Wharton Moneyball Podcast – 10-Year Anniversary Episode

Knowledge at Wharton

Ask about Knowledge at Wharton

Episode Description

Wharton Moneyball – the popular Sirius XM radio show hosted by Eric Bradlow, Vice Dean of Analytics at Wharton, Shane Jensen, Professor of Statistics and Data Science, and Faculty Co-Director of the Wharton Sports Analytics and Business Initiative Cade Massey and Adi Wyner – celebrated its 10-year anniversary in March, marking a decade of insightful, free-flowing, and data-backed conversations about the world of sports and statistics. To celebrate, the group got together to chat about their memories of the past and ambitions for the show’s future. For more on Wharton Moneyball, visit: https://whr.tn/3UKeBAU

----- 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. hi this is Kade Massie practice Professor here at the Wharton School sitting with three of my closest faculty
  2. colleagues and longtime Wharton Moneyball collaborators Wharton Moneyball is a show that's been on Sirus
  3. xm's Business Radio since its beginning 10 years ago and we decided coming up on
  4. the 10 year Ann we're about a week away we thought we'd pause take a moment and think back on those 10 years especially
  5. because the world's kind of changed in those 10 years what do we think we've learned from it we spent a lot of time
  6. with each other over the last 10 years kind of a shocking amount actually doing the show so we thought we'd gather a
  7. little bit and talk about what that 10 years has meant to us I'm sitting with aie Wier professor of Statistics Eric
  8. bradow professor of marketing and statistics and Shane Jensen professor of Statistics as well a bunch of stats guys
  9. and we've been doing sports analytics we're also data science now we've been added that since we began the show oh
  10. there's a change in the last 10 years statisticians have become data scientists or at least some of them have
  11. we can talk a little bit about that that's an interesting development in fact why don't we start there why don't
  12. we start with the question of how you feel the World of Sports analytics has changed in the 10 years our first show
  13. was March 2014 so it's been 10 years has we've seen a few changes we're gonna pitch questions out we're all going to
  14. take a little bit of a chance Adam and we'll wander through the reflections but let's start there what has changed in
  15. the World of Sports analytics in the last 10 years well I think I mean Audi kind of
  16. alluded to it a little bit here I think certainly the public discourse around analytics and data and everything I mean
  17. in the last 20 years of me being a faculty member in the last 10 years of the show I think just sort of the public
  18. consciousness of of of kind of data has really changed a lot you know I no longer have to go and explain to people
  19. like why why I'm a why I'm a statistician why I do why I work with data I think so I think that really has
  20. changed way of saying that it's much cooler now it's caught up to the coolness of Shane probably but I think
  21. also on the other side I think Sports has also become cooler within the sort of statistical academic community in the
  22. time I've you know in the last 10 years as well I was kind of looking back I started as a faculty in ' 04 and ' 05
  23. was the first issue of the Journal of quantitative analysis of sports it was the first I think before when I was
  24. first coming up through graduate school I was interested in sports but I was cautioned a lot like do not you know do
  25. Sports as an academic topic you know maybe have it as a side hobby but don't make it your main thing and now we've
  26. got Flagship journals in in in in in sports anls and I think in the 10 years we've been doing this show I think you
  27. know Sports has become gone from kind of being this sort of Black Sheep to this something where this is where you see a
  28. lot of the coolest kind of spatial temporal data situations so it's almost kind of on The Cutting Edge of
  29. methodology now it certainly is accepted I remember when you arrived that's when
  30. I just got tenured that time and that was the first time I decided to work on Sports I waited until I had job security
  31. before I could work with sports I wouldn't I wouldn't give that advice anymore I'd say Jump Right In I think
  32. your comment Shane though about the field is always going to chase the coolest data and that's really I think
  33. the difference that's happened I think the reality is sports has data that Rivals really any other industry today
  34. and I think that's why Academia is moving in that direction I think that's why student interest is there I think
  35. also it's a great testing ground for learning statistical methods yeah it just is you know one particular kind of
  36. data that's blown up almost exactly in the window we've been doing this show is space show temporal so the kind the dawn
  37. of that in my mind you can pick different Don but one of the most important Don was when Kirk goldsbury
  38. got the data at Harvard from the NBA and he ran across Luke Bourne and I believe
  39. that was 2012 when Luke first got got there as a faculty member so that was less than two years but that was really
  40. the beginning of and the NBA was before the NFL that was just preceding our show
  41. so we've been there kind of from the beginning and we saw this thing blow up and now it's secondhand well I will say
  42. that Shane and our first paper on baseball was spatial temporal data evaluating Fielding but it was crappy
  43. data right right done by video and human beings and the enormous change was leaving this crappy imprecise sparse ER
  44. filled in incomplete data to Now tracking but it Al also changed because in baseball the best data was always
  45. public and now the best data is private and this is this is one of the things that has changed I think not for the
  46. better we sitting here in Academia could produced Cutting Edge research that was
  47. a value to the teams back then and I think at this point if you don't have proprietary data you're not saying
  48. anything the teams want yeah I think it kind of it goes hand in hand because I think as the DAT has gotten richer or
  49. the data has gotten very expensive to collect and is is incredibly rich but is now kind of like often sort of owned by
  50. a league or something like that it's less kind of accessible I think maybe Kate to answer your question for me um I
  51. was hoping someone would say something I didn't say which Shane did um I think the part that's gotten most interesting
  52. to me is the problems meaning if you think about the original Money Ball right it was mainly about onfield stuff
  53. like how are we going to get you know let's go back to Billy Bean and Moneyball how are we going to get
  54. players that you know maximize our chances of winning but now if you think about the role of analytics I'm even
  55. thinking about just the papers the people here have written it's not just the onfield part it's about the business
  56. of sports it's about you know trying to make decisions on sleep and training patterns Etc which is to me the narrow
  57. set of problems which was let's attack should we go for it on fourth and one you know are walks as valuable as
  58. singles I'm not saying those aren't important problems but the problems have grown so diverse and that's the part
  59. that's exciting to me and I think that's what's really changed is that the business side like when I started
  60. working for the Eagles the onfield side wanted something to do with me but the business side wanted nothing to do with
  61. me and now both sides tend to have analytics so that to me when I was thinking about the question I think the
  62. problems are broader and a lot more interesting and the maybe also to build on Shane's earlier point and and what AI
  63. just said the Divide now between Academia and practice of all the fields that I see you know my role as Vice dean
  64. of analytics here I think Sports has the closest connection between Academia and
  65. practice that I've seen if I think about people analytics and Neuroscience or even big tech companies there's still
  66. this divide but I think now there's this real tightness between Academia and practice when it comes to sports and
  67. statistics and I think that's been fantastic I mean how many students are you sending to sports teams how many
  68. times are we working with people with sports teams whe it's the NFL Big Data bowl or some other problem I think that
  69. divide has really shrunk and that makes to me Sports Antics the most exciting field to be in so just for to finish up
  70. I would more or less say the same things statistics has always been very close to sports because we can develop
  71. new methods and really Advance the field of Statistics as while using Sports data
  72. so it really is a full-on statistics data science problem just using Sports and this has engaged so many people at
  73. all levels um but the I have to say just I can give you a number it wouldn't be a
  74. statistic show without a number when I first taught Moneyball Academy which was our high school sta statistics program
  75. um u high school sports analytics program over the summer the first summer we barely had more applicants than we
  76. had space and now we have which is how much space uh about 75 but the first time we ran we ran for 50 we had 55
  77. applications for 50 spots okay and this past year we have two rounds and we have
  78. already 350 and we I don't think we're going to do a second round right that's just show how it incredibly um the
