How to Measure Coaches

In college football, the "winningest" (coach with most wins) is a stupid statistic. It's much more a measure of quantity, not quality. If you coach a long time, even if you suck, you have a better shot at being winningest than a younger high-quality coach. It's a participation trophy.

What matters is your percentage of wins, right? Yes - but it's also true that teams differ in the strength of their recruiting. What I didn't know until I read this article at watchstadium.com is that the strength of each incoming class is measured and ranked, just like teams are. Not surprisingly, there's a relationship: the better your incoming class, the more your team wins.

But it's not an absolute relationship (otherwise, why play the season out?) So if we compare class quality against the final record, this gives us a good way to measure the strength of coaching. For coaches, your input is the quality of your players, and your output is your win-loss record. Coaching is converting talent into wins. That's what you have control over. Not the strength of the rest of your conference, not how screwed up the ranking polls and CFP committee are. (A committee to determine who gets into the playoff? Ever wonder why they don't just use a transparent formula? It's so the bowls make more money. It only has to do with how good the teams are to the extent that if it's too obvious they're actually maximizing for revenue, people will spend less on bowls.)

So if a coach consistently wins less often than the strength of his recruiting class predicts, he's screwing up. If he ends up winning more often than the recruiting class predicts, he's taking silver and and turning it into gold, and he's a good coach.

The article I'm linking to shows scatter plots of programs from the Power Five, plotting number of wins from 2015 through 2018 against average incoming class quality rank for the same period. They don't give an r-value, but the curve does look sigmoidal. This is interesting, because sigmoidal curves often suggest network behavior, and the output here (number of wins) for each team is dependent on other teams. They also point out the relationship is weaker in basketball for a number of reasons including transfers.

On the plot below, the further above the line, the better the coaching. Bottom line: Mike Leach at Washington State is the best coach in the Power 5, with his team finishing 6 wins higher during 2015-18 than would be expected based on his recruiting class. Paul Chryst at Wisconsin, Kirk Ferentz at Iowa, and Dave Doeren at NC State, all tie for second best coach in the Power 5, finishing on average 5 wins higher than their recruiting classes. These are the coaches who are converting talent into wins most effectively.

Which coaches waste the talent they bring in? UNC and Maryland (6 wins lower), along with Nebraska and Texas (5 wins lower.) The full list is available at the article.

For Pennsylvanians: Paterno's Record vs Engle and Franklin

November 2011 felt very much like Pennsylvania's own 9/11, especially to someone for whom Penn State was the closest thing to a religion that they were raised with. Nowadays my relationship to PSU football is like a lapsed Catholic. Sure I don't really follow the season or know the players, but I go once a year and went to the Rose Bowl two years ago - like a Christmas-and-Easter Christian I guess. And just like a Christmas-and-Easter Christian, don't you dare tell me I'm not really a fan! I just follow Penn State in my own way. In my heart!

You can't bring up Paterno without addressing the Sandusky scandal, or make people wonder why you're not addressing the Sandusky scandal. That's completely appropriate. This is a game - a game, like dodgeball - and kids' lives are infinitely more important than guys running around with a ball. Fair enough. Joe knew, and he could've been a hero by immediately going to the police, but he thought that an entertainment group (which is what all sports teams are, period, end of discussion) had an internal moral system that was more important than criminal justice. If there were another school sports team, and one of the assistant coaches was molesting kids, and the head coach not only knew about it but actively discouraged efforts to take it to the police - would you want your kid playing on that coach's team?

Without further ado: as I implied, Paterno is deified as a coach. Is he really that much greater than anyone else? Let's look at the numbers (and by the way, I expect some people to get irrationally upset that I'm daring to do something so brazen as to actually evaluate a coach's performance by numbers - much like religious people don't want the history of their holy text scrutinized too much. "It's just true! Now stop it and don't worry so much about how it was written!") To be honest, part of my motivation here is that I like Franklin - for one thing, instead of Joe's arrogant "I won't run up the score" position, Franklin realizes that his job is to increase Penn State's ranking, and that means run up the score when possible, given the (stupid) college ranking system. I always thought Paterno's take was selfish more than principled, and it certainly cost Penn State ranking spots and revenues (I may write a post about this later), which tends to be what pro-athletics people use to justify the distracting influence of football programs on the colleges that host them. Also, Franklin once coached at good old Kutztown University in old Berks County!

