though they got to game
  295. seven. Arguably, you know, should have or could have won that game if he doesn't get hurt, but they would
  296. definitely be one of the more mold-breaking champion teams if they had won that. And so now they're just going
  297. to be as bad as possible and be in the hunt for there's a number of good prospects in the draft next year and
  298. then they'll get Hallebertton back and then they'll have another guy that's you know poised to to make a difference
  299. coming up and they could be good again. But I think it's this confluence of like
  300. you're getting to tanking from like multiple different uh possible ways that are making them so bad. And then OKC,
  301. you know, we shouldn't downplay them and say they're a product of uh of these bad
  302. teams. I mean, they're a really great team. Uh Denver looks amazing as well. Houston with Kevin Durant, they they
  303. actually did make changes to improve their team. Then you have other teams like San Antonio, Detroit, where they're
  304. kind of emerging from they did do a tanking type of uh cycle, but now they've got the stars out of that and
  305. they're moving forward uh with that. So, it's like the confluence of a bunch of different like team timets syncing up at
  306. the same time in the league right now to produce this crazy lopsided distribution
  307. of teams. >> All right. Well, um, Eric, do you want to jump on this before we move?
  308. >> Yeah, I just had two follow-up questions to that, Neil. So, when do we get to a
  309. point quickly that you would, uh, predict OKC may break the record of 73 wins? They're certainly at a pace right
  310. now above it. And given the number of bad teams, um, you may you may get to that point. Um, and uh I don't remember
  311. my second question, but that was my first one. Oh, I remember my second one. I'm going to give you OKC,
  312. Denver, and Houston, and I'll take the rest of the league. How do you like that to win the finals?
  313. Uh, I, you know, could I swap out uh Cleveland for Houston? At least those are in my model. Those are the top three
  314. cuz they would have around a combined 60% chance. uh OKC, Cleveland, and Denver. But if you're giving me the
  315. Rockets, I still probably take it because it would be about 55 45. >> Well, I know Audi's going to take the
  316. field, but you're going to take OKC, Houston, and Cleveland. And all right, we're so good. We'll track this.
  317. >> OKC is really carrying that probability, right? >> It's about them at kind of a really high
  318. amount. That's all OKC based. Yeah, I think it's the majority OKC and then uh Cleveland and Denver around double like
  319. low double digits in there. Um what was the first question? I forgot OKC break the record. Can they get it to 73?
  320. >> Oh, so I think they need a little bit more. So the they're projected in my model for about 65 wins right now. I
  321. think their overunder at FanDuel when I looked at a couple days ago was like 65 and a half. So that tells me they need a
  322. little bit more. But probably if they sustain this for another month, they would be like their over under at
  323. FanDuel might be above 73 wins by then. >> Neil, I have one follow-up question. Um,
  324. you know, we talk about owners on the show a fair bit and thinking broadly people appreciate more now than they
  325. used to, how important owners are, but probably still don't appreciate how important owners are. Just observing the
  326. Pelicans owner is the same as the Saints owner. Am I wrong to be judgy about that
  327. particular ownership group? Well, in general, I think owners make a huge difference in the NBA and it it has
  328. to do with like setting the culture, convincing players to come there, the management, uh, like it just filters
  329. down from the top. So, I think we had always traditionally as analysts undervalued the importance of the owner
  330. in general. In terms of that ownership group, I mean, it's tough to say when it came to uh it probably they have been
  331. responsible for giving Zion such a long leash even when he's been out of shape, even when he's been injured every single
  332. year and they've just kind of pinned all their hopes on him rather than trying to
  333. think of like, okay, this Zion thing is not working out for us and we might need
  334. to think about moving on. It's like that lottery with Zion was a long time ago. like I was still at 538 and uh it was
  335. like pre- pandemic uh I want to say. So that tells you like you can get it in your head that a certain player is like
  336. I don't know you fall in love with them as a transformational talent and certainly when healthy Zion is that type
  337. of player but at a certain point you have to kind of understand and recognize that he probably is never going to live
  338. up to what you thought he was going to be and then uh move on >> yeah it's I think it's important to
  339. recognize there's a reason for why this happens in basketball and it has to do with the smallness of the team and the
  340. ability ability of a generational talent by himself to bring you to at championship or near it. And because of
  341. that, an owner just thinks they could they don't they could just do it right with and kind of get involved and put
  342. all your chips behind one or two players and tell them this is who we should be drafting. And
  343. >> all we could do, Audi, is forecast who that generational talent was going to be
  344. with certainty. We'd be all set, >> right? But, you know, they don't even come along that much. So, I think you
  345. wouldn't see it in baseball because it's such a damn big organization with with 26 on a roster and and 200 as a minor
  346. and football rosters are huge. No one would have that. I mean, I shouldn't say no one. People have very big egos, but I
  347. would imagine that that kind of hubris is ridiculous to try to imagine you can manage.