  79. interest in sports analytics even at the Young level is just exploded MH MH the one element that you alluded to Eric
  80. that I just I think is worth underscoring is that you're so right that in the beginning we were talking
  81. about more or less in-game decision making and Personnel really the two the play evaluation player evaluation and
  82. the I think the principle third that has evolved really only in the last seven or
  83. eight years is performance enhancement performance changes and and the classic examples of the Astros bringing in these
  84. pictures and teaching them to throw a different way and that's a use of Technology analytics that just wasn't
  85. happening at all and it's entirely different from the other from the other two or as we've had guests on the show
  86. where everything in your practice now these people have sensors all over their body so when is somebody tiring when is
  87. someone getting their Peak Performance and of course as scientist what we care about is how do you maximize performance
  88. which means if you can run an experiment or manipulate things like let's try different you know eating patterns
  89. sleeping patterns Etc and let's see how it manifests itself on the field there's
  90. no reason why that shouldn't be part part of sports analytics today you know sensors all over the body is really good
  91. data MH and I think that kind of immediate feedback you know they be being able to kind of train in a way
  92. where you know if you're a p hitter and you're trying to H learn how to hit a fast ball you can kind of just have like
  93. a lot of kind of data support like and and kind of immediately learn sort of like meth you know kind of mechanically
  94. what you're doing right or wrong or free throw shooting or whatever there's so many examples where we now kind of have
  95. almost IM you know immediate feedback coaching and these kind of really helps Player Development mhm
  96. MH okay that question was about how the field has changed this next one's more or less how y'all have changed so what
  97. have you learned in the 10 years doing the show because of the show what do you think you have what's an example of
  98. something that you've learned I can go I'll go first so um I've learned a lot by listening to all
  99. of you and uh the thing that kept coming into my mind since I knew you'd asked the question was base rates which is for
  100. me to move away from a base rate so for example let's imagine I took two arbitrary teams just playing each other
  101. I didn't know anything about them well 50/50 right okay so now I have these other factors I bring in there so the
  102. first thing is to move away from the base rate of Shane always talks about the coin flipping model to move away
  103. from that you better have a good reason and also I was probably one of those people that thought effect sizes like
  104. how much something really affected these probabilities was much bigger than I I now think it's much smaller like I used
  105. to think well it's a 9010 game you know Clayton kers on the Dodgers are playing I don't know the lowly Pirates 90% for
  106. the Dodgers no there's never you'd never predict you think you're different now than if we were sitting here 10 years
  107. ago I do I think I'm different because not you you were you were a cheered professor of marketing statistics 47
  108. years old 10 years ago and you were like why you keep bringing up the age what you were 47 10 years ago
  109. too we're all born in 1967 not I think I was only 37 I'm I'm not I'm not being critical at
  110. all I'm honestly querying that's phenomenal that that's been a consequence of these 10 years I do I
  111. think again it's both base rates in other words to move away from just or a always says if I have an empirical
  112. frequency let's say I've got say you know I've watched 10,000 baseball games over 40 years here's the empirical
  113. frequency something happens someone asked you to make a prediction that's a damn good starting place and to move
  114. very far away from that you better have a reason and I also think I adjusted too
  115. much like well it's a left-handed pitcher against this team momentum yeah well we'll momentum will get to I still
  116. believe in momentum that part hasn't changed but I think I believe in base rates and I don't move off them as far
  117. as I think I used to so let me let me just say what I've learned because it it kind of matches in some level what you
  118. were saying so when we started this show um I knew a lot about baseball I still know a lot about baseball but I was
  119. really terrified how we were going to feel two 2 hourss talking about sports that it can't be all about baseball
  120. because what am I going to say and uh I've certainly learned a lot about sports I think everyone is proud of my
  121. accomplishments in the football arena in particular um but all sports especially
  122. but the way you're able to do it is basically bring those base rate facts and questions to any sport so I might
  123. not know anything about golf but I can contribute by asking what's the mean what's the standard deviation what's the
  124. what's the a typical player doing these situations and sometimes what I found is
  125. my lack of knowledge about a sport works very well with the experts because it brings you back to Earth it it stops you
  126. from from overt telling the story and and forces you to ask a specific question and so one of the things that's
  127. amazing is that that it was I don't know how long it was but after a while we realize we can fill two hours no problem
  128. right AUD I just want to point out you're connecting to Something in Psychology called the inside view and
  129. the outside view the inside view people have very case-based detailed enriched models of the world the outside view is
  130. pretty much about Bas rights and often the inside view can get skewed by their cases and the outside view is really
  131. helpful correction in that way um Shane yeah I've been thinking about this too and actually it's a little related to
  132. what Eric was talking about a little bit too I kind of feel like I I when I with
  133. the show first started I mean you know I was trained as a statistician um I feel like I'm decent
  134. at probabilistic kind of thinking and thinking in the presence of probability but I have learned repeatedly over this
  135. decade of my own kind of how easy it is for like like biases and probabilistic thinking creep in I kind of I think of
  136. the kind of I don't know what inevitability kind of bias that I have you bring it up like every college
  137. football season right where we kind of like halfway through the season we're locked into a national championship game
  138. no not that when he asked at the beginning of the Season what probability do you think is assigned to these four
  139. teams and I'll say like 15% in case like it's like less than 1% like total overestimation and I think I've I've
  140. learned from myself that in situations especially like you know I think now of like you know the Kansas City Chiefs and
  141. the current Dynasty in in situations where there's kind of you know a dominant team or something like that my
  142. own mind I tend to like when you've got like something like a probability of0 75
  143. or 08 and you round it up to one or you round it to you you take a probability that's in like the 15% or 10% range you
  144. round you just your mind just BL rounds it down to zero and so kind of thinking about those sort of like you know like
  145. even intermediary probability events like a 20% you know a real mismatch in sports would be like an 80 to 20 thing
  146. thinking you know not thinking about that as in an inevitable Victory like you know thinking about that 20% and
  147. kind of processing it you know every time you see I think that's a lesson I've kind of learned and it's been
  148. instructive too to sort of see like me looking at the Kansas City Chiefs as an inevitable Dynasty it's because I'm not
  149. a fan following that team and seeing all remembering all the gut wrenches losses
  150. they've had over that time as well I think it's sort of like when you're really closely following when you're
  151. involved in a sport maybe you're less prone to that inevitability byus cuz you're kind of of seeing you know seeing
  152. those 20% turn up maybe that's interesting right yeah I'll give you a couple quick ones uh one I've I'm struck
  153. by how many times we figure things out collectively that we'll start kind of Meandering towards some subject and
  154. we'll talk about it for seven or eight minutes and after about 10 minutes we kind of have a better understanding
  155. collectively and it's connected to my other one which is everyone's wrong sometimes and I've learned that another
  156. environments one of the nice things about working in on faculty is that you see some really smart people be wrong
  157. it's helpful to see that we're all wrong sometimes you guys are some of the sharpest guys I have conversations with
  158. and you're all wrong sometimes well let's just take what the facts are I'm wrong more than all of you because you
  159. remember we've actually no no no for years remember we've picked outcomes of games and kept scores I think I have a I
  160. have momentum that I'm the lowest every single year we've done matter of fact I don't think we've done this in a few
  161. years matter I think it's cuz you guys want to keep me at the bottom I we need to go back