	Engle	Paterno	Franklin
Win %	66.7	74.4	67.9
% Winning Seasons	93.8	90.3	100
Longest win streak	15	21	4
% bowl seasons[2]	25	82	100
% bowl wins	75	64.9	50
% AP ranked seasons	31.3	51.6	50
final AP, when ranked[3]	14.8	8.7	7.5

[1] Note, "longest win streak" means longest streak of winning seasons. A season has to have more wins than losses to be a winning season.

[2] Since # of bowls has increased without a concomitant increase in # of teams, if the team remains the same, it would get higher as time goes on. Case in point, in 2014 PSU won the Pinstripe Bowl but was still un-ranked!

[3] I don't know if there are historical rankings beyond the top 25, which would give a more accurate picture (and I could just use that to give a picture of team quality instead of this and the previous category.) Though I don't calculate it here because it would take a lot longer to round up the data, a better measure might be start-to-end-of-season improvement in rank, beating expectations. This is unlikely for legacy teams with big media markets (think Notre Dame and Penn State) for cynical business reasons. Legacy teams tend to be consistently over-ranked in the preseason and under-perform during the season, ending up lower; for new hotshots in small markets it's the obvious (think Boise State or TCU.) This is the way the NCAA underhandedly seeds the bowls in order to maximize revenues. Since we're only dealing with one team this may not have any effect here, or rather, it only will to the extent that the rating powers-that-be are changing their opinion over time about how much they want PSU in a bowl, and/or the extent to which they're using this technique.


So what does this show? Paterno does have the highest win rate, but a lower % of winning seasons. That suggests that under Paterno, when Penn State was good, they were great, but when they sucked, they really sucked. (Specifically, they sucked in the 21st century, but more on this in a bit.) But the statistical fallacy is that there is more noise in Paterno's history because his tenure was so much longer than Franklin's so far. Franklin gets to more bowls (see footnote #1) but does worse in them.

Taking out the win streak because it's so heavily dependent on tenure, in these categories, Engle wins for % of bowl games he won, Paterno wins on win % and % of seasons ending up in the AP rankings, and Franklin wins on % of winning seasons, % of bowl seasons, and final AP when ranked. That is to say, Paterno was in the ranked teams list a little more often than Franklin at the end of the seasons, but ranked lower.

What does it look like when we compare the first four seasons of all three coaches - that is, when Engle and Paterno were in Franklin's position near the start of their careers, how did they look?

	Engle-4	Paterno-4	Franklin
Win %	62.2	81.4	67.9
% Winning Seasons	75	75	100
Longest win streak	4	3	4
% bowl seasons	0	75.0	100
% bowl wins	N/A	62.5	50
% AP ranked seasons	0	75.0	50.0
final AP, when ranked	N/A	4.7	7.5

Paterno got off to a strong start, again with high win percentage - that doesn't go away. Franklin spreads the wins out more and is more consistent season to season. It's worth noting that Paterno wasn't taking over a still-shocked program that had trouble recruiting.


What about the 2000's?

Paterno had his first losing season in 1988 after a 21-year streak of winning seasons. After that, others followed: 2000, 2001, 2003, 2004. You can see the trend over time here. (Note, his first non-winning season was a tie, not a losing season, and was also his very first as the head coach in 1966.) It may be worth pointing out that when the slide started in around 1998, Paterno was already 71.



It might seem a little mean to pile on a coach after he's gone, when there was no shortage of people in the aughts saying that Joe should retire (and it looks like they were right, for more reasons than they knew.) But Penn State continued its rabid fanbase, and does so even now after the scandal. (Here's the reason why people stick with unpopular or losing teams.)

Future possible project: analyze rankings movements during Paterno's career based on margins, and if there is some estimate of how many points off the margin you lose by putting in (second string, third string, etc.), and you correlate end-of-season rankings with revenues, you can actually calculate how much money you cost a university by refusing to run up the score.

Is There a Correlation Between NCAA Team Ranking and NFL Players Produced?

Some college teams disproportionately produce NFL players. It's not quite a Pareto distribution, but it's getting there: of about 3300 2016 starters listed by the NFL, they came from 355 college teams - and 20% of the teams produced 65% of the players.

You would expect that the better-performing teams would produce more players, but is this the case? And if there is a relationship, do some college teams produce even more NFL players than you would expect, given their average end-of-season rankings? (And vice versa.)