  348. >> Well, this is really neat. I've never heard that before. And it's an interaction between ownership hubris,
  349. which we know to be a thing, and structural differences across sports, which we know to be a thing. I've just
  350. never thought about the interaction between those two things. I think it's a really interesting observation. Why
  351. don't we wrap the first half there since we've been so slow to work through Neil's production. We'll keep him
  352. around. We'll do the second half with Neil, too. So, come back and join us after the break.
  353. Welcome back. Welcome back to Wharton Moneyball. Welcome back to the second half of this week's show. talking sports
  354. analytics as we always do as we have been for 11 and a half years now and I got the whole crew in here. Shane Jensen
  355. is here. Audi Winer's here. Eric Bradler's here. This is Kade Massie. We are talking with our guest Neil Payne.
  356. We like Neil so much. We're going to keep him around for the second half of the show. Also, he's so dang prolific.
  357. We have to keep him around to work through his stuff. By the way, I just realized I I was reading Barnwell this
  358. morning for the first time in a long time. He's another prolific guy, right? When he decides to write something, he's
  359. going to write something. Those are long pieces. super insightful, but really I thought Neil, y'all are kind of cousins
  360. in in the way you do your work. I mean, you're great writers, but also good analysts and you're very prolific and um
  361. I was thinking about you while I was reading this stuff. We um anyway, for what it's worth. All right.
  362. >> Good company to be in. >> It's good company for sure. Okay. We talked about two of Neil's recent
  363. pieces. We've talked about baseball and basketball. Shane is going to pick a thread, then I'm going to pick a thread.
  364. Shane, what do you got? Well, I wouldn't mind uh talking a little bit about the recent article you read on or wrote on
  365. the Rams, I guess, and basically talking about how it seems like I I mean, you could talk a little bit about the
  366. details. They're they're you know, they're kind of a the number one kind of team in in the foot in football kind of
  367. in terms of at least balance defense and offense. And it's kind of it's interesting to kind of I I mean I don't
  368. disagree, but one of the kind of points you make I think in the article is that it's not only that they're the Rams are
  369. good now relative to the rest of the 2025 NFL, but they're good kind of relative to past recent Rams kind of
  370. McVey Vay Rams teams and I think about those some of some of those old like other Rams McVey teams where they had a
  371. I feel like a lot more kind of talent skill kind of people like Donald and Cup and these types of players. So maybe
  372. maybe talk a little bit about how how the Rams are doing it. How this Rams team is somehow better than those kind
  373. those past teams. >> Yeah. Well, I thought it was really interesting. You know, the Rams uh they
  374. were among the teams, you know, they went to the playoffs last year, almost knocked off the Eagles. A little scary
  375. moment there. Uh at least speaking as a Philly fan. Uh and so, um, you know, they've kind of made this comeback, but
  376. it's just kind of built on itself. Uh, and this year you make a really strong case that they've been the best team in
  377. the NFL. They have the highest SRS simple rating system, which is just schedule adjusted point differential in
  378. the league, especially after beating Seattle, who's number two in the simple rating system. It was not like a super
  379. convincing win. it looked like it was going to be early on, but just uh you know the fact that they have emerged as
  380. the the possible best team in the league was something that I don't think people
  381. would have thought just a few years ago because if you think about the early version of the McVey Rams, it was all
  382. about almost this like I called it like a get-richqu scheme type of approach to team building, which was just like spend
  383. a lot on stars, lock in your your bigname stars if you've got them, go out, get free agents, trade away all
  384. your first round draft picks. They famously went, I think, seven straight years without a first round pick. And
  385. ultimately, it was like a win now at any cost type of approach. And they went to
  386. the Super Bowl. And then a few years later, after they uh traded from Goff to Stafford, they won the Super Bowl. And
  387. that seemed like that was going to be it though. It was going to be like, okay, you pushed all your chips into the
  388. middle, you won this big bet that you made, now the Bill has to come due. And you know, it was a nice run. And then I
  389. think they went 5 and 12 the year after which was one of the biggest drop offs by a defending Super Bowl champion ever.