  162. last place every year I just tell people I'm in fourth place and like that sounds
  163. impressive I don't tell them there's only four of us making these predictions well the redeeming thing is we haven't
  164. exactly been comprehensive in our record keeping so for all we know you might be
  165. number one and we just hav written down way well the last thing that I feel like
  166. I've learned is the power of doing something a little bit of something on a regular basis I think the the cumulative
  167. effect of doing this show even though it's only been two hours a week and now just one hour a week but we've done it
  168. 50 weeks a year for 10 years the cumulative effect in terms of what I know my relationship with y'all the
  169. people we know in industry is amazing we we what's the line like we tend to overestimate what we can do in one big
  170. push and underestimate what we can do with a sustained regular push I think this Show's been a like a tangible
  171. demonstration of that to me um all right what about themes that have emerged on our show we we have come
  172. to realize we didn't sit out with any themes we didn't know how we were going to talk about these things at all but
  173. after a few years we kind of come back to some of the same bits you guys have just talked about base rates for example
  174. that's definitely been a theme what else would you say or some patterns or some themes well come on it has to be what AI
  175. mentioned I mean you guys I ridicule is a fine word but in a professional nice way you guys ridicule me about momentum
  176. I mean that's definitely been and you know I'm a strong believer that um there is heavy if you want to call it a heavy
  177. degree of State dependence like given a success there's a much larger put a fancier term to make momentum
  178. sound more sophisticated no I think moment I still believe momentum exists but I'm willing to say that there's at
  179. least heavy non-stationarity and whether you want to call that momentum or not matter of fact
  180. i' I've joked as this I think you guys know but our listeners know like I'm from the old school basian World which
  181. means I believe in Cross unit crossers heterogenity what I think I'm going to spend the next 20 years of my career on
  182. is within person heterogenity which is if you think about something we talk about like a hidden Markov model like
  183. people go through hot and cold States well that's about a given unit that's not a cross-unit statement I think we in
  184. the show have spent a lot of time talking about whether you want to call it momentum non-stationarity Etc and I
  185. think it's a great thing because I think a lot of times we focus on differences across units and not as much sometimes
  186. there's Peak periods and weak periods within a unit let me jump in because that that's my favorite one that's
  187. emerged over time I can name a few others but that one I think is the most interesting because it really has been a
  188. slow thing to emerge and now it seems to me terrifically important we come back to it repeatedly you always named the
  189. connection to marketing there but within person variation and one of the reasons
  190. it means so much to me is that I think it's something that we miss like people in the world underestimate it we tend in
  191. the world to focus on betweeners differences and we're talking about mean differences and we underestimate the
  192. within-person variance well you know my favorite topic on that is when we talk about whether it's tennis player or not
  193. is that as players age you draw more from a bodal distribution so we're even know the greatest tennis player of all
  194. time we could debate it's probably Novak jokovic I begrudgingly have to admit that but at the end of the day as
  195. someone ages you start to see well maybe 95% of the time we'll see a draw from the good distribution and a 5% of the
  196. time it'll be bad and then maybe next year because he's 37 38 it'll be 90% 10% and that's what happens this is
  197. definitely an Eric theme this an Eric theme he's he's learning how toing from me for me he's but he's learning how to
  198. talk about it in a more structured way a real quickly but but this is something that you've given us an example of it's
  199. with pitcher this is an empirical observation you have about baseball pitchers that we tend to think oh this
  200. is how you rank them who's the best who's the next Who's the differences and you're like yeah yeah fine but a given
  201. picture looks very different on different days right they do not only do they look very different on any given
  202. day but any really almost any pitcher can have a fantastic game in any in any one moment and this is the kind of
  203. things we we've Ted aled about variability across Sports this is the kind of thing that happens with baseball
  204. pitching constantly doesn't happen so much with with quarterbacks you expect good quarterbacks to have fantastic
  205. games obviously it does happen every now and then but that kind of comparison across Sports amount of variability but
  206. if I were to answer like what what theme that we've I mean is and this is I guess so important in
  207. in basic statistics we think about regression and that just drives just about everything we do and I don't mean
  208. it just in regression as the forecasting tool because the word regression is generally used today to mean your
  209. prediction your predicted value what is you how did you regress this but the word actually comes from to regress
  210. which is to go back down and in sports they use the word constantly and you and we it's a tool for forecasting but our
  211. so much of our dialogue is is concentrated on how much to regress and H and how what data we use to to figure
  212. out and and what are we regressing to in fact our just our last show what are we
  213. regressing to to individuals rate to the base rate of the of the of the of the the the community the group and some of
  214. our finest conversations have to do with trying to understand how much regression
  215. and that of course intercepts with our basian conversations like how do we how do we integrate priors into that and so
  216. it's a really a great amalgam of some of the most important themes of Statistics
  217. regression basian shrinkage is is the more the more complex way to think about it and we've become artists in
  218. describing these hard I think so hard Concepts in our Sports and I think it's made me a much better teacher
  219. because I use it in class even not necessarily with sports data has it made you a better Reasoner I I think it does
  220. do you think you're more you're less prone to make that mistake you're more prone to remember to strength your
  221. forecast because of how often we talk begrudgingly sometimes particularly when it comes to my
  222. Yankees I think and I think I've also kind of like I I guess you could call it maybe a cynicism or something like that
  223. but I've kind of developed sort of like a little bit more of I think a protection against the sort of some of
  224. the recency biases like when you you know when when you see some unusual performance or some new thing hit a
  225. sport you know are you just looking at kind of the random variation of a particular pitcher like suddenly having
  226. a stand up performance is that an example of a non-stationary thing where we can see that you know we can predict
  227. that for that pitcher going forward or are we just kind of biased by the recency of a really stand out
  228. performance I think back one of our um you know back 2015 I think when the Kansas City Royals won the World Series
  229. we spent like weeks after that talking about how they've revolutionized baseball with the way they like got to
  230. the World Series with like you know shut down relief pitching and fielding and all these types of things and you know I
  231. mean they did do all that but how much that that actually you know how much we should have really regressed even more
  232. those kind of Trends towards you know the mean because it was it it wasn't maintained you know can you can you go
  233. back 10 years can you remember the first show we're in a basement here at the Wharton School we were in a different
  234. basement when we started before the Studio existed what were you thinking as we Eric rounded us up let's give credit
  235. to Eric I remember sitting his office the first time and then we end up in the studio over there Dion was there on day
  236. one what were youall thinking on day one remember I'll tell you what I was thinking because I remember it I
  237. remember two things first of all I remember being nervous speaking live remember it's been a while since we've
  238. done a live show but the very first show was a live radio broadcast there's no opportunity to fix what we did and I
  239. remember being concerned about that remember being concerned about saying something stupid that that you couldn't
  240. take back or or have some tick or whatever so that was the first thing the second thing was really bothered me I'm
  241. curious to know what your thoughts I didn't know how we were going to fill two hours um and and that's of course
  242. been the the most pleasant surprise I feel like we can we can fill two hours in fact there was a a crazy time where
  243. none of you were there and all the the entire sound went out and there was no Communications with and it was I was
  244. left alone it was and I it was live and I had to talk for a half an hour to no one about nothing for a half hour and it