Methodology: source comes from sportingcharts.com and Coaches Poll end of season ranking. NFL Players Association says the average NFL career is about 3.5 years, so for the 2016 season, most of the players would likely be those who were seniors in 2012-2015. Therefore the "ranking" here is the averaged end-of-season ranking for those years. The perennial problem of these analyses is how to count unranked years and most other analyses assign "26" for ranking when the team is unranked, which is what I do here.


Yes, there's a trend, and it's as you'd expect. For every one-step improvement in the average ranking over that four year period, you send on average 1.7 more players to the NFL. You can see that there are outliers:

Michigan, Oregon, and Navy under-produce NFL players, relative to their rankings.

Florida, Florida State, LSU and and Georgia over-produce players, given their rankings.

So if you go to an SEC school, it doesn't even matter if you win!

To see if my methodology is sound, I should do multiple four-year sets of ranking averages (2011-2014, 2010-2013, etc.) compared to the 2015 NFL roster and 2014 roster respectively, etc. and then see if the fit is WORSE when mismatched (i.e., 2011-2014 should determine the 2015 roster better than 2014 and 2016) but I'm not getting paid for this am I.

I could speculate about the reasons for this - whether under-producing means good coaching (or bad), same for recruiting, or non-NFL (or non-football) careers pursued by athletes at the "under-producing" schools, etc. But the under- and over-production by conference is interesting in itself.

Why Do People Remain Loyal to a Losing Team?

Cross-posted to the Late Enlightenment and Cognition and Evolution.

tl;dr Sports fan behavior is explained by a combination of constant identity-forming team loyalty which is an end in itself, and status signaling by association which is modulated by team performance. These two factors differ between individuals and are associated with different cognitive styles, with constant loyalty more associated with moral foundations and intransitive preferences.

It's been observed that you can tell who a team's true fans are by noticing who remains loyal to the team even when that team is losing. I think this is meaningful, but it does beg the question: what are those fans getting out of it?[1] Of course any speculation about this must mention the very real example of the Cleveland Browns, who over the past 2 years have a 1-31 record, and this year after going 0-16 they were on the receiving end of a sarcastic "perfect season" parade.

Humans get utility from associating with others with high status. Much of the happiness that a sports fan gets from their emotional connection to their team derives from this, and many observations are consistent with what a status-by-association theory would predict: fans are happier when their teams win because they feel high status and can signal higher status, they engage in extreme dominance displays when their teams win important contests (i.e., people acting like idiots as they come out of a championship game if their team won, yelling, jumping on cars, setting off fireworks) but not if they didn't win, they attend games more when the team is winning and less when the team is losing, and they wear branded gear to identify themselves with the team and otherwise let others know of their association.[2]

But this theory falls short of explaining why, for example, there is any such thing as a team's consistent fanbase. By this model, everyone should just cheer for the best team, game by game (or even play by play!) It especially doesn't explain why the the Cleveland Browns have any fans left at all; supposedly they're a football team but I've seen a number of convincing arguments against that, for instance, every game of the 2017 season. During an 0-16 season you would expect that if fandom is about fully rational people maximizing utility by associating with high status teams, the fans would stop posting on forums, they would put their gear away and deny to others that they were fans, and the stadium would not just have lower attendance, it would be completely empty. Yet this is not what happened.

I think the answer here very likely has to do with the gap we see between two types of beliefs/behaviors that often produce apparent impasses in other domains of life, especially religion and politics, the intensity of which differs between individuals. This gap in rational and more instinctual behavior will seem very familiar to readers of books like Jonathan Haidt's Righteous Mind, or Simler and Hanson's Elephant in the Brain. Humans demonstrate some domains in their cognition which are inflexible and impervious to reason - to use Haidt's categories, harm, fairness, loyalty, authority, and purity. By "inflexible" I mean "not open to discussion, or conversion into money or other goods/services." For example, you likely do not believe that murdering children is morally acceptable. Are you interested in hearing arguments about why it might be morally acceptable? If you would never consider such a thing, and you're uncomfortable that I would even suggest it in a thought experiment, you're showing inflexibility in discussing it. Okay - would you kill an adult for $50,000? I see that also upset you, I'm sorry to have opened with such a low offer! $75,000 then? You're being inflexible (I hope!) in reacting by thinking "It's not about the number!" Okay, what's the conversion rate between adults and children? Forget murder, how about urinating on a picture of your family for money? etc., you get the point. "Inflexible" means it can't even be suggested as open for discussion, which includes not being allowed to convert between moral-foundation-violating acts and money, or between different types immoral acts. (A favorite of action movies or dramas to demonstrate the extreme evil of an antagonist is to have them force someone declare the relative value of immoral acts, e.g. Sophie's Choice.) To connect to the abstract - the philosophical term for having values that cannot be negotiated, and for which there is no relative value like this, is that they are intransitive.