  390. But it really defied the expectations that then they were able to kind of retool off of that. They didn't even um
  391. rebuild. They just reloaded and they kept pushing forward. And so I was like digging into why that happened. Some of
  392. it is Matthew Stafford has been awesome this season and he has been a very inconsistent quarterback over the years.
  393. a lot of it. You know, when he was in Detroit, he had not that much supporting talent aside from Calvin Johnson to kind
  394. of work with, but uh he has always in my mind early in his career, he had this reputation of being a guy that like the
  395. itest like better than the stats. Uh and you'd think like he has Mahomes type talent, but where's the where's the
  396. production? But he in LA he has kind of evolved and especially this year he has been rightly considered one of the MVPs.
  397. Puka Nakua helps as well. I did a little thing where I did a principal component
  398. analysis. You guys will like this on a variety of different receiver stats trying to kind of compress them down
  399. into two dimensions. One of which was like usage and one was efficiency. And that could be efficiency on like deep as
  400. a deep threat or as like an underneath guy. But basically it's like how are you doing your production and then how often
  401. does the team call on you as like a target share. And Puka showed up as being one of the highest guys in terms
  402. of combining efficiency mixing both the attributes of an underneath guy and a deep guy and then having one of the
  403. highest uh target shares I think just behind Jackson Smith injigba uh on the season. And so, uh, I thought
  404. that that's like, okay, that's an underrated part of Stafford. But then the main thing that I think separates
  405. this Rams team is that their defense has been really good, which was not a hallmark of the early gooff, Todd
  406. Gurley, you know, that type of Rams team. Even when they went to the Super Bowl, they were outclassed. Even though
  407. they had Aaron Donald, they were outclassed by that Bellich Patriots defense in that Super Bowl. And it was
  408. this idea of like, well, if you're this high-flying offensive team, sometimes defenses can just shut you down if you
  409. don't really have that other gear. But they've been one of the best defensive teams in the league. And they've done it
  410. almost entirely through like lower round draft picks, undrafted guys, guys that they got from other teams. Only two of
  411. their defensive starters this year were players that they drafted themselves with a pick in the first two rounds. So
  412. to me that's like conditional on them making this turnaround despite trading away all those picks and spending all
  413. that money uh you know and not being kind of capped out. The only way they were ever going to be able to make this
  414. pivot was Stafford finding the fountain of youth and then discovering a bunch of
  415. unsung guys. I mean Puka Nua is also a fifth round draft pick and you look at that guy now in his production he's like
  416. he's probably the best receiver in the league and you were able to get him at at such a bargain. So, it's a testament
  417. to the scouting that they did, the coaching, the cap management, and you got to get a little lucky to make to hit
  418. on so many guys, but it's really worked out for them so far this year. And I'll just sorry again, I'll just add on top
  419. of that one more thing just because you talked several about them kind of mortgaging their future and past drafts.
