NFL Analytics Preview, QB Forecasts, and Team Rankings for 2025

Knowledge at Wharton

Ask about Knowledge at Wharton

Episode Description

ABOUT THE EPISODE Moneyball podcast hosts and Wharton Professors Eric Bradlow and Cade Massey talk with Aaron Schatz about 2025 NFL outlooks, including why the Ravens and Bills rank among the best teams. Aaron Schatz highlights Jaden Daniels’ standout rookie play, expected Eagles regression, and the challenge of modeling coaching. They also discuss Ichiro’s hit streak, rising WNBA viewership, and an LPGA record for most consecutive unique winners.

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.

#NFL2025 #FootballAnalytics #NFLPredictions #SportsAnalytics ----- 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, welcome to Wharton Moneyball. Welcome to a full hour of sports analytics here on the Wharton podcast
  2. network. This is Kate Massie hosting this week along with my longtime colleague, collaborator, co-host, and
  3. friend Eric Bradlo. Our two other co-hosts, Shane Jensen and Audi Winer, are out this week. They're doing Shane
  4. and Audie things. They will be back. Some combination of us are here almost every week of the year and have been for
  5. more than 11 years now. Delighted to be here with you. We are going to run through a regular schedule. We're going
  6. to record, we are recording on Tuesday afternoon. The show will go up on Wednesday. We're going to do about an
  7. hour. We're going to do a guest in the first half as we have been lately and then some open topics in the second
  8. half. We have back on the show for the eenth time and um a very happy annual marker of sorts. Aaron Shots is here
  9. with us which means football season is just around the corner. Delighted to see you Aaron Shots. Welcome back.
  10. >> Hey, we made it through another off seasonason. Football people any I I am aware there are other sports but
  11. football people made it through another offseason and I'm ready to go. I'm ready
  12. for a new season. I'm ready. training camps are open and I'm ready to uh completely get obsessed with whether
  13. guys throw interceptions to their teammates and pretend that it means something.
  14. >> Good. Good. Well, we appreciate the enthusiasm. That's what you're here for. Most of you guys know Aaron. If you
  15. don't, let me give you just a little bit of background. He is presently the chief
  16. analytics officer of FTN Fantasy. He also writes for ESPN. He's very well known for founding Football Outsiders.
  17. many years ago, more than 20 years ago, he was early into the game on advanced football analytics. He's the creator of
  18. DVOA, which has been doing a lot of work in our community for a long time. And um
  19. he's always a delight to talk to. I I always like to note also that he is the his his writing tree, his analysting
  20. tree is as impressive as it gets in football, if only that perhaps the two leading lights. He I give you credit for
  21. finding Bill Connley and Bill Barnwwell. If I think if you do word count right now, the leading word count producer on
  22. the NFL side for ESPN would be Barnwell and the leading word producer on the college football side of ESPN would be
  23. Connley. So Shots is the journalistic father of both of those guys. >> A lot of my writers in the past have
  24. turned out to be quite verbose. Uh Bill Connelly has a new book coming out by the way, so I hope you'll talk to him
  25. about that. It's really exciting. And uh yeah, my coaching tree includes Doug Ferrar from Athlon and Mike Taneer who
  26. has his own site now and Ryan Wilson from CBS, Michael David Smith from Pro Football Talk. It's a lot of people.
  27. >> You're forgetting FEI, I think. >> Yeah, Brian Frimo, who does the FBI ratings? Yep,
  28. >> exactly. Freo as well. That's a heck of a tree. Heck of a tree, Aaron. Okay, man. Let's get into it. This is uh NFL
  29. talk one for us and so God knows we're going to kill you with questions. I know you've got some stuff you want to talk
  30. about, too. I'm going to start at the top with a personal question. All right. I'm in a in a dangerous position this
  31. year. My favorite professional team is at the top of many people's list and my favorite college team is at the top of
  32. many people's list. It's uncomfortable. You'd rather be there than anywhere else, but it's a little uncomfortable. I
  33. need to know, Aaron. I need to know what's the case for the Ravens. We're used to them being like second, third,
  34. fourth on most people's list. There's usually somebody like the Chiefs or the Bills or the defending champion Eagles
  35. up. They're look looking up at them. But now everyone, it's ridiculous how many people are saying the Ravens are the
  36. best positioned in the league. Now I know there it's tight. >> You say Kate, there was a time. Was
  37. there not a time that the Bills were your favorite team? Like when I first met you almost 30 years ago now.
  38. >> Well, you spend a couple years in Buffalo and and you get you get soft for the Bills and they're still way high on
  39. the list for sure. But there I I discovered, Eric, last year, remember they played in the playoffs and it's
  40. like I didn't know what was going to happen internally when I went to know >> that's how you found out and it was
  41. clear to me. I'm sorry, dear Buffalo friends. It was clear to me that recent recency matters and it's the Ravens. So
  42. Aaron, I need to hear your thoughts on the Ravens for this year's NFL. >> Well, if anybody who follows my numbers
  43. knows that they have absolutely loved the Ravens for the last two years. Uh they have been our number one team for
  44. two straight years and number one in a historically great way like among the top 10 or 12 teams ever measured in the
  45. regular season and then they go to the playoffs and they fall on their faces. >> So the idea that the Ravens are the best
  46. team is based on the idea that falling on your faces in the playoffs is not a real thing. that history is filled with
  47. teams and players that couldn't do it in the playoffs until the year they finally
  48. did >> get it done. >> And certainly we know that there's no magic sauce that John Har uh John
  49. Harbaugh can't win in the playoffs because the man won a Super Bowl 13 years ago.
  50. >> Is there some sort of magic sauce that Lamar Jackson can't win in the playoffs?
  51. I honestly >> don't think so. Has he struggled so far? Yes. Does it mean he'll struggle
  52. forever? Just ask Pton Manning or Jason Tatum or Barry Bonds or a number of different athletes if you struggle in
  53. the playoffs for your entire career or maybe it just happens early and then you kind of get with it. Now, the downside
  54. for the Ravens is that they are desperately trying to get past the Kansas City hump,
  55. except they also have to get past another team that is also desperately trying to get past the Kansas City hump,
  56. right? >> Which is the Buffalo Bills. It's a really interesting table we have in the
  57. book. Mhm. >> The top DVOA average of a five-year team that doesn't make the Super Bowl belongs
  58. to the last five years of the Buffalo Bills. >> The second highest average belongs to
  59. the last five years of the Baltimore Ravens. So these are what great teams would say historically get past a hump
  60. >> over what period of time are we talking about? >> This is over the uh since 1978.
  61. >> Oh my gosh. The best 5ear spans since 1978 to not make a Super Bowl are the Bills and
  62. Ravens of the last five years. >> Yeah. And we just place that on Andy Reid's doorstep and and Pat Mahomes
  63. doorstep. Geez. Good lord. Okay. Well, tell us why it's different this year. Why would it be different this year for
  64. either of those teams? >> I mean, it's just a question of, you know, when is somebody gonna get it done
  65. and and when is Kansas City not gonna have the brakes work for them? And I mean, I think Kansas City is a very well
  66. coached, very well quarterback team, but I don't think that they're invincible. They did show a couple of years ago that
  67. if somebody could get past them because Cincinnati did. And I think that they're
  68. not going to have the same luck they had last year. I think they may be better than they were last year, but they're
  69. not going to have the same luck. So, they're not necessarily going to be the number one seed in the playoffs. I think
  70. the most likely team to be the number one seed is the Buffalo Bills because they have the easiest schedule out of
  71. those three AFC top teams, >> right? I mean, Ravens visit Buffalo week one and then they go to Kansas City
  72. first half of the year sometime as well. It's >> that week one game is hugely important
  73. for playoff seating. It >> you also pointed out um it doesn't have to be this is what I've always thought.