I took you on this little tour of moral darkness to illustrate that morally normal humans do not adhere to consistent rationality, and the ones that actually do are psychopaths.[3] (You may be interested to know that Haidt found that when he surveyed the business students he was teaching, they scored low on every single moral dimension, taught as they are that everything is negotiable.) So what does all this have to do with the Cleveland Browns? Many of us have noticed that "hardcore" sports fans - the ones who stick around with long faces even when the Browns are losing, and falsify the first model above - tend to have certain personality and cultural characteristics that fit well with some of these inflexible moral foundations: they tend to be more religious, more nationalistic, more conservative and more valuing of loyalty and authority.[4] Sports fans rarely become hardcore about a team after entering adulthood, and very often there is a family lineage of fandom - and these are exactly the times and ways in which characteristics of core identity are formed. Also telling, while there were about 3,000 people who showed up for the Cleveland Browns parade, there were many fans who were quite angry about it - but online objections were mostly that it was "embarrassing". (No mention of the 0-16 record that inspired the parade.)

Before I put into words what might be motivating them and make predictions, here's a summary of the two kinds of of beliefs, producing two kinds of motivation. While these beliefs exist in everyone, there is going to be a distribution in the population, with one category of beliefs dominating the fandom-related cognition of some fans, and the other category dominating that of others.

HARDCORE FAN	CASUAL FAN
motivated by moral foundations	by utility calculations
end in themselves	deliberate, external goal-oriented
higher value on loyalty	lower value on loyalty
adopted in childhood, maybe from family	adopted voluntarily in adulthood
not negotiable	negotiable
central to identity	not central to identity
unwilling or unable to verbalize	position clearly verbalized
more often encountered in person	more often encountered online
sees casual fans as untrustworthy, sleazy	sees hardcore fans as stupid, gullible

Of course it's a spectrum, and every fan is somewhere on this spectrum, but many of us clearly lean toward one or the other end. (If you're reading this, you're more likely in the right column than the left.)

To summarize the hardcore fan: he is motivated by more basic, instinctual moral drives, especially loyalty. Being a good fan is an end in itself, and an offer to burn a team jersey, to cheer for the other team, etc. in exchange for money is likely to not only be immediately refused but to provoke active offense. These fans consider their fandom a crucial part of their identity, to the extent of including team-related themes in their weddings or mentioning it in obituaries ("he lives and dies by the Browns"; "a Browns fan to the core.") He can get uncomfortable when the business aspects of a professional sport are discussed and overshadow the games on the field. Asking him to explain his fandom will be met with puzzlement, anger, or a jumbled set of team cheers and slogans, in the same manner as a person asked to explain why they are patriotic or follow a certain religion - "If I have to explain it to you, you'll never understand." And finally, because tribal loyalty sentiments are more warning-barks or team cheers than any kind of actionable proposition, you're more likely to hear such sentiments when talking to him in person, where the nonverbal (affect-laden and irrational) part of communication dominates. He will be a fan for life. When the bandwagon people disappear during losing seasons the hardcore fan says "Good riddance, good-time Charlie."

To summarize the casual fan: he is motivated by utility calculations about external goals (this team might win this year so I'll cheer for them, maybe I can make friends this way, maybe I'll look successful if I follow a good team.) He doesn't see what's impressive about staying loyal to losers, and really doesn't understand why making fun of your team when they lose is shameful or embarrassing. He probably picked up his fandom after college, maybe when he moved to a new city. He probably don't care either way about the business dealings of the team. If someone offered him money to stay home from a game or burn team logos, he would seriously consider the offer. He doesn't introduce himself to strangers as a fan, and five years from now he might not be following the team, or might not be following the sport at all. He can give clear reasons why he started following the team, and you're more likely to hear from people like him online. He shakes his head at the hardcores who keep shelling out cash for losing teams' jerseys.