  420. They actually now are a team with, I think, two draft pick. They have two first rounders. Yeah, they flipped it
  421. and Atlanta's their their other pick. And Atlanta, that draft that draft slot's moving up the board as as we see
  422. it is. and they also have some of the most cap space for next year. So they've totally flipped that identity as being
  423. now they're kind of a balanced measured we're being running this team responsibly type of operation.
  424. >> That it's interesting that they think they have both gears. You know, most front offices have one gear or the
  425. other. And so that that's impressive, but everything you said is impressive about the front office. And what I the
  426. main thing I'm taking away because I'm kind of at a distance from this is I would have thought any story about the
  427. Rams would have centered around McVey, but McVey is the offensive guy. And so for you to say the defense has been kind
  428. of a surprising and important part of this is say oh well it's gonna I mean he affects the whole team but his real
  429. caches on the offensive side of the ball. So it really puts more of the premium especially based on the player
  430. acquisition that you've talked about on the front office. What else? >> And MC, >> oh, I was just going to say McVey, uh,
  431. you know, one of his best qualities seems to be as a manager. We've seen so many times that these guys that are
  432. really great tactician coordinators when they become head coaches, they screw it
  433. up and they can't really get it right. I mean, that's like the story of a lot of
  434. guys in the league. Even guys from McVeyy's own pipeline uh, have have done that and certainly other coaches trees.
  435. Uh McVey seems to be really good at like delegating to the right people, managing
  436. the team, and just running it as a professional operation, especially for somebody so young. And he is shooting up
  437. the list. I was looking at the list of like career winning percentage. And it's like, you know, he's I think top 30 in
  438. wins above average or wins above 500 uh all time in NFL history at this point. So, it's sort of like pretty soon we're
  439. going to look up and we're going to see it's Shawn McVey is not this like babyfaced kid that we think of him as,
  440. but instead it's like, oh, he's like literally one of the best coaches of all time.
  441. >> That's a neat stat you just quoted win percentage, but you said wins above 500.
  442. So, that combines both rate and duration essentially. >> Yeah, exactly. Yeah, it's like they have
  443. uh it's it's like our wins above average that we were talking about in baseball,
  444. but for coaches um they have that at uh Pro Football Reference, which I think is
  445. a really cool stat because like you said, you can't if you do winning percentage, you have to set some kind of
  446. threshold and it's like how many games should be in that. So, it's a more organic way to measure that.
  447. >> Awesome. Eric's trying to get in. Eric? >> Yeah. So, football's an interesting
  448. construction in the playoffs. So, let me ask you the following question, Neil. Right now, I think most predictions
  449. would have to have the Eagles as the one seed besides strength of schedule remaining. Um, they also beat the Rams,
  450. which we could question. I mean, they beat the Rams on a block field goal, but ignoring that they beat the Rams.
  451. If the Eagles played the Rams right now, Eagles have home field, let's say, who would be favored in that game given what
  452. you just said that the Rams, By the way, I looked at the It's interesting that I
  453. I I saw what you were working on, but I had done some stuff independently. I mean, the Eagles, I understand you did
  454. schedule adjusted point differential, which I think is great. Eagles are only 33. The Rams are plus 100 this year.
  455. They're not even close. How much stronger are let's forget that the Eagles would get a buy. Let's imagine
  456. they meet in the NFC Championship game like one versus two, but the Eagles are at home. Are the Rams that much be what
  457. would be the line? What would be the win probability for the Rams in that game? >> Well, so I like to look if you're
  458. talking about hypothetical uh lines, I like to look at Mike Buoy's inpredictable. I know you guys uh know
  459. that site, but they have the sort of like generic points favored metric uh based on the lines from recent games.