  74. It doesn't have to be the Bills or the Ravens that defeat the Chiefs. As far as I know, Kansas City goes to
  75. San Diego or the whatever Harbaugh team. The other >> Harbaugh Chargers, LA Chargers. Yeah.
  76. No, I mean, we think Denver is very good this year. We think Cincinnati is good.
  77. >> Absolutely. The Bengals beat them. The Bengals beat them and then the Ravens and Bills, one of them wins. That's what
  78. I think is actually going to happen this year. Also, uh, it's guaranteed that some team from the AFC South will make
  79. the playoffs. >> That's Yep. That's the way the NFL works. Sure enough. Um, well, remind me
  80. what we learned from the Super Bowl. If I don't misreall, the Eagles beat down the Chiefs. I mean, not just a little
  81. bit. So, tell me what we learned from that about getting past the Chiefs. uh that offensive line is a weak link unit.
  82. >> Say what you mean more. Say more about what you mean by that. >> The idea is that your offensive line is
  83. generally only as good as its weakest lake. >> And therefore you need to be good at
  84. every position. And the Chiefs had Caliando at left guard and they had Tuni playing out of position at left tackle.
  85. And the Eagles pass rush destroyed them. And the other thing we learned, Vic Vangio is a very good defensive
  86. coordinator. >> That Eagles defense from week six onwards would have been one of the top 10
  87. defenses since 1978 by DVOA if you if you take out the first four games of the year. That's how good that defense was
  88. last year. >> All right. Well, let's use that as a pivot to the hometown Philadelphia
  89. Eagles. Are people sleeping on them this year? I know they're roughly a a a top,
  90. you know, I don't know, five team or so in most people's expectations, but some think Detroit going to clip them this
  91. year. What What is your What does the air and shots take? >> I'm the opposite. I actually think the
  92. Eagles are not quite going to be as good as people think they are. They are, I mean, a very good team, but this is
  93. something we talk about a lot. Offense is more predictable and more consistent than defense. The Eagles had the number
  94. one defense in the league. It is very unlikely they will have the number one defense in the league again, especially
  95. because using the equation that I use to look at personnel changes, they lost the
  96. most veteran personnel of any defense in the NFL this off seasonason. Miami was second. So, with all the players they've
  97. lost, they're depending on a lot of young guys. They're probably not going to be number one on defense again. Their
  98. offense in the regular season was not as good as it was in the playoffs. So, let's say that they have a little bit
  99. better offense than last year. And like the fourth best defense in the league, they also have the second hardest
  100. projected schedule after the Giants. >> That is not a 13 uh win team. >> That's a 10 or 11 win team.
  101. >> Okay. Before we go on, I want to go back to the first thing you said there, which
  102. was that offense is more predictable yeartoear than is defense. This is kind of the can this is canon now, I suppose,
  103. for football analytics, but it's everybody may not know that because it's only been observed, I don't know, the
  104. last five or 10 years perhaps, but can you give us the intuition for that? Because I think most people might have
  105. the opposite intuition. Something about offense feels more volatile, so you might expect it to be less predictable.
  106. Why is it that season to season defense is less reliable, less consistent than a
  107. team's offense is? >> We don't. >> Is it Is it turnovers? Because that's one stat that they've said is very hard
  108. to predict from year to year. Is Is it that part of defense, Aaron, or is it some other part?
  109. >> Yeah, turnovers regress towards the mean strongly and stronger on defense than on
  110. offense. That is absolutely true. A big part of it is the quarterback position because so much of the offense is
  111. managed by one player, right? It's very hard to imagine that a team quarterbacked by Patrick Mahomes is
  112. going to have a below average offense. Yet, a defense with Miles Garrett as its best player can be one of the top three
  113. defenses in the league one year and one of the bottom three defenses in the league the next year, despite the fact
  114. that Miles Garrett is still awesome in both seasons. That quarterback having so much control over the offense leads to
  115. much more consistency. >> Super interesting. So, it's more about offense reliability as opposed to
  116. defensive unreliability. Super interesting. Okay, I want to come back to that. I know you you worry a lot
  117. about a hard thing in analytics, which is how do we forecast a quarterback when he moves positions, but let's hold that
  118. question for one second because we left the team previews one team too early. We
  119. were talking about Eagles and if you don't like them as one of the top teams, you might even like the Commanders
  120. better in the East. They are coming and they've got one of the most exciting, if
  121. not the the most exciting young quarterbacks in the league in Jaylen Daniels. What do you expect from them
  122. this year? Yeah, that's that's our big surprise projection this year. We project the Commanders over the Eagles
  123. to win the NFC East. That is not to say that the Eagles are not a great team. I think they'll be the number one wildcard
  124. team. >> Okay. >> Um the Commanders are going to not be as good on fourth down. They were
  125. shockingly good on fourth down. That is going to regress. But I think they have balanced that out with some of the
  126. additions they've made on the offensive side of the ball. Debo Samuel and especially Laramie Tonsel to improve
  127. their offensive line. We think their defense can be a little bit better than last year. They were actually 23rd on
  128. our defensive ratings last year. So, we have them projected average like 17 >> and they have an easier schedule than
  129. the Eagles because of which teams in which divisions finished first and second last year
  130. >> and therefore we do actually have them projected ahead of the Eagles this year.
  131. >> Okay. >> So, Aaron, just two quick follow-ups to that. Um, a how much weight do you put?
  132. Um, I was at every Eagle playoff game this year. Um, the Eagles absolutely destroyed the Commanders in the
  133. playoffs. uh they ran through them. It I forget if they won by 30 or whatever the
  134. number was. How much weight does that put into anything? The second thing is the team that I didn't think the Eagles
  135. were much better than last year and got lucky to win. I was at that game too was
  136. the Rams. So where do you put the Rams? I mean Matt Stafford in that snowstorm, he was one pass away from them going to
  137. the Super Bowl. So I'm interested to hear, do you put any weight on the fact that the Eagles blew the Commanders out
  138. in the playoffs? And what do you think of the Rams? >> First of all, that's only two games.
  139. Let's say that there really is something about the way the Eagles play that the Commanders do not match up well with.
  140. There's still 15 other games that determine who wins the division. >> So, it could be
  141. >> predicting necessarily that they're going to beat them in the playoffs. You're talking about the division like
  142. >> Yes, I'm talking about the winning the division during the regular season. >> That's totally different. I agree.