Both the hardcores and non-hardcores gain utility in proportion to the team's performance. A team's performance can be negative, causing you to lose utility by associating with them.[5] But there must be another source of utility for the hardcores, who somehow gain utility from the association no matter the team's performance - and that source of utility is a constant ability to demonstrate loyalty, period, to others as well as to themselves to reinforce their own identity. And this signal is most informative when your side is losing.[6] Speaking quantitatively, in the utility equation for this model, there are two terms, loyalty (a constant for everyone, hardcore or not), plus the product of team performance times associative utility. Associative utility is how much your utility changes per team winningness. Both loyalty and associative utility vary by individuals, and team performance of course is determined by the team. The equation looks like this:

Total utility = Loyalty-based utility + (Team performance * associative utility)

Team performance can be positive or negative. For the hardcores, loyalty is such a large term that it doesn't matter how negative team performance is, loyalty will alway be greater and the total utility will always be positive (this could be the definition of "hardcore", "rain or shine", etc.) Further toward the other end of the spectrum, the value of loyalty signaling decreases and the team performance makes more of a difference in whether people keep following the team. It's also worth pointing out that this explains people who don't care about sports at all, because they have zero loyalty and zero associative utility - that is, it doesn't matter how much the team wins, they still won't care.


PREDICTIONS

Many of these predictions seem trivial, but the point is to relate these predictions to specific components of the hardcore fan's motivation structure as noted in the table above, which would be more informative.

• While utility is hard to measure directly, there are good proxies for it, like revenues, attendance, or Nielsen ratings. Given that there will be a distribution of hardcore to non-hardcore fans, there will be a non-zero floor to revenues so even 0-16 teams don't go to zero, as we observed. If we graph all of the teams on performance vs utility proxy, I would expect a mostly linear-looking scatter plot with an increase in the slope at the good end, for those teams with some expectation of a national championship, and possibly a flattening at the bottom. This may depend more on expected utility (if fans are pleasantly surprised by a win vs. they expect their team always to win.) I plan to try to collect some kind of utility-proxy data and see if this is in fact the case.
• In general a sport will be more successful in inspiring loyalty, the more similar it is to tribal warfare (always a reliable revenue stream for every team); maybe this is why football has eclipsed baseball as the national pastime.
• The more hardcore, the more they will pay attention to the outside charity activities of their own team, and the more outraged they will be by disloyalty-demonstrating acts, e.g. kneeling during the national anthem. They will also be more interested in the moral failings of opposing teams, especially rivals.
• The more hardcore, the less they will be interested in statistics, especially of other teams, even ones their teams are playing in important games.
• The more hardcore, the greater the difference in their interest in a player when he is on their team, vs. after he is traded. That is, hardcores think each of their players is a great person on and off the field - when he plays for their team - and any suggestion that they'll stop caring about him the second he is traded is likely to be met with hostility, but in fact this is the behavior they demonstrate. (He will also be annoyed when asked why, or when Seinfeld is cited - "Essentially you're cheering for clothing.")
• The more hardcore, the more they will feel sad or angry after a loss, and the more likely they are to attend or watch the next game despite having been very sad or angry at the last game's outcome.
• The more hardcore, the less tolerant they will be of fans behaving negatively toward the team, even when the team loses (very concrete and contra expectations here: you might expect hardcore fans to support a parade showing anger against the people making their Browns lose, but it seems to be exactly the opposite. Parallels to gay marriage here too: how exactly does the 0-16 parade degrade your fandom, when you didn't attend?)
• The more hardcore, the more they will confuse the team with a government agency or public good (i.e., demanding that the city finance a new stadium.)[7] More recent teams with cities that have highly educated and/or mobile populations (i.e. the coastal Pacific) will therefore find that they can't get what they want from those cities, because the voters don't care (Seattle, San Francisco, San Diego) where other cities filled with less mobile, less educated people would crucify their mayor for allowing a team to leave on their watch.
• It's often been noted that the Midwest with its brutal early winters has far more rabid sports fans than the mild West Coast. One possibility is that the loyalty-demonstration value of attending every game is diminished when all of those games are 70 F and sunny, vs some of them being freezing cold. (Think of the people who still wait in line in the cold and dark on Black Friday morning to buy things for their families. They do know that Amazon exists. So what do you think they're really doing?) Of course there could be a climate-independent cultural difference between east and west coast, but the model's prediction would be that Miami has equally low loyalty.
• The more hardcore, the more they will be upset if a star player leaves for another franchise, or the whole team moves to another city, and they use words like "betrayal."[7]
• The more hardcore, the less tolerant they will be of long-term, off-field strategies, especially ones that alter play and result in on-field losses. (Both the 2008 Detroit Lions and 2017 Cleveland Browns had 4-0 preseasons, then went 0-16. Tanking (here and here) and/or salary cap manipulation? Difficult to explain as mere incompetence. And if it were confirmed that this is what is happening, the hardcore fans would be angry; casual fans might say "Huh, that's kind of clever, although it means you've been putting a bad product on the field." "My team is not a 'product'!" the hardcore fan says.)
• I'm not sure what to predict about the impact of hardcoreness on betting. The hardcores' loyalty may make them become overconfident in their team's performance. On the other hand, moral foundations-related beliefs are often kept carefully separate from anything affecting real-world decision-making. By that I mean: sacred beliefs are often more tribal chant than actionable proposition, and in general, people desperately avoid any bet that touches their moral foundations (next time someone makes a verifiable statement about religion or politics that you disagree with, offer to bet them, and see what happens. Typically they backtrack to a non-verifiable version of what they said, and/or get very offended that you would "cheapen" such an important matter by betting on it - which are all moves to avoid testing their belief.) Then again, the hardcore fans presumably know more about their team than most others, which means they should be more confident in their predictions, and be more willing to bet. Consequently they may be less willing to bet proportional to their claimed confidence, than would a casual fan with equal knowledge of the team would be. In my one test of this during March Madness, I found that self-identified fans did more accurately predict the outcome of a game involving their team than non-fans, but I collected no information on willingness to bet.