  460. And the Rams are a plus 6.3. The Eagles are plus 5.1. So if that was like a neutral site, the Rams would be like a
  461. point and a half or one one point to one and a half point favorite. Uh but to your point, Eric, the the stats from
  462. this season would not have it anywhere near that close. the the Rams would be massive favorites. But I think there's
  463. an element of a little of what we're talking about in the NBA of this sort of like teams pacing themselves. The Eagles
  464. haven't revealed their true talent this season. It's only been half a season. We
  465. have to regress to their prior, you know, all these things of like we know the Eagles, the Eagles seem to win every
  466. game and I watch all every Eagles game. It's like some of the ugliest games possible. I know some of that is like
  467. the tush push and being able to just, you know, you can grind out wins whenever you need to, especially on like
  468. closing drives and, you know, when things are are tight, but the the stats like SRS and even all of our predictive
  469. metrics, they are geared toward blowing teams out. They want to see you impress in your wins. And the Eagles have been
  470. highly unimpressive, but in a different way than like maybe the Broncos, uh, who
  471. have also been highly unimpressive in their wins or the Chiefs last year where theirs feel like a string of like fluky
  472. circumstances just repeatedly like they flipped heads 10 straight times and it ca, you know, they called heads and it
  473. came up heads 10 straight times. With the Eagles, it does feel like almost the coin is weighted. And maybe I'm biased
  474. in saying that, but it does feel like they know they can play a particular style of game and win it if it gets to
  475. be close at the end. But the question is, can you really rely on that 100%. And don't we need to see some level of
  476. dominance at some point? I would love to see some amount of dominance. >> Yeah, for sure.
  477. >> Start watching the Patriots, bro. >> I know. That's the crazy thing. Talk about regressing, you know, to their
  478. prior though. >> The Patriots, it's like 2015 or something like that in the in the AFC.
  479. >> It's remarkable. >> It is crazy. >> It's an interesting interesting season. You know, the thing I was reading from
  480. Barnwell was about the Chiefs actually, and they're sitting there at 500 right now. He's like, "Do we count them out?"
  481. And so, he really makes cases both directions on counting them out or not. And I'm in I I'm in a thinking I'm not
  482. counting them out. Those >> No, I'm not counting them out. Better they better indie at KC. They better uh
  483. Indie at I'd still put them above 50% to make the playoffs. Whether what happens
  484. from there, I don't know. >> Well, for you know, anything can happen if you if you get if you get in there,
  485. especially if you got somebody like Mahomes. All right, last thread. I'm gonna go to the Connor Baddard thread,
  486. which is is this young supposed star, former number one pick, finally showing that he's going to be a great player.
  487. So, I'm interested in Bedard in particular, but I'm let's do a little hockey. We've done the other three major
  488. sports. Let's do the fourth. But I think it generalizes to this question of when
  489. do we know a young player is going to be or not be what we hope him to be. So, for example, JJ McCarthy with the
  490. Vikings right now. Pennix just goes out with a knee injury. these young quarterbacks, there's always a question
  491. of how much do we need to see Sam Darnold? I mean, we got we thought for years we had that one picked. So, this
  492. is an interesting question in general in sports. So, Baddard, for those of you who don't know, number one pick, massive
  493. attention. Blackhawks got him three drafts ago. And then, you know, we have another one right behind him. I I I
  494. don't know if you wrote about it, but Celibbrini playing his first full season with the Sharks. He was consensus number
  495. one. So when you get these consensus number ones, when do we know whether they live up to it or not?
  496. >> Yeah. And that is uh something that is kind of pertinent across all the different sports because we had one year
  497. where we had Bedard, we had Wimby in the same like year. Uh a couple years earlier we had Trevor Lawrence. That was
  498. kind of the NFL version of that. These like generational prospects. And then we had Caitlyn Clark in the um WNBA as
  499. well. So, it's like there's there's number one picks and then there's like these people that get talked about
  500. generationally and that gets thrown around way too much, I realize. But, it is it is true that like Wimi was being
  501. talked about on a totally different level than even like Cooper Flag or someone like that. Like there's just
  502. this like there's a regular number one pick and then there's transformational. And Connor Bernard was talked about as
  503. being transformational as well, just because he had this combination of like the most ridiculous shot that you've
  504. ever seen in your whole life, and then also this like awareness of of, you know, the trademark of every great
  505. center in the in hockey of like knowing where everyone is, where the puck's going to be, who how to set up your
  506. teammates and everything like that. Uh Baddard, you know, he's fine his first couple years. Uh I I had kind of written
  507. things that were defending him because people were you know the people love to start creeping in with the bus talk or
  508. the kind of hey maybe we overestimated. The only thing we love more than building these people up is to then tear
  509. them down when they don't perform like immediately at the level of like one of the best players in the league. Uh and
  510. so I had written about like hey calm down. He's still one of the best uh young players like through age 20 of all
  511. time. you know, maybe not quite as good as Crosby was at the same age, but like close or McDavid, but close to that. And
  512. then this year, he has been one of the best players in the league, which is why I wrote this story because it's like
  513. this breakout moment that I think you have to have within the first handful of seasons. I know it varies by sport, but
  514. I talk about Trevor Lawrence. We have not seen that yet from him. And I think we would all have to agree that our
  515. estimations of him being the next Pton Manning or, you know, someone like that or Andrew Luck, they're just not there.