  143. That's entirely different. >> And the Rams, we like the Rams. We do think their defense is going to take a
  144. step back. There's a lot of questions in the secondary. Um, their defense improved a lot last year and there's a
  145. sort of a plexiglass principle thing where they may come back to the pack a little bit. Uh, so we don't have the
  146. Rams. That's a very strong division. So, we also like the Cardinals this year and
  147. we also really like the 49ers this year and the Seahawks may be better than we think because their defense because Mike
  148. McDonald in Baltimore at least had his defense improve a lot in his second year there. So, that's a very strong
  149. division, but the Rams are a very strong team. >> Okay, let's let's talk QBs. Um Jaylen
  150. Daniels um after a spectacular year last year, it's it's you know because quarterbacks are so important to offense
  151. and offense is so important to a team. Understanding where a quarterback's going to be is the single most important
  152. component in forecasting a team. And yet especially early in a guy's career, it's
  153. hard to know what kind of steps he's going to take. What tell talk about Daniels? Like what do you see in
  154. Daniels? What do you expect in Daniels long term? What do you expect in Daniels this year? I mean, the best thing I can
  155. say about Daniels is that when a player is that good as a rookie, you don't expect that they'll take a step forward
  156. in year two because they were already so good as a rookie that they're not going
  157. to necessarily get better. >> Okay. >> Uh Marino was like the exception, but he's not necessarily going to get worse
  158. either. I think I mean he's really good last year. He's a great pocket passer and then he's an amazing scrambler. He's
  159. accurate. He's got a good clock in his head to not take sacks, right? You compare that to Caleb Williams. Now, the
  160. quarterback projection system that we use to try to project quarterbacks from college to the pros liked Caleb Williams
  161. better, but the fact is Caleb Williams just he made too much of his own pressure, took too many sacks, and
  162. didn't have anywhere near the kind of year that Jaden Daniels did. It's so hard to predict. Erin, tell me, talk
  163. about that because Caleb Williams, I mean, I would say I don't, we'd have to go through through the numbers, but he
  164. wasn't quite forecasted as a generational guy, but the consensus around him >> was he was number one,
  165. >> was so strong. I mean, I could just the consensus was as strong as it has been
  166. in many years. >> I agree with Aaron, by the way. I've seen both quarterbacks live, including
  167. JD Daniels first ever game, which was uh Commanders at Buccaneers, and the Bucks
  168. won the game. But I remember saying as I was leaving the stadium, "Thank the Lord the Bucks played them in
  169. week one. This guy's clock is amazing. This guy's accuracy. He's making the right reads." You a lot of times, as you
  170. know, Aaron, you go to games and you say, "This guy's open all over the field. What was the quarterback
  171. thinking?" I don't remember saying that once about Jaden Daniels during the game. I think I think you hit it on the
  172. head. I I don't know why that would regress. Why would his ability to find receivers and his I'll call it internal
  173. clock not holding on to the ball too long knowing when to run. I'm not sure why that would regress even though
  174. defense is quote unquote adjust. >> I mean the biggest reason why listen CJ Strad regressed a lot but he mostly
  175. regressed because the team around him he had injuries with his receivers and his
  176. offensive line completely fell apart. Washington has tried to make sure that does not happen by going out and getting
  177. the only good player from the Houston offensive line to improve themselves at left tackle, knocking last year's left
  178. tackle, Brandon Coleman, into left guard. So, they're hoping for a better blocking to make sure that that Daniels
  179. doesn't take the step back that Straoud took back. >> What What about Caleb Williams in
  180. Chicago? you know, expectations. It I I will say that as hard as it is to forecast a QB coming out of college, it
  181. seems to take us a number of years before we really understand a quarterback in the pros. We we get short
  182. on guys that end up turning it around at a second or third stop and we get long on guys that end up progressing three,
  183. four years into their career. What is your current position on Williams? What is your current forecast for having a
  184. >> I think he can be better. I think we may be making too much out of the new offensive lineman that Chicago got.
  185. They're gonna have better blocking, but the problem is still Caleb Williams as far as the pressure goes. It's not the
  186. offensive line. That being said, Ben Johnson is a very smart offensive coordinator. Now, head coach is a
  187. different job. We have no idea, none, how good Ben Johnson is at being a head coach, but part of his job is that he
  188. will be running the offense. That we know he's very good at. So, I mean, we do think Chicago in our projections,
  189. they come out as an above average offense this year. >> Okay. Okay. Well, we we've talked about
  190. projecting guys coming out of college and the challenge of that. Number one pick last year, Cam Ward going to be
  191. playing in the NFL for the first time. What hess in general? And then what do you see for
  192. Cam Ward? We like Cam Ward better than I think uh the consensus was that if Cam Ward had come out a year ago, he might
  193. have been the seventh quarterback taken. >> Yeah. Right. >> Our projections like him better than
  194. that. We have him more like third or fourth, like after Williams and Daniels. Um it, you know, it's based on the
  195. college performance. It's based on mobility in college. It's based a little bit on experience and a little bit on
  196. age. Now, the age thing is interesting because the change in how college is working means that age and experience
  197. may not mean the same things they've meant in the past. Guys are staying longer. The COVID year really
  198. complicated this, but now NIL complicates this. It used to be you'd see a guy was like 24 in college and
  199. you'd be like, "That guy is beating up on 18 and 19 year olds. How does that necessarily translate to the NFL?" Now,
  200. that guy is 24 in college, and a lot of the other players are going to be 24, too, because they're staying in college
  201. because of the NIL money. So, overaged prospects may not be as overaged as we've thought in the past. Cam Ward is
  202. not, by the way. Cam Ward is still a young guy, but this was an issue with Daniels. This was an issue with Bo
  203. Nicks. This was an issue with Michael Penn, etc. >> So, Aaron, let me ask you, as someone
  204. that does mathemat mathematical modeling, how do you then address that issue? I'm sure a lot of our listeners
  205. like in some sense we've seen a regime change in the NFL where you know the age distribution has changed but like we
  206. can't just project from the past because we haven't seen that age distribution before how do you even think about
  207. addressing it from an analytical perspective or do you assumption do you assume that you look
  208. at the 24 year olds in the past and you try to project them forward what do you do
  209. >> I think early on like now you just do your projection questions the way you've
  210. always done them and you just put words next to them. I mean, that's why we have
  211. words and not just tables in our book. You know, we explain when we think, you know, things are changing and and what
  212. it's going to mean now after we've seen three or four years of this. We'll have to see. Wide receivers is another
  213. position where this comes into play. Like we've always found wide receivers who come out as juniors are just better
  214. in the pros than similar wide receivers who come out as seniors. They just are. But what if guys stay for senior years
  215. because they can make $2 million and get college checks? Like you know like what
  216. that it's not that they're staying till their senior year because they're not as
  217. good. It's they're staying till their senior year because they're making lots of money. So, it's going to change how
  218. we see things. >> So, is that just kind of the art of it? And yet, I know, you know, as modelers,
  219. we try not to put our thumbs on the scale, but you do put your thumbs on the scale when you know your model's missing
  220. something. >> I mean, we we all right, I I've heard people talk about this. As much as we
  221. want to believe that our models are objective, the way you build your model is in its way its own way sort of
  222. subjective. So, yeah, I mean, we try not to put our thumb on the scale, but we also want to be as true to life as
  223. possible. and life has changed. >> That's right. That's right. That's the humility of it really. Um you recognize
  224. that it's imperfect and you got to sometimes bring a little subjectivity. We could talk quarterbacks all day long.
  225. I mean, another interesting one is JJ McCarthy. So, he's not I mean, he's not a rookie, but he is a rookie. And what
  226. what is a year sitting on the sidelines? He's not even getting snaps, right, because he's injured, but maybe that
  227. year of mental prep makes a difference. What what and the team is now completely
  228. dependent on him. What are your expectations for McCarthy? rejecting him like a rookie. There's never been
  229. anything like this before. >> Mhm. >> You've never had a quarterback who missed his entire first year because of
  230. injury. You've had guys who sat on the bench for their first year, but you never had a guy who was supposed to
  231. start and then got hurt and so you don't know what to expect from him. I mean, obviously there are guys who sat on the
  232. bench for a year and Patrick Mahomes in his second year, right, was insane. But in general, I think we have to treat
  233. him like he's a rookie. And that's in our projections. We treat him like he's a rookie.
  234. >> Okay. But you he came in in a strong QB draft. If he'd come out a year later, if
  235. he'd come out in the Cam Ward 2025 draft, >> he might have been the number one pick.