Footnotes

[1] This very article is diagnostic. By trying to dissect loyalty, instead of taking it as an obvious good and discussing it in the context of a specific team, I mark myself as someone with a small loyalty term in my equation - whereas people whose sports utility equation is dominated by loyalty would not understand, and/or be actively be offended by, a question like "What do you get out of being a fan of your team?"

[2] One might argue that a purely rational human being would ignore sports altogether - what do a bunch of guys chasing a ball on a field somewhere else in my city have anything to do with me, I've never even met them! - and I'm sympathetic to that argument.

[3] I hope no one read the paragraph about the price of murder and thought, "Hmmm...What is my price to kill someone?" In the case of exemplar psychopath Richard Kuklinski, he got positive utility from harming people so he kept doing it even after he ran out of work.

[4] When people do not have VNM-consistent rationality (that is, they have these inflexible, non-negotiable, non-fungible beliefs - i.e., intransitive preferences) - they can be turned into money pumps, by observant and unscrupulous characters who can carve their motivation structure at the joints, i.e. focusing on the the inconsistencies. While this has been reproduced now in artificial settings, not only salespeople but politicians have been doing it since the dawn of civilization. The NFL and in particular the Cleveland Browns are doing exactly this to the fans by exploiting the intransitive preference of loyalty, and I would be very surprised if their marketing does not already have a model of their fans and spending patterns similar to what I've described here. Another follow-up is to look for literature on whether psychopathy allows one to see these disconnects more easily, or (hopefully) the ability to see them and the willingness to act on them are unrelated and therefore form a mercifully narrower sliver on a Venn diagram of the population.

[5] There's probably a Markovian/hedonic treadmill effect here too, where the utility multiplier from a team's win is not constant but rather influenced by expectations based on the team's record. Next year if the Patriots go 9-3, fans leaving a game after a win won't be as happy as Browns fans if the Browns have the same record.

[6] Remember Karl Rove dragging out the 2012 election night broadcast and refusing to accept the outcome, seeming a little nuts? But simultaneously advertising to ten million Republicans watching that he never ever gives up. Say what you will about Karl Rove, but "bad strategic thinker" was not among the many epithets hurled at him.

[7] When the Baltimore Colts were about to move to Indianapolis in 1984, the city actually tried to pass an eminent domain act (!) to take over the team, but the Colts escaped with the team's property under cover of darkness the night before. Other teams like the Chargers have found a much more lukewarm reaction on threatening to leave, and found themselves without many fans.

[8] While I wrote this post I was wearing a Garfunkel and Oates sportsball T-shirt, so you can guess which end of the spectrum I'm near.