  516. Like we have to downgrade what we think the rest of his career is going to look like cuz he just hasn't, you know, quite
  517. elevated himself to that level of like he he should be where in the rankings where Matthew Stafford is right now, you
  518. know, if if he was that type of guy. Uh, and so it's always encouraging when we see like Wimi at the near the top of the
  519. NBA rankings and Connor Baddard at the top of the hockey rankings because then you can kind of say like, okay, we're
  520. probably right about these guys and their generational uh potential. And when I looked at this for hockey
  521. especially, it seemed like probably by age 20 you need to have like this bump. I looked at a number of other guys like
  522. Gretzky, McDavid, Crosby, Yarm, Yagger, Ovuchetkin, Mariel Lemieux. All those guys, a couple of them had shown signs
  523. of it even before age 20, but they all had a bump in their performance at age 20. Now, you could still make the Hall
  524. of Fame without that, which I thought was kind of interesting. like the this is truly the definition of like value
  525. above baseline Hall of Famer where it's like uh if you're talking about being uh
  526. on that inner circle level, you really need to produce uh one of the best seasons in the NHL, you know, for that
  527. season by age 20. If you don't do that, you're probably just you your ceiling is
  528. probably just regular Hall of Famer, which is like, "Oh, no. I'm only a normal Hall of Famer instead of an inner
  529. circle hall of famer." But it does matter. It really does seem to matter. >> Interesting. Eric's trying to get in.
  530. Eric, >> no. I was going to ask the question that Neil just said, which is are what are
  531. you looking for? Are you looking for like an extraordinary peak performance year? like if you know uh if Connor
  532. Bdard had three let's call them.95 distribution years before age 20 be like yeah he's targeting towards a regular
  533. hall of famer he needs a one year or at least that's a like a real n you know 99th percentile year for you to project
  534. him as you're right generational talent. So it's not it's not the summation you're looking for a peak performance.
  535. Yeah, a little bit. Especially since like peak is kind of all you have when somebody's 20, you know, especially as
  536. they kind of get toward that level of like the years in which we'd really expect them to have at least figured out
  537. what it's like to be an NHL player. You give them a little grace in those first few years and then it's like, okay,
  538. you've been here for a while. You know what this is like. Now it's time to kind of take off and do that. And yeah,
  539. you're right about the percentiles because it's like extraordinary claims require extraordinary evidence. And our
  540. claim going in was that this guy is the next McDavid. And so you have to provide
  541. extraordinary evidence of that. >> Neil, it strikes me that you're talking about the sport where we see the
  542. youngest player on the full professional stage of the m of the four major sports.
  543. So baseball is famously long pipeline. Football, they just get drafted later. They get thrown to the wolves mostly,
  544. but they get drafted later. Basketball, they they mostly get drafted at a little
  545. bit later age. these high school these these uh hockey players get drafted basically coming out of high school. A
  546. lot of them go into some kind of development work. The team maintains their rights but they go to college and
  547. they go to one of those smaller leagues. The very best players and this is my question like how did they develop Bard?