  236. >> He's he's that that's how high you have him coming out of their >> I mean that it's more like people are
  237. not as excited about Cam Ward. Remember Cam Ward, and this is another modern thing. He played for three different
  238. schools included including Incarnate Ward, >> right? Like he he's kind of a late
  239. bloomer. >> Yeah. >> So I think that a lot of scouty type people would have put McCarthy ahead of
  240. him. >> Okay. Okay. Okay. Another of these challenges from a modeling perspective
  241. is when coaches change and a lot of us have tried to do coaching modeling and and we've seen some efforts here and
  242. there. for a for a long time we just kind of called the residual coaching effects.
  243. It's somewhere in the residual but it can't be the entire residual. What are you doing with coaching these days? And
  244. when we see coaches move from one franchise to another what what can we say? >> It's hard because you don't have a
  245. history of except for a guy like Fangio, you don't have a history of, hey, he's coordinated 10 different defenses and
  246. this is what's happened to them, right? You're usually with a coach who's on a new team, your sample size of his past
  247. performance is either one or zero, right? Like we've never seen Ben Johnson run a team. We've never seen Aaron Glenn
  248. run a team. We have seen Mike Vrabel run a team, but only one team, only the Tennessee Titans. So, we know something
  249. about Mike Vrabel, but we don't necessarily know what was those Titans players and what was Vrabel. Now, Vable
  250. is an interesting uh one to talk about though because there's a flip of this which is there's no way to model like
  251. really bad coaches. Like there's no numbers that say that Gerrod Mayo was completely in over his head even though
  252. everybody says Gerrod Mayo was completely in over his head. So, it's possible that we are underplaying the
  253. change in coaching, not because Rael is more important than we think he is, but because Mayo was so bad.
  254. >> What if every Aaron ju just say, let me ask a question. What if every statistical model, what about if all of
  255. your models, all the other models, Massie, Peabody, whatever it is, had the Patriots last year at seven and 10 and
  256. they went three and whatever they went three and 14. Wouldn't that give some suggestion? I I wouldn't say it's
  257. perfect, but wouldn't we have some information that maybe Jared Mayo severely underperformed?
  258. >> Yeah, I think so. But here's an interesting thing. Um, so I have an equation called the postgame win
  259. expectancy. And what it does is it looks at the stats of a game and says based on the in
  260. the stats of this game, which team would we have expected to win? >> I love And there are certain coaches who
  261. have a history of outperforming what that equation says, like Mike Tomlin and Andy Reid. There's also a coach who has
  262. a really amazing history of underplaying what the postgame win expectancy would say, and he's considered one of the best
  263. coaches in the game. And that's Kyle Shanahan. >> Oh, I was going to guess Shawn McVey,
  264. but okay. >> No, it's Kyle Shanahan. So, what if does that mean Kyle Shanahan's not a good co?
  265. Right? Like there's a lot of measurements that you try to put into coaching, but I do think that if you
  266. have a consistent record of winning close games and winning more games than your underlying stats would suggest,
  267. it's probably a sign you're a good coach. And the opposite is also probably true. But that's not true of the
  268. Patriots last year. The Patriots did not tremendously underperform the underlying
  269. stats. The underlying stats were bad. Well, that's the thing. The coach gets some responsibility for that
  270. fundamental, >> right? I mean, when you hear that like Gerrod Mayo was not even aware of like
  271. which meetings he should be holding on which days, like you have to expect that that impacted the underlying stats.
  272. >> Yeah. Yeah, that's right. But that is an interesting that is one glimpse into
  273. coaching effects. It's not the whole picture and it's not perfect as you said, but as you say, as soon as you get
  274. a somewhat sufficient sample size, it says something. It says something about the coaching.
  275. >> Let me ask you, Aaron at all because I find this whole topic fascinating. Just
  276. one quick followup to that. >> Do you find that the same stats help you predict, let's call it expected outcome
  277. at different parts of the distribution? Like in other words, can I just have one
  278. set of stats that would predict a really bad team's outcome, a good team's outcome, and a great team's outcome or
  279. >> Yeah, I think it does. I mean I think using the same stats it uses you know DVOA it looks at penalties it looks at
  280. how many plays you run right like DVOA is more predictive which is an efficiency metric uh than how many plays
  281. you run but in a single game how many plays you run is important right like you want to run more plays that's not
  282. necessarily predictive for future performance but it helps you win that game but I think those underlying stats
  283. for like determining single games are basically the same whether you're a really good team or a bad team.
  284. >> Let me ask you one other follow-up question just for because we may have a big machine learning audience, blackbox
  285. audience out there. Um, when you think about these models or even you and your disciples, it's it was great to hear
  286. from you guys early on what your disciple tree is. Um, would you be happy today just building a machine learning
  287. black box with, you know, kind of throw away whatever you know about the NFL and
  288. whatever does best out of sample prediction, you're just going to jam it into some black box or do you see it as
  289. a blend of the two? Like what do you feel most comfortable with? >> I would say a blend. I will say that the
  290. uh the big data bowl a couple years ago when they did uh trying to predict the rushing yards on a play based on where
  291. the players were, the winners were two dudes from Europe who knew nothing about football and just put everything into a
  292. black box. So that was kind of interesting. I'm also I mean going to just be really honest here, which is I
  293. am not a trained data scientist. Even though I've been doing this for over 20 years, I'm very self-taught. I have an
  294. economics degree. There was no such thing as a data science degree in the mid1 1990s. The people who have come
  295. after me are much more adept at teaching machine learning to a computer than I would be. But I love the answer though
  296. um because it highlights that there are different kinds of problems in sports, never mind different kind of problem.
  297. There are even different kinds of problems within the NFL. So the example you gave was a really cool year where
  298. they were predicting the yards from from at any given moment on a rush rushing play like stop the action and say what's
  299. the expected yards from here. That's very different from saying are the 49ers going to be any good in 2025 with the
  300. rushing yardage forecast. You've got not infinite data but you've got a lot of data. You've got micros secondsonds as
  301. your unit of of analysis and you've got as much videotape from history as you can cram in essentially because there
  302. hasn't been that much change. That's a good environment for blackbox models. When it comes to are the 49ers going to
  303. be any good in 2025, the unit of analysis is essentially a season, a a team season and that's really chunky and
  304. really coarse and we don't have much data. That's not great for blackbox models.
  305. >> Yeah. I mean, I think there's a lot of subjective things when you look at whether teams are going to be better or
  306. worse from year to year and things like trying to figure out which personnel changes count and how to count them.