  245. was no problem that might be a professorial skill that you've developed over time
  246. it's possible yeah I was never worried about us filling up time at all not in the slightest because in my case I just
  247. watch so many different sports and so much Sports and I'm just used to also taking notes when I'm watching sports on
  248. things that also to me the beauty of sports for me is that it's the ultimate Trojan Horse to talk about any
  249. statistical topic like should you go for it on fourth and one should the coach do
  250. this at this point in time should you go for three should you foul in the last 10
  251. seconds but every one of those I could give and I'm sure you guys could do the same could give an entire lecture just
  252. on that topic so I was never ever worried about it and also my my view was um the one thing I've learned through
  253. teaching and you mentioned about its effect on teaching and and the radio's really changed it in that way too is
  254. that I figured if I'm having fun the audience is having fun and that's always been my motto for teaching as well and
  255. part of the reason I brought this group together I mean we've known each other longer than I've actually known both of
  256. them and my comment was is that we were going to have fun no matter what talking
  257. about sports and statistics and if we had fun I figured it would be fine no whether Sirius XM would keep us on the
  258. air for a month a year or now 10 years I actually wasn't worried about that I was
  259. just like if we're having a good conversation and we're having not only fun but learning something if we're
  260. learning something the audience will learn something too but Eric you taught that to us I mean I learned that from
  261. you that we can talk about sports at a high level and teach statistics and do this for two hours with no problem but I
  262. remember not knowing that at the time no and I remember that I I can Echo the nervousness of you know I mean again
  263. we've learned since then also that you're you're encyclopedic knowledge of sports I can I can say stupid sh every
  264. show and you're there to kind of as you know sort of a uh instant fact Checker so that that was reassuring and I also
  265. do think I I kind of worried for a while that we would sort of be able to kind of
  266. generate to our you know like that amount of material every week but as we started doing the show and we started
  267. seeing these kind of various themes popping up I started seeing them everywhere like I you know I would be
  268. you know I'd be you start looking for them you start looking for them and you and you you do sort of like kind of
  269. enabled me to sort of see connections be between sports at like kind of the base
  270. you know between versus Within ation these kind of basic levels that you know basically allowed me to kind of I think
  271. Branch out to sports like tennis that I didn't have much experience with before the show even started Shane you're
  272. talking about between sports makes me think about um tournament design as a a a late answer to an emerging theme I
  273. never knew I was so interested in tournament design until this show came along what about favorite moments we
  274. talked about how you were thinking about it 10 years ago 10 years of shows any favorite moments come to mind I'm sure
  275. we've got lot but like what comes to mind without no one's keep keeping track or score but what might come to mind
  276. well just cuz we were talking about the early days in vance H I think one of the
  277. I think maybe the first Super Bowl kind of re I think it was Super Bowl 49 the the Malcolm Butler interception was one
  278. of my favorite shows we came in I mean obviously you know my my team won it all but it was more coming in and just sort
  279. of like it was early on in our history it just seemed like just kind of the perfect time to talk about something
  280. that had like obviously a lot of consequence but also like a lot of analytical strategy discuss and just
  281. seeing that kind of moment being discussed in the like 10 years to follow and how how that kind of discussion of
  282. that particular like play has been informed by 10 years of kind of you know increasingly sophisticated analysis that
  283. that play has stayed the test of time let me just say it's still discussed today I I think uh I told AI just before
  284. the show taping here what I was going to talk about but part of his just I love seeing your reaction each time Kade I'm
  285. going to talk about Joey Chestnut no no yes to bring that up just I mean again I
  286. want to say for our listeners here I want to say it again Hot Dog Eating is a real sport you can train for it and I'll
  287. tell you the thing I like about it as well is that you can measure atypicality and that's the thing is that
  288. you can measure how many how much how many more hot dogs can this person eat than that one and when Joey chessnut can
  289. eat 20 25 more than other people and people have now especially now he's been like the 15 time mustard belt Champion
  290. why haven't people caught up to him I'm just saying you have to you have to admire someone who's his exceedence is
  291. so far greater than everybody else that to me so part of it is just seeing your reaction I just love it because I in my
  292. heart I'm not just saying it to piss you off I really think it's a sport and I love talking and when he came on the air
  293. and he talks about his training methods and everything else I'm like this a real
  294. scientist working here 20 years from at our 20th anniversary Kate have totally turned around on this I think no he
  295. won't you're not you're not there yet I can tell already so any other favorite moments a so I have a bunch of favorite
  296. moments that so it's interesting because I don't not really concentrating on the
  297. guest per se because we've had some recurrent guests who are incredible um and but the moments for me are when I
  298. learned something that's shockingly new to me so I remember when Rick Peterson oh told something to us remind us Rick
  299. Peterson was was a was a pitching coach longtime pitching coach was the Oakland A's and he was on our show quite
  300. frequently in the beginning regular guest as a regular guest and uh we used to bring him in for about 10 minutes and
  301. as a baseball analyst one of the big strategic questions are why you not bringing in your relievers at the high
  302. leverage moments and we analysts were always saying this is just dumb how come they have this s conventional way of
  303. doing things and you should you should do some things differently and he said you know you guys that's that's fine to
  304. do this but you realize that there's all this actual onfield issues that you don't have any idea about like a pitcher
  305. has to warm up and once you've warmed them up and you don't bring them in we might not be able to use them again and
  306. we were sitting here listening to this and going oh and I remember feeling really schooled by an actual baseball
  307. profession like a real Egghead moment you and like boy I think we we have a lot to say but we can't Implement
  308. without actual onfield expertise okay and that's just a general lesson that we should be not just knowing ourselves but
  309. preaching to other statisticians and data that's right it's a humility that that Rick was saying get over yourselves
  310. on this you don't have the full story and that was that moment of course there are other moments as well like when mie
  311. Betts told us that the real reason why Bill Buckner let the ball go through his legs because he was running and it got
  312. him so nervous I enjoyed that moment wilsonon sorry everyone I screwed that up mie Wilson
  313. yes you're talking about uh Rick Peterson and and giving some background that that was I opening to you reminds
  314. me out of the blue of what Chris Collinsworth told us when we talked to him about Bill bellich he said he was at
  315. some practice the Patriots practice early in the season and he came away thinking that belich was actually
  316. burning some wins early in the season in order to rotate defensive strategies and
  317. to train his players to play multiple strategies at a way that might cost them early season in order to learn to be
  318. better later in the season this was Collinsworth interpretation of what belich did and it was like oh wow that's
  319. interesting and it did I do remember that now that you say it one of those really eye opening um I have another one
  320. too and it's a surprising one for me it's one of our horse racing interviews no it was the one where we
  321. were talking about like you know what really take makes a horse like a a true rener he's like oh it's the horse with
  322. the biggest heart and we were all like all of us cynically were like oh you know we've heard that kind of wants it
  323. the horse wants it the most but he was literally talking about biologically the horse with the largest heart is the one
  324. most likely to win the race and I thought that was you know that was is kind of a fun little that's a top three
  325. that's a consensus top three all time also in that same interview he mentioned that also every horse slows down during
  326. the race just who slows down the least it's kind of like the sinking fastball kind of which also was a great
  327. moment one of my favorites was I think a category that I like is um when our guests have little either they share
  328. they have an epiphany or they share a moment and we had one I think the first time we ever had Brian Burke on the show