  548. How quickly do they put them on the ice? Because it's kind of it's kind of um striking to hear you say they got to do
  549. it by 20s. Like my god that guy's a kid. >> I know. Yeah. It's it's a little like
  550. Olympic gymnast or something, right? Like there's just so much younger. But I do think in some ways the the hockey
  551. development pipeline is a little bit more accelerated like you were talking about. They uh often the best ones they
  552. don't go to college. Instead they go through juniors and they kind of are playing up uh compared with their age
  553. against players that are older or better or just kind of they're 16 and they're going uh almost playing professional at
  554. that age. Uh and and it really kind of accelerates the development. And it's similar to like in baseball, somebody
  555. can be drafted and in, you know, the summer and then like play in the major leagues like that year. Uh in in
  556. baseball sometimes like we've seen this with pitchers. Um >> rarely, right? Rarely.
  557. >> Yeah. Sometimes. Oh, well, if it's like a Paul Ske, who I think is sort of also
  558. in that conversation as well. So, I I do think that hockey kind of combines all the different elements that would cause
  559. you to see something at a younger age than other sports like the accelerated pipeline, the ability to jump right in.
  560. And I'm really fascinated that the NHL, you mentioned Mlin Celibbrini, Matthew Schaefer as well, he was the top pick
  561. for um the Islanders. He is now having one of the best young seasons by a defenseman I think since Bobby our which
  562. really also says something. So, uh, before the season in the NHL, I had done this thing where I looked at the fact
  563. that the average age of the league in terms of just weighted by playing time or or player value had gone up for seven
  564. consecutive years, which uh was sort of spoke to the stagnation of the league. We'd also had like the same finalists in
  565. multiple seasons and all these things. And it was like when are we going to get this this change in in the hockey right
  566. now? And it seems to have come this season led by Baddard, but also those other guys where you're seeing a
  567. disproportionate number of the top players in the league in terms of like value over replacement or just scoring
  568. uh be the younger stars. Now, it's like this next generation is finally taking over the game. Now, that doesn't mean
  569. that Crosby is being pushed out. The Penguins have actually been amazing, surprisingly, shockingly, this season.
  570. uh or McDavid, you know, he's still in the prime of his career, but you've got these young guys kind of nipping at the
  571. heels of the of the established players in a way that we really had not seen in a number of years in hockey.
  572. >> All right, good fun. Well, listen, why don't we wrap it there? We managed unintentionally to cover the four major
  573. North American sports in our four topics. Many, many thanks to Neil Payne. Neil, you can find him. We can't
  574. recommend his substract strongly enough. which is fantastic writing, keeping you
  575. as apprised and interesting uh across across the board in sports. We didn't get, for example, to your piece on
  576. NASCAR, which I was mocking for, but I wanted to hear about because he talks about tournament design essentially a
  577. thematic topic for us here over the years. We'll pick that up at some other point. For the whole crew, Eric Bradlo
  578. for Shane Jensen, for Winer, this has been Kate Massie. Many thanks to Neil Payne, many thanks to Dion Simpkins, D
  579. Patel, Mariss, Marissa Rena, the whole team. Many thanks to everybody and to you guys for listening. Come back and
  580. join us next time. Between now and then, enjoy your sports.

Brief Summary

Neil Payne discusses baseball WAR, Rams' success, NBA dynamics, and young hockey stars' performances.

Key points

  • Welcome Back, Neil Payne. Neil Payne returns to the show after a year, bringing his insights on sports analytics.
  • The Significance of 10 WAR. Discussion on the rarity and importance of achieving a 10 WAR season in baseball.
  • Measuring True Talent. Exploring the complexities of measuring player performance and true talent in baseball analytics.
  • MVP Award Debate. The discussion shifts to the MVP award, with a focus on its narrative aspects.
  • Historic NBA Team Performance. OKC is on track to break the record for wins, highlighting a unique season.
  • Tanking Strategies. The conversation explores different approaches to tanking in the NBA, including situational strategies.

Episode highlights

Tension

45/ 100Some tension

Related episodes