  307. >> Yeah. Right. Huge. And that's why I think that's why we're going to be having these kinds of conversations for
  308. the rest of our lives anyway. We're not going to have these things solved anytime soon. Okay, Erin, that's we've
  309. already taken more time than we should. Let's ask one last question. Give us one
  310. more storyline from 2025 NFL. one more storyline that you think is gonna be interesting, you're gonna have your eye
  311. on as the season unfolds that you think we should pay attention to, too. >> And Kate, I'm gonna get one more
  312. question in here that I really want to know the answer to, too. >> Okay. >> I'll mention, by the way, that all of
  313. these projections and stuff, we do the preseason book we've done for 21 years. So, I do want to point out to people the
  314. FTN football almanac 2025 is available on Amazon or a PDF version which costs a little less. You can get at
  315. ftnfantasy.com/almanac. Uh, one of my interesting picks for this year is that we like Jacksonville as our
  316. favorites in the AFC South. Partly for the same reason that we like Washington in the NFC East, which is the Eagles
  317. were the number one defense in the league last year. It's very unlikely they'll do that again. The Houston
  318. Texans were the number two defense in the league last year. Okay, >> it's very unlikely we'll do that again.
  319. We like the Jaguars defense to improve a little bit. We like the Jaguars offense
  320. to improve a little bit and we may be underelling what Liam can Cohen can do with that offense if he's a very good
  321. head coach. But the big variable here is Travis Hunter playing both ways, which is the
  322. most interesting story of the year as far as I'm concerned because it's it's only been done a couple of times in the
  323. last 30 years. Roy Green had a year where he played both wide receiver and safety. And Deion Sanders had a year
  324. where he played both wide receiver and cornerback. Nobody's ever tried to do it to the extent that Travis Hunter has
  325. talked about doing it for multiple years. It's fascinating. >> Okay. You did the whole Jack
  326. Jacksonville bit and didn't mention Trevor Lawrence. That's amazing. >> Well, that's the thing is the question.
  327. Can Liam Cohen unlock Trevor Lawrence like he unlocked Baker Mayfield? >> Okay, that's cool. That's good. That's
  328. great. All right, Eric, you last question here. >> Yeah, I just wanted to ask him. So,
  329. these are yes, no questions. We always say half the NFL teams that made the playoffs last year typically don't make
  330. it this year. So, if I'll just list the quickly, you just tell me yes, no. Do they make the playoffs? The Chiefs?
  331. >> Yes. >> The Bills? >> Yes. >> The Ravens? >> Yes. >> The Texans? >> No. >> The Chargers?
  332. >> No. >> The Steelers? >> No. The Broncos, >> yes. >> The Lions, >> yes. >> Eagles,
  333. >> yes. >> Rams, >> yes. >> Buccaneers, >> yes. >> Vikings, >> no. >> Commanders,
  334. >> yes. >> Packers, >> yes. >> That's >> That's the whole field. That's the whole
  335. field, man. >> That's great. >> All right, Eric, I just realized you're wearing your Bucks colors. This you may
  336. intentionally not >> no this is an R this is not >> Tampa Bay >> Mike Tir's chapter talks about Tampa Bay
  337. deserves more respect as a really good team and not just as a team that gets to go because somebody from that division
  338. gets to go. >> Okay, I'll believe that when I see it. But okay. >> And we love Baker Mayfield, let me tell
  339. you. >> All right. >> We love Levante David. We love Levante David. >> Erin Shots, great to talk to you, man.
  340. Great preview. Wish you the best. Fun time of year. Have fun for the next month as we ramp up.
  341. >> Absolutely, man. I hope everybody enjoys the NFL season. ftnfantasy.com/almanac
  342. to get the book. >> There you go. That has been Aaron Shots, formerly football outsiders, one of the
  343. real pioneers in football analytics. Welcome back to Wharton Moneyball. Welcome to the second half of this
  344. week's show. Kade Massie here, hosting with my longtime colleague, co-host Eric Bradlo. We are gonna do the second half
  345. of the show as an open topics half number of things to talk about but we just got off the line with Aaron Shots.
  346. Aaron is a frequent contributor here at Wharton Moneyball has been for all of our 11 years I'd say and is good sign
  347. that NFL is upon us. Football is upon us. And he does a great preview. I think Aaron does a fantastic preview. Super
  348. helpful, super insightful. Eric, any observations coming out or or let's say anything left to say about the NFL here
  349. about a month out? No, I I like I loved what he said. I mean, he said that offense is more persistent. Um,
  350. partially he said defense is less persistent because of turnovers. We've talked about that. There's actually
  351. reversion, mean reversion that typically happens in turnovers. Um, offense is more persistent because quarterbacks
  352. don't bo and weave that much. And so, I thought that was a very good insight. And so what he basically said was is
  353. that if you have a team that was average offensively but great defensively and like for example the Eagles and that's
  354. what made them a great team. You'd have to forecast some sort of regression towards that. And I think that's the
  355. thing I like about it too is, you know, it's in all the complex models and worlds we have now, that's something you
  356. can explain to somebody and it it doesn't it doesn't it's not just empirically true, but it makes sense.
  357. >> Yep. Um I found his team by team breakdowns helpful. It's interesting that they're a little short on the
  358. Eagles. If you go to the markets, our friend Neil Payne sometime the last two weeks summarized the market forecast
  359. like the probability of a team winning a Super Bowl. He got market forecasts from
  360. two places, FanDuel and Poly Market. This interesting new market, Poly Market, the for the numbers are very
  361. highly correlated. And you can look at what it says going into this year. We've got five teams at like 10% or higher.
  362. No, four teams. Four teams 10% or higher. And that's more than we've had in a few years. It's definitely more
  363. bunchy at the top. And despite Aaron's take, the Eagles are the highest, at least on the poly market side, at about
  364. 12% chance of winning the Super Bowl. >> Yeah. So, I love this bunchiness um that
  365. there's more teams at the top. Um I could debate just because I was there and you know, again, this is a bias that
  366. I have. I think people are underelling the Rams. Um, that was a damn good football team that came into
  367. Philadelphia and was one play away from going to the Super Bowl. Um, they're not
  368. that low, you know, they're like the seventh or eighth best team. >> Well, let's say what that is, but but as
  369. a as a fraction, it is so much smaller. It's like 4% or something. Yeah. Yeah. Yeah.
  370. >> Yeah. The other part that's interesting is a lot of that bunchiness, this is what we talked about with Aaron, is the
  371. Chiefs, Bills, and the Ravens, and only one of them Super Bowl. >> That's right. That's why the Eagles
  372. That's why the Eagles are popped up above them because that market poly market understands that and they're they
  373. just have less competition in their competition. >> Actually, you brought up a really
  374. important point because not a novice, but someone that's not thinking about it as carefully might say, "Well, of course
  375. that means the Eagles are the strongest team." No, the Eagles are in a weaker division. They have really one serious
  376. contender. Not that they don't have others, but the Lions I think people would consider in that top tier.
  377. >> Yeah. in the AFC. I mean, there's a lot of teams. There's a lot of teams that
  378. could be competitive and certainly the Ravens, Bills, and Chiefs. That big three,
  379. >> you know, three versus two tier, three versus two top tier teams changes your
  380. probability of going to the Super Bowl and winning dramatically. If you have one other top tier team, you have to
  381. beat one. If you have three, depending on your break and the playoff uh seating, you may have to beat two. And
  382. that's a big difference. Well, that that concentration is so high as well. So, three AFC teams above 10% to win the
  383. whole thing and then nobody down until somewhere below four. So, it's just heavily heavily weighted to the top. And
  384. and Aaron, maybe the most stunning thing Aaron said was that since 1978, so almost 50 years worth of data, the the
  385. two highest five-year runs of DVOA, which is his advanced analytics measure for quality of a football team, the two
  386. highest five-year runs without appearing in a Super Bowl belong to the current Bills and Ravens in the last 47 years.
  387. It's absolutely stunning. That's because Chiefs >> the last two years, you you might have a
  388. different opinion. I think it's hard to argue. I'm not arguing with the outcome.
  389. The outcome was the outcome. >> I think it's hard to argue the Chiefs were a better football team than the
  390. Ravens or Bills the last two years. I don't I'm happy to go back three, four, five years and say that, but not the
  391. last two. And I don't expect I do not expect the Chiefs to be a better football team. And let's even talk about
  392. Aaron also talked about postgame matchup beating expectations. I think the Chiefs are going to excel at
  393. that. You even mentioned they got Andy Reid. I mean, come on. That's worth a lot. I'm not saying John Harbaugh is not
  394. a very good coach. I'm not saying um Sean uh why can't I think of his last name? The Bills coach.