Predicting Steelers Games

[Added at the end of the season: this ended up being very fun, but not an effective way to test whether sheer team loyalty and knowledge versus general unbiased football knowledge are better for predicting game outcomes. The reason it wasn't a good test of this was that early on, my mother realized that her approach of "I like the Steelers so I will always predict that they win" was in conflict with her moral value of "I don't like losing money to my son". She therefore started asking me to tell her what the official spread was, and more or less based her prediction on that (so much for information-free statements about team loyalty when you're pinned down by a measureable statement with an actual consequence! Not that surprising.) But, that's where I also learned I had a conflict; namely, my moral values of "I don't like losing money to my mother" and "Don't lie to your mother" were in conflict, which I promptly solved by always lying to her about the spread so that I could win. Once she realized her son was the kind of terrible person who would lie to his mother for money (clearly because of his bad upbringing) she started looking the spread up on her own (who knew, your mother will use the internet to avoid losing bets!) So after 4-5 games her bets were mostly informed by the Vegas spread just as mine were, not really reflecting the accuracy of loyalty vs expert estimates. But most importantly, I still came out 2 games ahead.]

(If you're a math-minded sports fan much of this may seem like a Neanderthal re-discovering fire - if so, please don't hesitate to comment and point me to resources or critique errors in my thinking.)

This isn't really a team sports blog but occasionally there's some fun with statistics - for example, showing how the BCS Bowl System seeds high-earning teams into its bowls through its bullsh*t preseason rankings, or seeing if there's bias in predicting March Madness games among fans vs. impartial predictors. Now I have a new and exciting project: betting with my mom on Pittsburgh Steelers games.

I don't know that much about football and I certainly don't have loyalty to any team, but I like the statistics of it. And I really like taking money and pride away from the wonderful woman who brought me into this world. But I also want to see how the spread we settle on for our bets, differs from the Vegas line, and whether it does so systematically. (The only way I can swindle her is if she doesn't have the same information I do, i.e. doesn't know the Vegas line. I got her a smartphone with my own money, and if she would use the damn thing she could look it up herself. So unless my take is more than the price of the smartphone I'm not a bad son.)

Statistically, the point spread is the handicap needed for a 50:50 outcome. I've always been interested in the effect of team loyalty on predictions - so in other words, does my mom's loyalty to the Steelers make her think they are more favored than they really are? And of course, in a zero sum game, you want the other person to be irrational in deciding where the 50:50 cut off is. So far, our spreads have been 1.3 points more in the Steelers' favor than Vegas, and I'm 2-1, and it's a dollar a game so she owes me 3.5 zloty. (I only accept Polish currency.)

What I'll be looking for as the season progresses:

• Do the Vegas spreads converge on actual scores as the season progresses, i.e. as there's more information available about the teams? So far this season, they've been way off. Taking the spread and betting on the underdogs, this week you'd have made good money.
• What is the relationship between Elo rating and spread? 538 says it's about one game point per 25 Elo points.
• Finally, what's closer to the actual score, the Vegas line or the K10 bet - in terms of both average difference, and standard deviation? Most importantly, how were the bet outcomes for each spread?

ADDENDUM

When I looked up the score for the game (why sit there and watch the game when you can just get the score at the end! Waste of time!) I ended up reading about this week's opponent, the Cleveland Browns. It's easy to develop a morbid fascination with bad, bad teams. The Browns are certainly that, and the franchise has been drowning in toilet water for a decade. Cleveland almost became a 0-16 team last year. I really can't say more about how bad the Browns are, and how much any affiliation with them will cause you to suffer (much like a sports fan's version of the puzzle box from Hellraiser) than this Deadspin article. But my favorite part of their last season is as follows. Trying to salvage some humor from the Browns' suckitude and passive-aggressively punish them, their fans had gone so far as to start raising money for an oh-and-16 parade. And guess what? After fourteen long, stupid, torturous games, on Christmas Eve 2016 the Browns narrowly defeated the San Diego Chargers. THAT is when they finally decided to get their act together, when it would ruin their fans' plans, and after it could not possibly matter anymore in terms of the playoffs, and (my favorite part) against a team that technically would no longer exist after the end of the season (the Chargers moved to LA after that season.) So not only was the win completely pointless, it robbed the fans of their one moment of levity. It was worse than worthless. I really like that.

Sub-2-Hour Marathon by 2019?