  329. so Brian is an ESPN analyst and one of the guys who had been on The Cutting Edge of football analytics for a while
  330. Brian had been an engineer by training he flew for the military and then he got into football analytics but his
  331. background had him very black and white and when he got into football analytics he started seeing Shades of Gray he
  332. literally said I had to start thinking probabilistically for the first time in my life and this is a you know highly
  333. accomplished mid-30s probably guy and he was saying that's what stats and thinking about sports analytics did for
  334. me it turned me from black and white Reasoner to Shades of Gray and probability so what you just said maybe
  335. think about something that I'd love Audi to talk about the work you're doing with
  336. Ryan Brill on the fourth down in just a second but the thing that we also I surprised it didn't come up as one of
  337. our themes also is is the role of uncertainty which is even if you put a probability on something that's your
  338. point estimate of the probability like probability is 60% you should go for it on fourth and one okay but the
  339. uncertainty is plus or minus 10% on and so one of the things I think I meaning with a coin we know with Precision what
  340. that probability is but with these real world team events we're estimating something and and in general the you
  341. know the classic rule of thumb is well in fact I think what I'm about to say is probably a theorem but audio correct me
  342. which is by definition you're always underestimating the probability like in order you could say well a fair coin's
  343. 5050 maybe I don't know that like maybe the way you flip it I'm just saying you can
  344. always say there's some Ed factor that you're not putting in there but I think just to me the role of uncertainty even
  345. in things that are uncertain we don't take into account enough yeah and we have to evaluate that and we've learned
  346. to do that over the years remarkably um and but sometimes you have to do it fast
  347. but in particular you're talking about so bril Wier right so we we we came to this idea saying looking at to figure
  348. out like what is the what is the uncertainty in the models and what we really disc models you're talking about
  349. are the fourth down models we really discovered that the data isn't really that rich which means that given data is
  350. a constraint and as a result you can come up with an estimate but if you had gotten a different set of data different
  351. 2,000 games you might ask well how might the estimate change and that's really what What's called the confidence
  352. syndrome we realized that those things were really wide almost shockingly wide because there's not much data in
  353. football because there's just not that much data in football which led us to the which connected with the Rick
  354. Peterson comment which says wait a minute if we don't really know we have to have some humility about what we
  355. don't know just cuz I've got a point estimate doesn't mean that that's the correct estimate and I can't oversell it
  356. to the team because they have information that I don't have and that information should be more important
  357. particularly when I don't have good information and I've got to let them let them go for it okay this connects to two
  358. other of my favorite Point moments on the show one is it connects back to our regress your forecast and one of my
  359. favorite standout moments me in memory was the last bit of the show the last show before the 2016 presidential
  360. election so we've done a fair bit of analytics on the presidential election leading up to that and if yall remember
  361. 538 famously had Hillary Clinton as I don't know 80% or something likely to win that election and the last thing
  362. that happened on that show was AI saying ah you know I'm gonna I'm gonna go with
  363. something I'm be like really like how much less I don't 67 I play that clip to my class every election year yes I do I
  364. put it on there because it's so 538s was 80% but almost every other forecast was
  365. like 99% yeah they were actually the most humble Sil just said I just don't feel like it
  366. and he did something completely unpraised dung for Hillary Clinton but something
  367. doesn't feel right and we had a conversation about what doesn't feel right on the show and I was trying to
  368. argue that it's got to be a lot lower her probability of winning is a lot lower and although we had no I had no
  369. hard data to go on I was adjust looking at previous elections within that calendar cycle in other countries where
  370. the populous candidate was coming out big and it had and and overturning the forecast that was a great moment so I I
  371. what I like about it is that that's probably that's 2016 so we're only two years into the show we're talking about
  372. we're we're all cultivating this tendency to regress and pay attention to base rates and the question was always
  373. like are we like be you is that true because in that moment I wasn't there's no way I was thinking regress that
  374. forecast it was only when AI said I'm going to regress that forecast and now because of that moment and because of
  375. the reinforcement that that came about from the it does start maybe bringing us to that
  376. habit but it's so easy to say it intellectually but in the moment we don't tend to do it you did it in that
  377. moment and that was one of my kind of what I talked about inevitability bias or rounding those 80% up to one that was
  378. a mo that was a real lesson for me you know that was kind of a a real learning moment I want to name one other that's
  379. connected to this and that is uh and it may be another question we can pursue in
  380. a seconds like we've read some books every now and then for some of the people we interview and we read a book
  381. called called escape from model Land by Erica Thompson and this is written by a modeler cautioning modelers to not get
  382. too caught up in your model like you got to recognize what we all love our models
  383. and we begin to think the world is our model but the model is only an imperfect capture of the world and we have we have
  384. to get out of that and it connects exactly what ai's saying about what's your data you know how how sure you of
  385. these estimates you got to have that humility it really comes back to that humility well you may also remember for
  386. years I asked Our Guest the following question I always used to end when my when I got my time to speak I'd always
  387. ask the following question you could have one of three things better you could have better data better model or
  388. better let's call it internal buyin I never in all our years doing it I never heard someone say a better model like
  389. wow you're right my random Forest my XG boost models just not working well enough no no I need a better what you
  390. always hear is either it would be greater to have more impact at my company because there's better
  391. connection or I'd rather have a randomized experiment or you know I have Spar data on fourth and one I'd really
  392. rather have an INF never did someone say a more sophisticated model that can give
  393. me another third significant digit on something ever that's you I didn't I've heard you answer that asked that
  394. question in lots of environments on the show and off I've stolen it on occasion myself I didn't know that you had done
  395. it enough to begin drawing inference about it and I love that no one ever said the model ever matter I'm pretty
  396. sure the number could be zero but it's interesting because you talked to all the the representatives of of startup
  397. companies that are selling data and selling and selling pipelines like we just talked to people from data bricks
  398. and we have talked to another company uh just called shot quality they aren't selling models nobody's selling models
  399. they just and and I always wonder why they're not because there's a there's an edge there right they're selling you
  400. know serious strategy advice and data and they just layer on top of it somebody's random forest or deep net
  401. deep neural net just something we're sitting here 10 years in D Wharton Moneyball all the all the
  402. co-hosts longtime collaborators here taking a moment to think about how the world has changed over those 10 years
  403. how we've changed some of the things we've learned maybe some of the highlights we were just talking about
  404. Erica Thompson's escape from modeland which came out she was studying partly covid and talking about models in covid
  405. there was this moment in time guys where we were almost actually we were pretty much full-time covid analysts so March
  406. 2020 hit and we start talking about Co making sense of Co at some point we decided we're going to stay on this and
  407. we're going to dedicate the first half hour I think of every week to co what are your Reflections now that's been
  408. we've we shifted back more than two years ago now but it was a long stretch when you think back on it now three
  409. months was an hour we didn't have any sports remember that that the beginning that's right the first three months it
  410. was all pandemic all the time what Reflections do you have on that stretch of Wharton I I remember it
  411. extremely well because I really dived into the data deeply you became our expert and um but just I mean the lesson
  412. that modeland the book which came out much later we learned on the Fly because there were modelers coming out with