  395. >> Yeah, I never can do names either. Sean. >> Yeah. He's obviously a very good He's
  396. obviously a very good coach, but he's not Andy Reid. And so I think the Chiefs may be the third best team in the in the
  397. AFC. And I think right now I would put them as again since they still have Pat Mahomes, they're still the favorite to
  398. make the Super Bowl. Why not? >> All right, so we got we got we started our football previews a little bit early
  399. this year. We're going to we're going to slow walk them in. We're going to stage
  400. them in over the next few weeks. We've got a five weeks or so before we launch NFL. We've got some college football
  401. previews coming up in future weeks. Eric, I know you're just back from your annual pilgrimage with your cousin and
  402. the boys to the Hall of Fame inductions for baseball. How'd it go? What stood out to you?
  403. >> Well, couple interesting things that I heard. First, um, Billy, there's only eight, what was it, eight relievers that
  404. are in the Hall of Fame, full-time relievers. Billy Wagner's one of them. Billy Wagner's right-handed. He broke
  405. his right arm as a kid, so learned to pitch left-handed. So, he's a left-handed pitcher that made the Hall
  406. of Fame who's not left-handed. So, that was one thing I thought was interesting.
  407. >> I'm sorry. I don't understand what you're saying. You lost. >> He's right-handed.
  408. >> You mean pitches lefty? >> He's right-handed outside of baseball, but he pitches he pitches with his left
  409. hand. Okay, got it. >> That is correct. That's because he broke his arm as a youngster and couldn't
  410. pitch with his right arm. >> Okay. Okay. That was one thing. The second thing is that um
  411. Ichiro besides giving a hilarious speech um he had no hits until age 27. >> No major league baseball hits.
  412. >> Correct. >> Yeah. Yeah. Yeah. Yeah. >> And they actually published a list. They
  413. showed it actually up on the screen. I'll make it up. Like Tai Cobb had like 1,600.
  414. You know, there were 11 players that had a thousand of their 3,000 hits by the time they hit that age six. Wow.
  415. >> And Icho had zero. >> And of course, his first 10 seasons record, obviously, all 200 hit seasons,
  416. including his fourth year, we had set the all-time record of 262 hits in a season. Um, he averaged 236 hits per
  417. season in his first 10 seasons. So, that's a lot. Let me just real real quickly. 236
  418. is is pushing one and a half a game. And how many at bats do these guys expect to
  419. get? Not that many more than three. So, it's almost half of his at bats he's getting hits. Maybe maybe they get four
  420. at bats. Maybe it's >> maybe it would have to be. And actually, I have to look at his batting average
  421. after his first 10 seasons. I think his career batting average was 331. I'm just
  422. doing it by recollection. I'm not sure, but it could very well have been between
  423. 340 and 350, which would be exactly your math. If he got, you know, four at bats
  424. a game, 640 a season, and he got one and a half hits per game, that's hitting 350
  425. out of four. Okay. So, that could very well be. He may have had a 350 lifetime batting average after 10 seasons.
  426. >> What do you think his longest hitless streak was, but in games or or at bats?
  427. Well, I mean, if he's basically hitting onethird of the time, then the probability of not getting a hit in a
  428. game is uh well, he'd have to it's 2/3 to the fourth. So, 8 over 8 over 9 8 over 8110th probability. He doesn't get
  429. a hit in a game. So if you call that a geometric distribution, his the expect I don't know 10 I don't know uh games
  430. without a hit. I was going to say 10, but that seems way too high. >> That seems too high, right? The way you
  431. just went through it, that seems too high to be more than 50%. >> It does seem way too high because
  432. there's a 10% chance roughly that he won't, you know, 2/3* 2/3* 2/3 times 2/3, which is not hit, not hit, not hit,
  433. not hit. That's eight out of roughly that's one10enth. If you put a geometric distribution on that, like I'm flipping
  434. a one out of 10 coin. Um, oh, the waiting time. Yeah, the waiting. I I don't know. Maybe I'm computing the
  435. distribution of the maximum wrong. That's probably what I'm doing. I'm just probably thinking about the distribution
  436. of the maximum wrong. I don't know what it is. >> Okay, let's stop. Let's stop the math
  437. now and give an intuition. Like six games or so. >> Yeah, I'm going to say five or six
  438. games. >> Yeah, something like that. That's that's extraordinary. Okay. Um, but I
  439. interrupted your Hall of Fame thing. >> No, no, no. The only other thing, the only other thing I would say was that I
  440. also had forgotten when I looked at his stats how great a player Dave Parker was. You know, he was close. Not I don't
  441. know. He had 20 over 2700 career hits. I assumed his stats were much worse. And so I now have to wonder why he wasn't
  442. potentially in the Hall of Fame earlier. Um because just because I maybe it was a
  443. different era when people were voting and everything else like that, but I compare him to a lot of the players that
  444. have gotten in recently and you know to me he was as good I saw them both. He was as good as Andre Dawson. He was as
  445. good as Andre Dawson and you know he was as good as a lot of people that are currently in the Hall of Fame. So that
  446. was the other thing that struck me. >> Eric, remind me remind me what was so special about those Pirates teams back
  447. then with Willie Starel and Dave Parker like that. There's something that like the fish that ate Pittsburgh or whatever
  448. the what was so crazy about what was going on with those guys. I forget what was going on with those guys.
  449. >> Well, I think a bunch of it was just they were very interesting big time personalities. I think that was part of
  450. it. Also, they just had an amazing, you know, an amazing pitching staff, amazing
  451. hitting team, and it was also just, as you remember, Pittsburgh was title town back then, right? That was the Steelers
  452. uh era. >> Don't remind me. Don't remind me. No, I'm just saying that was a big era for
  453. the city of Pittsburgh. Um, and they were just I think um Pittsburgh again Dave Parker obviously just passed away
  454. two months ago. His son spoke. It was he was just saying it was a team with a lot
  455. of personality. >> Yeah. Fun. Fun. Okay. All right. Let's do a little bit of contemporary baseball
  456. because your boy Judge just hit the IR or whatever they call it and DL whichever one it is. And meantime that
  457. pesky catcher out of Seattle dinged a couple more. I think I'm liking my I think I'm liking my position on the on
  458. the Raleigh versus Judge bet. >> Well, now all of a sudden you're looking good. >> Hey, who's that? You could you had to
  459. factor in injuries. Judge is not exactly injury robust. >> Yeah, of course Judge's injury was from
  460. throwing the ball. And so the interesting part from the Yankees perspective, and you I forget if we
  461. talked about this last week or not, >> if Judge can't throw, then Judge can't play the outfield, which means Judge has
  462. to DH, which means you now put Stanton in the outfield because you you got to play Stanton. He's the second or third
  463. best hitter on the Yankees, maybe fourth. So you got to play him. And so now you put him in the outfield and then
  464. you get defensively much worse. So the they've got a lot of trouble. Um, I now put it at at best. I think Raleigh has
  465. 40. I think Judge has 37. I may be wrong, but I think he has 37. Judge is on the 10day injured list. If he comes
  466. back, I still like Judge, but not by much. He might be five behind by the time it happens. And there'll be only
  467. 55. Actually, I I don't know how I can say I like Judge. If he's five behind with 55 to play, his rate, well, it
  468. would have to be 0.1 more per game, which means over 162, I'd have to think he's a 16 better home run hitter than
  469. Raleigh, and that seems unlikely. >> It's a little strong, but on the >> on the on the other hand, these things
  470. do seem to be chunky. So, there's a lot of chance there, just you might not expect it. Um, but I think Rley got to
  471. 40. I may have this wrong. I thought he got two in a row relatively quickly. I thought he was at 41, but perhaps
  472. >> Oh, he could be. I just the last time I heard he was at 40. >> Okay, Eric, I want to give you some bits
  473. and bobs from around sports. >> 41. >> All right. So, he did sneak out there a little bit. Um, the things that have
  474. caught my eye are a little unusual. Maybe it's because it's the time of year, but we're a little bit in a in a
  475. in a Eb in major sports, but some of the interesting bits have been a little bit
  476. more businessy. We don't do a lot of sports business here, but some of these are businesses. Some of them are stuff
  477. from women's professional leagues. So, for example, I just saw, you know, the the the Euro the women's Euro finals was
  478. this past weekend, and England won. We saw some headlines around that, but I just saw the TV numbers for that. And
  479. the the the US numbers were the were a record for a women's soccer game. There's something like I'm missing the
  480. exact same like 1.4 four 1.5 million viewers, which is a 50% increase apparently over the 2022 Women's World
  481. Cup. And even the the the whole tournament, not just the final, but the whole tournament was averaging 458,000
  482. or something like that, which is a 100% increase over 2022. And people are thinking that this is because the NWSL
  483. has become so popular around here. And then those players go overseas and play in some of these um national tournaments
  484. and their interest goes with them. But so how when you just like raw numbers, this is one thing we ought to do more of
  485. Eric and I and I think we've been trying some in recent years is just to be calibrated on the number of eyeballs on
  486. various sporting events because that's what drives revenue in sports. And so we need to have some sense of what's a big
  487. event, what's not a big event. When you hear 1.4 4 1.5 viewers and it peaked at like 1.9 for the women's Euro final in
  488. the US. How does that strike you? >> First of all, I love women's soccer. I love watching it. I love the game. Um,
  489. in some ways I can um not associate, but I can like watch the game and you know identify more with it than the men's
  490. game which I can't identify with at all. Um, the women's game I think is a beautiful game to watch.
  491. >> So, hold on. Let me Rick, let me jump in on that. Um, my advisor used to say, he's a golfer. He used to
  492. say, I think we ought to all watch more women's golf instead of men's golf because we can't really expect to do
  493. what the men do, but our swings, if we if we tried to mimic the women's swing, we'd be better off.
  494. >> I completely agree with that. And I I played golf this weekend. I need to mimic something other than my own swing.
  495. But um you know I'm I'm surprised at that number. I'm not surprised at 1.35 million. I'm surprised that it's a
  496. record for women's soccer because you know if we think that there's I don't knowund 100 million US households that's
  497. still only a and you know I don't know how many TV viewing households there are. you know, I don't like I think the
  498. average NHL game gets about that number and so and so it's not large compared to
  499. other sports and premier games from other sports, but I'm glad to see that it's growing. And you're right, we are a
  500. business show as well and the business of sports is generated by TV revenue. That's the business of sports. The
  501. reason the NFL players get paid well, MLB players, etc. are the TV deals and packages. That's the primary source of
  502. revenue. And I hope this trend continues. If you would asked me to guess what the highest ever
  503. this game, I would have probably guessed the number around five to six million. >> Right. I definitely would have been
  504. three or four. And you make a very good point. I mean, can you imagine that in the, you know, the US women's
  505. championship in what was that 2000, the the the World Cup, 2000 against China, we had fewer than one and a half million
  506. Americans watching that match. That's really surprising, but apparently apparently true. Um, but the increases
  507. also the increase from the last Euro um championship are notable that that that we're jumping that high essentially
  508. year-over-year or tournament over tournament. >> Yeah. Just to let you know I this is the
  509. wonderful thing about large language models. So the average Major League Baseball viewership per game on Fox,
  510. ESPN, TBS, the national networks is 1.84 million viewers. >> 1.84. So the average MLB
  511. that's the average tele like all in a given night like all the games they're all they're a on a random Wednesday
  512. night or whatever they're averaging 1.8. That's probably better than I would have
  513. expected. >> And NFL, by the way, what's your guess for the NFL? >> Oh, it's an entirely different thing.
  514. So, I know I just saw the stat. You see variations on it, but like 72 of the highest 100 programs in 2024 were NFL
  515. games. >> Yeah, it's 17.5 million >> for the average NFL game. Yeah, exactly. Okay, Eric, what I think we're doing and
  516. we ought to do more of is basically the analytics of sports business. So, we like analytics, right? I don't care that
  517. much about business, but if we can make it more interesting analytically, but this is just fundamental calibration.
  518. This is this is what drives dollars. It drives signings. It drives everything. So, just be calibrated. We we can keep
  519. on trying to to be calibrated all. So, let me give you I don't have the numbers on this. Actually, did while I'm asking
  520. you this question, Eric, do the research on the average um TV audience for the WNBA game. Give us that. And then let me
  521. make the following observation. I just learned this. So, we know that the new TV deal is coming for NBA, WNBA, and
  522. they're expecting a whopping um increase. So, I think they're expecting something at two or $300 million. And
  523. that's more than a two-fold increase. It's I'm I'm getting the numbers wrong. It's a massive increase this coming
  524. season. And here's the interesting bit, Eric, and this is what I just learned in
  525. the past week. Because players and agents have known that that's when the new TV contract was going to come up.
  526. And because they've anticipated the numbers going much higher, no, everybody has set their contracts to expire this
  527. off season. So essentially, unless you're on a rookie deal, >> the entire the entire veteran crop of
  528. WNBA players is going to be up for free agency this off season. So, you're we're
  529. going to see something in the WNBA we've never seen before, which is a I don't know if it I think it's the majority of
  530. players are going to be up for signing with whatever team. It's going to be a super super interesting offseason.
  531. >> So, so if I had asked you to guess the ratio of NBA to WNBA average viewer, what would be your guess on the ratio?
  532. I'm not even asking you for the number, the ratio. >> Well, I'm not I don't walk around with
  533. high expectations for NBA audiences. My sense is that they're down a fair bit and they keep on coming down. So I think
  534. they may be as low as like 2 million or something. Um and so and I'm working to the ratio, Eric. I have to do it this
  535. way with basketball, the women's basketball, I'm going to go in the high six digits. And so I'm going to put it
  536. at about 3 to one. >> You're very well calibrated. You're not off by much. You're you're off, but not
  537. by much. Men's is at 2.68 million. higher than I thought. >> Women's is at 1.32. So it's two
  538. >> also higher than I thought. >> But also but I'm saying though I mean if you would asked me I would have guessed
  539. five or seven to one not just the NB WNBA. I would way overestimated the NBA. >> Yeah. Yeah. That No, I've been that's a
  540. story over the last few years that just keeps on coming down and people are amazed at the valuations keep on going
  541. up. No, but your point is is that you know the shirts that you and I both support that the women's shirts were pay
  542. us the damn money we deserve. If the ratio is 2 to1 the salaries might be 100 to one, 50 to1, 25.
  543. >> Yeah. Yeah. So that's that that's a that's a big topic and one we probably ought to tackle. Whenever they get into
  544. those formal negotiations, we should talk about it because it's a really interesting question. The biggest issue
  545. is the val the club franchise values are really going up in the WNBA. That's where the owners are doing great on
  546. franchise values. Revenue is is moving and it's about to move. It's about to move again this offseason, but it's not
  547. anything like the changes in franchise values. >> No, but Eric, hold on, hold on, hold on.