  413. forecasts for the spread of covid and forecasts on diseases and on hospitalizations and everything that we
  414. wanted and policy was being built on it and those forecasts were terrible and it
  415. was the best Minds doing the best job with the data they had and you couldn't do anything couldn't do it and it was
  416. just an incredible humbling lesson um where I was disappointed was that people would put out models with their
  417. prediction intervals and their prediction intervals nobody's overlapped and that said to me this is a problem if
  418. we have 10 different forecasts about where we're supposed to be going down the line really time series forecasts
  419. with confidence bands or prediction bands on top of them and they were all so depressingly narrow and we looked at
  420. them and this is just bad so not only were the forecasts bad but much more importantly the uncertainty was horribly
  421. underestimated and that just kept appearing over and over again as we move through the pandemic it was almost
  422. ridiculously tragic how little we we could say with any accuracy I was going to say the one thing I remember though
  423. is that some things were relatively stable like what I do remember is because I remember each week I was the
  424. one that was looking at the CDC data and just seeing like in some sense like the
  425. death rate conditional in you're getting covid was actually fairly constant for a
  426. long period of time now of course the question is how many people are going to get covid which people are going to get
  427. covid what's the risk for a certain population but there was a period where I think the number was somewhere around
  428. as I have a vague recollection maybe it was 1% or 08 of 1% where if you looked at every country and you said how many
  429. people have Co reported report hug I understand reported and how many people were dying that number had a very narrow
  430. Bandon interval for a fairly long period of time so one thing that was interesting because we talk about
  431. moments when David Spiegel halter came in as a guest later not there that day um which he was amazing he was one of
  432. the sharpest Minds on covid early earliest on he had a beautiful infection fatality curve that was by age that
  433. turned out to be just about exactly right and he published that in like March of 2020 and we we publicized it on
  434. our show and we talked about it and many people came up to us that you were on we
  435. provided solid good data on Co that they weren't getting anywhere else I think that's what I'd like to hear a little
  436. bit more from you is like setting aside the vagaries of modeling the pandemic how what what do we do well what do we
  437. not do well in that show I me did we stay too long do we spend too much time to little time like how did we because
  438. that was a incredible moment of uncertainty and I think one of the things we did was we worked together to
  439. reduce that uncertainty we were trying to make sense of it like together that's I think one of the virtues of what we
  440. did what what are your thoughts on and it gave us a vantage point at least for me I mean it was obviously I think
  441. probably a frustrating time for all of us but it was particularly frustrating for me kind of along kind of you know
  442. Eric's way of kind of thinking about things would you rather have good data good models or or Buy in that was a time
  443. in society where although we you know could talk and we were mostly focused on evaluting the models what we really
  444. needed was better data and what we really needed was actual buyin you know on on public policy kind of initiatives
  445. and so it was kind of I mean a little bit of therap therapy for me but also like you know actual like instructive to
  446. kind of come in and hear mostly from Audi like kind of the what was really kind of going on in the kind of modeling
  447. sphere and also to kind of just just be able to kind of I I don't know Express both of my kind of frustration but also
  448. interest in in what was kind of going on at a societal level at the time through
  449. kind of the lens of analytics and through the lens of trying to kind of think about what we could actually
  450. measure you know without bias Etc I have to admit though I also I think back to that period I also have to think about
  451. um how I remember thinking to myself how you know we're in we're exactly the same
  452. age and so I remember there were times where you would say yeah you know what I would go out and have outside I'd have
  453. some sort of party with friends and I remember I said say it on the air like oh I would never have done that our
  454. internal no no no that was fascinating to me and and by the way I don't think of myself as overly risk averse in any
  455. particular way and and you were you were our expert so all most of the information I was getting I was getting
  456. from you just my interpretation of it was like I'm not going to go for that really even it's a low probability event
  457. it's just not worth it and that to me I I will remember all the decisions you guys made and like like I was proud like
  458. wow I got my like 80th shot not's like I don't know if you should be getting that
  459. many shots and I'm like I've got 80 go ahead oh I was just going to say I can't really it was a lesson how that
  460. distribution of like kind of risk tolerance versus risk adversity how much there is across people within a person
  461. even over time in that kind of in in in that it was fascinating one of the things we learned about covid during
  462. that time which has been really useful is the importance of observational data is distinct from experimental data and
  463. in sports we talk about confounding it's just the most important Concept in all any evaluation of a a complex game where
  464. everybody interacts with each other particularly basketball soccer football much less so baseball but even that to a
  465. degree and what this forced us to constantly ask when we looked at a study well these aren't these aren't
  466. experiments and so much public policy and so much discussion around covid was built around observational data some of
  467. it terribly done some of it better done and we were able to explain that to to our listeners to each other and make
  468. sense out of it and this was this was really one of the most and I think it really carried us through it it puts
  469. fart in Center the scientific method which is at heart really what statistics is serving the the acquisition of
  470. knowledge and truth okay so I everything you said I have to agree with entirely of course but it also feels like one of
  471. the things that happened in the pandemic is that experimental evidence by itself was
  472. too isolated to really explain what happens in the complicated World it goes the other direction as well we have
  473. experiments well you can't experiment with you have a hard time experimenting with societal uptake of policies and one
  474. and the biggest confounder especially early on was people didn't act the way they were
  475. supposed to act in the models and we wouldn't have known that I don't think from an experiment so in this case I
  476. think it goes the difficult is you need to go both ways we had to explain this that why these weren't predictable what
  477. was the issues and we did I think that was one of our our strongest points with within the field of Statistics as you
  478. know nobody would consider me an expert specifically in causal inference that's not really the domain I operate in but I
  479. kind of felt like you know with was happen with my friends and kind of with with the L people I was talking to like
  480. having to explain just kind of very basic what you can learn from an experiment what you can not learn from
  481. you know observ study is controlled observational study all right guys let's shift out of the pandemic and do a
  482. lightning round before we end up with a couple of final questions lightening round back to favorite moments but let's
  483. look at some specific uh some specific narrow aspects of the show um any favorite guest moments or interviews any
  484. any things jump out to you over the years we've talked about some so far I mean you've had some big baseball
  485. guys that you just love having bunny chance we had Ron Bloomberg which I enjoyed
  486. immensely we I we I don't know if Matt was the producer then but Sam Gwyn wrote a book about HAL mummy the the the
  487. famous football coach who was kind of the beginning of the air raid and Sam Gwen and so I had him on the show I
  488. didn't know it was going to be on the show I had just read an earlier book of his that is really about the history of
  489. Texas that was profound for me and I come in some Wednesday morning and Matt's lined up Sam Gwen for the
  490. interview seg like oh my God I just this guy just wrote this amazing book just random you know pleasure any other guest
  491. jump out well when we uh were on radio roll for uh one of the Super Bowl I think the Super Bowl in Miami um and we
  492. got to interview Justin Tuck on the show and I was wearing the same Tom Brady coat and just the absolute look of
  493. disgust on his face was kind of a a Gade Mo that was a great interview to because
  494. again Justin Tuck you know is not I don't think the most you know necessarily analytically minded like
  495. athlete but it was a really informative kind of discussion because we talked a lot about you know I mean he was you