  548. This is the key point. Historically, >> professional athletes haven't participated in franchise increases.
  549. They participate through revenue. And it's a whole it's a whole question of whether they can get a part of franchise
  550. and that you that now we're talking about are we talking about labor versus capital. So it's this interesting
  551. philosophical question. >> No, no, I just loved your previous I I I appreciate it now even more than 30
  552. seconds ago or two minutes ago when you said it like you'd be a fool to sign a long-term WNBA contract right now. Why
  553. would you sign it now? I want my contract to be expiring. No, they that's and everyone saw it and so it's just
  554. going to be this amazing offseason moment where it's it's like be a free-for-all essentially resigning all
  555. these veteran players. All right, last one for you, Eric. And this is instead of a a business question, I'm go back to
  556. statistics, but I'm going to stay in the women's professional world. It's a little detail I saw in the bottom of
  557. this article. So, there's a young rookie who just made her debut, a PGA debut at
  558. the Scottish Open this past week, Lahi Wde. Lahi W. She's done real well as an amateur. She's done ridiculously well as
  559. an amateur, but this was her first um professional tournament and it counts as a PGA tournament even though it's the
  560. Scottish Open. She wins the dang thing. All right. So, that's great. We should all keep an eye on Lahi W. And by the
  561. way, the the women's I think it's the women's is it the US Open is this weekend? I think it's No, no, no. We've
  562. already had the US Open. Is it the British Open? Essentially, the Europe there's a major this weekend and I'm
  563. shame on me for not knowing exactly. I think that's coming up this weekend. But here's the question.
  564. >> She just won Scottish. It's probably the British because they probably do what
  565. the men do. Scottish than British, >> right? Okay. Here's my question for you. I get your I want to hear you talk this
  566. one out loud. What do you think in the 75 years of the LPGA? What's the longest streak of a new
  567. winner? What's the longest streak of new of between a a winner repeating anybody
  568. repeating? What's the longest streak of um not not first- time winners? I mean, just the longest streak of not having a
  569. repeat winner in 75 years of LPGA. And I'm tell >> you're talking about a given person the
  570. distance between their first win and second win. >> No. >> What are you referring to?
  571. >> I'm saying for anybody to repeat. So if you know, so Lahi W. If she if she won
  572. five weeks from now, that would be a four-week streak of No. Well, let's say four new winners and then Lahi
  573. Wins again. That would be a four-week streak of there not being a repeat winner.
  574. >> Oh. Oh. What's the longest streak of there not being a repeat winner? >> Yes. Yes. Yes.
  575. >> Oh. In the LPGA. >> In the LPGA. This is a harder problem than I usually give you guys. So, let's
  576. just talk about how what's your expectation. I'm going to tell you that they just hit their record mark in 75
  577. years. and I would have had no idea how to think about how to calculate. I'll give you the answer, but I first want to
  578. hear you talk about estimating. >> Don't give me the answer first. Let me think tell you how I think about it. So,
  579. um I don't I would imagine this may be this may already end me and doom me for too low a number. Um I doubt there's
  580. been any year on the LPGA tour where there hasn't been someone that's won twice. Now, if that's true, if that's
  581. true, I'm not saying it's true. If that's true, then the number has to be less than the number of tournaments
  582. in a year on the LPA LPGA tour. So what I'm computing the stat is the maximum number of wins by a player in a given
  583. season. And if that number is two for every season, the maximum of your statistic can't be more than about 25 or
  584. 27 or however number of tournaments there are in a given year. So that's why I'm going to guess somewhere in the high
  585. 20s because I don't think there's been an LPGA season whether it's Nelly Corda, Anukica Soren Stam, Nancy Lopez, Ro Jang
  586. or whoever the hottest golfer was in a given year. Somebody's won twice in that year and therefore that's how I'm going
  587. to bound the maximum from above. >> Well, that was my first thing I was going to say. You you didn't come up
  588. with a real answer. You just bound it. So you're saying less less than that >> down from there a little But but it's
  589. interesting to me how top down you went because usually we build things very very bottom up and you you just kind of
  590. went cluji in the middle which is fine. It gave you a bound. >> I could do the other thing too.
  591. >> Yeah. Let's try to go bottom up to come what? Let's give us a more precise number underneath that bound.
  592. >> Well, if I went from the bottom up, let's just say there's someone let's say a very good golfer has a 5% chance to
  593. win a given tournament. Right? So the expected weight time until the next win is geometric with probability 0.05. So
  594. the expected number is 20 20 weeks 20 events it would take that's not the distribution of the maximum number of
  595. weeks but I also have a large number of players I mean I could simulate this in about two seconds and see what the
  596. answer is. Um >> but that's hold on let's say because what you just did is really interesting.
  597. You said for a good player your expectation is that they might win every 20 tournaments or so, but then so
  598. but then we're talking about records, right? So we're like how what's the most extreme?
  599. >> I know. >> And and we're also talking about multiple players and so it kind of those
  600. go in different directions. >> Exactly. >> Multiple players are going to shorten
  601. that but the maximum is going to lengthen it. >> Tell me the answer is something like 60
  602. or something like that. >> No, no, no, no. Like look what you just did. For example,
  603. >> I'm staying with my number. I'm staying my bottom up and top down seem to have
  604. some sort of convergent validity and I'm going to stay with a number in the 20s somewhere.
  605. >> Well, I want to I want to say again that you're you you you did a real simple
  606. thing. You said one player, decent good player, 5% chance. I would expect that player to win every 20 tournaments. And
  607. then I said, well, there are two other factors. One, we're not looking for the expected. We're looking for that how
  608. what's the extreme version of that? The max of that that pushes it longer. On the other hand, we've got more than one
  609. player and so somebody's going to do it below their expectations. So that factor
  610. goes the other way. You might say just kind of heristically since I've got factors going both directions, I'm going
  611. to stand on 20. And had you done that, you'd have been shockingly close to the answer. Their record as of this week, it
  612. may go longer. It's not done. It could go longer. The record is 19. >> Yeah. So, but I also forget whether I
  613. was close or not close. I like well I was I like the fact that the top down and the bottom up both led to or the
  614. upper bound and then the more granular probabilistic led to similar answers. >> Well, cool. That was it's always fun for
  615. me to hear you hardcore basian statistitians work through the way you think of these things and sometimes it's
  616. very involved and sometimes it's heristic and I think both can be instructive. All right, why don't we end
  617. it there? Eric enjoyed the visit. Thank you for being here. Thanks in absentia from our colleagues Shane Jensen, Ali
  618. Winer. Thank you to the big boss Dion Simpkins for keeping the wheels on the track. We hope all is well today as you
  619. are away doing doing Deion Simpkins things. Thank you guys for listening as well. Come back and join us next time
  620. between now and then. Enjoy your sports. [Music]

Brief Summary

Aaron Shots discusses NFL analytics, team projections, and quarterback performances ahead of the upcoming season.

Key points

  • Welcome Back Aaron Shots. Aaron Shots returns to discuss the upcoming football season and analytics.
  • Ravens' Surprising Position. The Ravens are being touted as a top team this season, which is unusual.
  • Predictability in Football. Aaron Shots explains why offense is more predictable than defense in football analytics.
  • Commanders Over Eagles?. Aaron predicts the Commanders will win the NFC East over the Eagles this year.
  • Caleb Williams' Future in Chicago. Experts discuss the challenges of forecasting Caleb Williams' success in the NFL.
  • Cam Ward's NFL Potential. Cam Ward is projected to be a top quarterback despite mixed opinions.

Episode highlights

Tension

45/ 100Some tension

Related episodes