  496. know on the defense that went up you know went up against some great teams in the Super Bowl and just sort of the way
  497. he thought about you know battling Brady and the way he thought about other quarterbacks I just remember it being a
  498. really cool discussion I loved also the moments we've talked about we could call
  499. them the more secondary sports like when we've had people on golf talk about the
  500. golf Analytics and you know like the smarter players know where to forget they know they're going to be inaccurate
  501. so they choose a zone that they not going to cause thems as much damage like you could go for the left side of the
  502. green but you know you should shoot for the right side of the green because here's how much variance there is in
  503. your shot that's something that really has always stuck with me that was one of my most eye-opening interviews you i'
  504. had forgotten about that that was one of those moments where I'm listening to this guy talk and it's Scott faucet
  505. Scott faucet is the golf golf coach he played some golf golf coach now in the Dallas Fort Worth area and we chase him
  506. down because of an article in the newspaper and he starts talked on our show about these guys have to act like
  507. they're going to hit it perfectly because you got to have confidence but they have to plan knowing that they
  508. can't hit it perfectly even the best golfers in the world he was talking about accepting uncertainty and
  509. accommodating uncertain in your decision- making it was absolutely profound moment loved it yeah so I have
  510. a few but I remember some of my favorites were David Epstein he's written two of my favorite books on
  511. Sports one is a sports Gene and I interviewed him about that book I may have been alone this was a long time ago
  512. and then he had a second book range which he talked to us about the importance of athletes trying out
  513. different sports and how that's important for for Success that was just those were incredible interviews and
  514. then of course I remember um Annie Duke's first interview with us she came live in the studio she she was writing a
  515. book uh thinking in bets and she talked about poker and reading opponents and I remember one thing she said to us and
  516. and we've she's been on many times she comes from our our our my Moneyball Academy every summer um but she talked
  517. about how the importance of exactly the opposite of what we do which is you know
  518. calmly think about data and regress but when you're in a procer tournament you got to size up that opponent immediately
  519. you don't have time to wait and watch their re their play and accumulate data and move yourself off the prior you've
  520. got to look at everything they they do how they touch their chips how they talk to each other to quickly figure out what
  521. type of type almost cluster type of player because if you get it wrong you're going to get hammered and it's
  522. like very observable just the expertise that goes into being a professional poker player that's much more Beyond
  523. just the the strategy of just do doing I remember that too it was super insightful I I I kind of found myself
  524. coming back to that because you know it was really about what she was talking about is like very body language that as
  525. a as kind of a professional in the field you learn you know you really immerse yourself in when coming out of Co one
  526. thing or going through Co and coming out of it I kind of from my own teaching I got the I realized just how much I feed
  527. off the body like you you are subtly measuring the body language of people like teaching in front of people in
  528. person is a very different kind of endeavor because you can read the body language of people and very s you know I
  529. I was I more kind of I I could perceive what I was picking up on in very sort of
  530. subtle signs and I think that's probably you know for poker players that kind of
  531. become that as their vocation that almost becomes probably second nature in the way I didn't even realize I was
  532. doing it as a teacher a favorite offsite moment well well I can start with one you mentioned Justin Tuck before but on
  533. our offsite moment I was quoting I think Rufus Peabody or Kade Massie talking about how you predict
  534. future football games and how no no player can change the line when they go down except for a quarterback otherwise
  535. it just doesn't affect it we're not saying those players aren't important we just we can't measure it and I told this
  536. to Justus tuck and he looked at me and this very imposing human being looked like he was ready to take me down like
  537. what are you saying I don't matter and here I am trying to scramble wait a minute of course that's what I'm saying
  538. yeah yeah one of one of mine was at the Miami Super Bowl we ended up at a bar on South
  539. Beach late after dinner and Eric eager was there with somebody of his from the Kansas City area yeah but this was
  540. January or early February 2020 which means pandemic was already in the air but Eric had been using some um some
  541. diffusion model based on pandemics in one of his Sports papers and so he was studying pandemic models in the months
  542. prior to this and he sits there late at night drinking some monster Margarita or
  543. whatever it was on South Beach and he said I think y'all should be looking at this pandemic thing but you know what I
  544. took out of that lesson that that conversation I think we should be talking talking to Eric eager that's
  545. right he's not only been on our guest many times but he's now a longtime collaborator with us at Mone Moneyball
  546. Academy so that was an incredible offsite moment that's right that's right honestly I was thinking of that uh
  547. actual I think that same evening talking with Eric eager because I was thinking about that kind of in the cont Tex cuz
  548. that was 2020 right before the pandemic that was right before you know the chiefs were about to play the 49ers I
  549. think that was the Super Bowl and he was already talking about a Chief's Dynasty
  550. and I'm like this guy is this guy's getting a little ahead of himself what you know what does he know this guy's a
  551. blow heart and you know it turns out well I went to dinner that night with you guys and the pre- dinner bar but I
  552. went home before you guys went out with Eric eager so I don't remember that you traveling with family I was traveling
  553. with family but what I do I when I think about it is actually really I give a lot
  554. of credit to SiriusXM I remember the it was probably the time we interviewed Justin Chuck and like we were there on
  555. radio row and I have to admit we looked legitimate like there was a big Serious XM sign behind us and we were up there
  556. and interviewers looking over like these are some serious broadcast I was wearing
  557. this Cod so I mean you guys it actually made me feel good that like you know I make it up CBS Sports
  558. was across the way and like and there's us four guys Wharton Money Ball and we're there in radio row I I felt like a
  559. legitimate broadcaster for a minute do you remember when they had the NFL draft here in Philadelphia I do and we went
  560. did the show on site beforehand but then we started walking around backstage and
  561. we end up walking through the dadgum Green Room players sitting around with their families were like yeah we're not
  562. supposed to be in here well it's funny because some of my favorite off-site moments was going down to spring
  563. training bringing our equipment talk about being somewhere you're not supposed to be right and and I get
  564. myself press access and I wander into the Yankee Dugout and the Yankee Dugout and
  565. the Yankee Clubhouse and I'm walking around with all the real reporters I'm going what do I do with myself so I go
  566. find CC Sabathia and I'm like no one's talking to him he's which they were I talked to him and I'm trying to explain
  567. the opener to CC Sabathia did you tell us remind us what the opener so the opener is an idea that I've talked about
  568. on the show for years is that the baseball team can get an advantage by beginning the game with one of their
  569. good very good relief pitchers for influence perhaps when he was at Toronto not Toronto Tampa Bay Rays I talked to him
  570. just they were the first team to really do it but I was trying to evangelize on it and I said to Cece you know you
  571. should be in favor of the opener and he's like no we're starters we we're we're trained to start the game and he's
  572. an imposing human being and and it's just a ridiculous idea that I would start the game um in anything other than
  573. the first inning and I said well you know has some advantage and I start to explain them and I one of the things I
  574. said was you'll have an opport opportunity to get the win without having to pitch five innings and he just
  575. without even missing a beat he says I thought you stat heads don't value wins that is
  576. awesome there's so much in this story absolutely incredible um okay last lightning round question Eric you
  577. mentioned the less popular sports like golf but what about non sports are there the pandemic we do sometimes wander off
  578. the playing fields into other areas any favorite non sport moment and to give yall a moment to think about I'll tell
  579. you I'll tell you my favorite it's the Moneyball for fire guys the forest fire guys we we just them well we
  580. hypothesized that they must this was the summer of I think it might have been 20120 and bad forest fires in the
  581. American Northwest and we thought someone is running analytics on that and we asked Matt to run it down and he's
  582. like shed I found these guys who were publishing papers one of which is called Moneyball for fire and now we have a
  583. multi-year relationship with these guys we've had on the the show multiple times
  584. and they fight the same battles the exact same culture battles that we see fought in football organizations
  585. basketball organizations baseball organizations it's incredible yeah I mean talk about a cool spatial temporal
  586. data situation that basically pops up every year and links to so much of the kind of climatic things that are going
  587. on right now as well it's one of my favorite non non-sports topics was the traffic engineer we had on years years
  588. ago right before Thanksgiving you may not remember this I don't think I was on this show you may not have been on it I
  589. wasn't alone I know there was a few of us there remember um and he was talking about money essentially Moneyball for
  590. Designing highways and one of his comments was that we in America don't do things properly we just build wider
  591. roads when we want to accommodate more traffic but that can be accomplished by learning how to
  592. drive really we just don't drive properly huh and he explained the various ways in principle he was saying
  593. stay out of the left lane unless you're passing and how much trouble that causes
  594. and and he explained how the models show how much how your roads become inefficient when people drive improperly
  595. and he was of course saying but in Germany where he was a German engineer we don't build wide roads we teach
  596. people to drive properly speaking of Germans we have the DI the diet researcher out of Duke whose name I'm
  597. going to forget but he's given us some great stuff over the years on calories really good stuff all right we need to
  598. wrap up and so I want to ask one last question we've been talking about the last 10 years of course and
  599. appropriately hopefully we'll do another 10 years and More in front of us what is
  600. one hope you have or ambition you have for the show over the next 10 years I think it's going to be really cool five
  601. six seven years from now when we all have Hall of Fame votes and we can on the air we argue about the Hall of Fame
  602. for weeks every year won't it be great as voting members to do that I'd like us to do an offsite at the hall of fame I
  603. mean I'm saying that self Baseball Hall of Fame but a matter of fact we could do
  604. all of them I mean actually I have you can actually do the baseball basketball Football Hall of Fame in one swing I
  605. mean there's not that far from each other the major sports I don't even know where's the Hockey Hall of Fame but
  606. Toronto all right well it's not it's not totally out of the question sounds either way um I have such a passion for
  607. the history of baseball especially and the Baseball Hall of Fame that I think it would be great to do a show from
  608. there and I'm confident that we could get a number of Hall of Famers to come on the air and I'm sure they would talk
  609. to us about the role that analytics played in their careers and it would be great to get a spectrum you know getting
  610. someone whose major career was in the 70s and 80s in the pre-analytics ER then get someone in the 90s and 2000s when it
  611. was emerging and then get even the most recent Hall of Famers that would be a ton of fun yeah I mean I sorry to just
  612. piggyback on that like I I I as I was sort of thinking about the the the last 10 years I feel like we we do kind of
  613. you know because we've been doing this a while I think we have a historical perspective things on things I would
  614. like this to do even more of a kind of like you know this year like you know 10 years ago what was going on both in like
  615. sports itself and also in analytics you know I think so many you know it the the
  616. field has moved so quickly in say the last 20 years that you know kind of like you know we're the last keepers of the
  617. generation without internet or whatever we have we we have an obligation to remind people what it was like and where
  618. we Shane what we may find out in the next 10 years is you know Audi was talking about you know running more
  619. sophisticated models maybe artificial intelligence and these large language models are going to change the way we do
  620. even Sports analytics like maybe we're going to be able to just upload a massive amount of video and other stuff
  621. and just say you know run you know compress this using a Vari variation encoder and just run a large just run
  622. some analysis on this and it'll get automatically done I have a feeling that 10 years from now the the use of video
  623. data and other types of data will be much more prevalent in sports because of all these technology and AI that's going
  624. on I think that's a fascinating uh prediction um I of course I have to second Eric's Hall of Fame show I mean
  625. we're kind of a force in this department it's my obsession and and just because loving baseball and baseball history in
  626. particular so I'm just seconding that one um but if I just say I'd like to see something in the next 10 years um I'd
  627. like us to be to reflect a little bit about what's happened on on the sports analytics side kind of uh on the
  628. research end and kind of reflect on our particular area which is the academic contribution of of sports analytics and
  629. methods and just reflect on some of the great papers the great contributors interview some of the scientists
  630. involved we haven't done too much of that um we do the new stuff we're very good with new but we haven't had the
  631. opportunity maybe it's my bias of reflecting historically about what's happened and and kind of not a Time
  632. based moment but uh you know talk to some of the people who are really involved in in pushing our our subject
  633. forward that's imminently doable right that's great yeah the road trip I'll pitch for is the is a combine so the
  634. combine week right now in Indianapolis um the base save me any time CU you know I watch the whole thing
  635. anyway we you have to bring your second and third and fourth screens for other sports that's true you got to pick up on
  636. that subtle body language of being in person that we've discussed already though you don't get that across the TV
  637. there's a baseball combine now that could be interesting yeah um but I do think it'd be you know we were we saw
  638. this technology evolve over the last 10 years and we were having conversations with people when it first came out I'm
  639. curious can we remain on that Cutting Edge and what does it take for us to remain on that Cutting Edge so you just
  640. mentioned Ai and you talked about you know dumping a bunch of video data but it may be even earlier that what AI does
  641. it levels the coding playing field Y and other people can get into analytics without having to go through years
  642. basically of coding training could be that could be one of the big things are we going to be doing the interviews and
  643. the research and the reading to really know what happens next we we we did it the last 10 years we need to keep on
  644. whatever that was that kept us at the front edge we need to keep on doing or all the companies like Pro footb Focus
  645. that do a lot of human coding well once we have a lot of human coded data we have supervised learning we can do so
  646. now we we're going to use human coders but we're going to use them to train our models and then we're just going to let
  647. our models rip on all the dat we can collect in some large automated way as a matter of fact over the years we've had
  648. a lot of people talk about the different you know value of having cameras and locations and other stuff in stadiums I
  649. think the volume of especially video data we're going to have to figure out a way to deal with that in a large
  650. automated way all right guys thanks to our audience and thanks uh big thanks shout out before we give uh before we
  651. leave to the crew here helping us bring this to life our boss man matd dats overseeing even the live show Michelle
  652. young standing in as co-director and Deion Simkins always vital to the show on behalf of my colleagues and Friends
  653. longtime collaborator Shane Jensen Eric bradow AI wer thank you guys it's been a
  654. great 10 years hope we stick around for the next 10

Brief Summary

The hosts reflect on 10 years of Wharton Moneyball, discussing changes in sports analytics and personal growth through their discussions.

Key points

  • The Evolution of Sports Analytics. Over the last decade, the perception of sports analytics has transformed dramatically, making it a cooler field for statisticians.
  • Exploding Interest in Sports Analytics. The number of applicants for sports analytics programs has surged, showcasing the growing interest in the field.
  • The Emergence of Themes. After years of discussion, the show revisits key themes like momentum and state dependence.
  • The Importance of Within-Person Variance. Exploring how individual differences can be underestimated in sports analysis.
  • Memorable Moments from the Show. Reflecting on favorite moments from ten years of broadcasting, including iconic plays.
  • The Science of Hot Dog Eating. A defense of competitive eating as a legitimate sport, highlighting Joey Chestnut's dominance.

Episode highlights

Tension

40/ 100Low tension

Related episodes