Here I predicted (with data) a sub-2-hour marathon by 2038, based on the rate at which previous world records have fallen. (Which by the way is amazingly logarithmic.) [Note - I was obviously wrong. Kudos to Eliud Kipchoge of Kenya for beating the record in 2019! And to his coach Patrick Sang of Kenya. A great day for humans!] A very ambitious Greek running coach and scientist (Yannis Pitsiladis) is trying to produce the first sub-2 by 2019. Here is the article in the New York Times. Their figure showing record marathons (and fastest marathons in each year) has more data and is prettier than mine, so I reproduce it here: Not super easy to see if I fit the whole thing into this blog's format, but: the Y-axis is marathon time, with the lower edge being 2 hours. X-axis is year, in 5 year increments. The colored dots are the world records, the text and arrow at the bottom middle of the graph is Pitsiladis's goal, and the shaded area is the trajectory records are likely to take going forward. The earlier edge hits the 2-hour level around 2033.

If the record is beaten much earlier than this, I suspect it would be some combination of the following two things: undetected doping, or genetics. Genetics further breaks down to good luck (a new mutation; see earlier article about how this may actually confound endurance sports); better recruitment (many credit Germany's dominance of soccer to this, but running as a sport just doesn't command the profits to accomplish the same thing); eugenics; and discovery of pre-existing genes in previously isolated populations. Maybe there's a group in Tibet or highland New Guinea just waiting to whoop the marathon world's ass! What I don't think will cause an early sub-2 is non-doping training innovations. We mostly seem to be chipping at margins of mature training techniques. A little taking advantage of altitude here, a little better nutrition there...will that really get us there so soon?

As an aside, the importance of human capital in the modern world is demonstrated by this caption from a photo in the article: "The biomedical lab at Addis Ababa University contains hundreds of thousands of dollars of equipment. But because of insufficient funding and a lack of available experts to operate it, much of it is unplugged and covered by tablecloths."

(Watch Kipchoge cross the finish line here.)

Where Should You Live? Critter Fatalities By State From 1900

Added later: a NYT article shows that we do indeed spend too much time in the U.S. worrying about things with venomous things and things with big teeth, as opposed to dogs and wasps. And even considering wiki statistics, based on Wiki statistics from the 2010s for snake bite deaths and population, a person in Australia is indeed more than 11 times as likely to be killed by a snake in Australia than the U.S. - but even in Australia you still didn't even have a 1 in 13 million chance of being a victim from 2011 until today. Your morning commute is far more dangerous than living in Australia.

Also added later: here's a guy screwing up my statistics who was also bitten by a shark in Hawaii just a month after being attacked by a bear in Colorado. If I were this guy I would have business cards made up that said this.


In light of the recent seasnake that washed up on a beach in Oxnard (pleasantly, a few days after we were swimming across the channel from there), I looked up the statistics for snakebite fatalities in the U.S. (No recorded seasnake fatalities.) Interestingly, at the bottom of the wiki article for snakebite fatalities, they have links to mountain lion, bear, alligator, and shark fatalities because they know there are people like me running around loose who will look those up next anyway. (Hereinafter these five types of animals are referred to collectively as "critters".)

And of these critters, how many deaths do you think they've caused in the U.S. since the turn of the last century? 226, over the last 115 years. Not even two a year, not many! Compare that to the about 31 deaths per year caused by dogs (actually, 42 in 2014). And of course (being responsible) if you're out in the wild you are much much much more likely to be killed or injured by falls, drowning, exposure to heat or cold, dehydration, or other people than by wild animals. Of course drowning doesn't have pointy poisonous teeth that rush at us from the darkness, so our dumb Type-1-error making brains pay less attention to rushing water rising around our ankles than a twig snapping in a forest.

Still, I decided to compile absolute critter fatalities by state. In calculating this, I took out BS ones (i.e. snake handlers and dummies who keep non-native poisonous snakes at home don't count as a legitimate bite. Come on guys.) Interestingly, Texas has fatalities from the most types of critters (four of the five) - they finally got a gator fatality on the board this summer - and in fact Texas is the only state where the range for all five critter types overlaps because there are sharks and gators on the Gulf Coast, snakes all over, mountain lions in the west, and bears in the Big Bend part of the state. (Come on guys, you can do it!) There are 22 states with no fatalities from these critters since 1900. If all these critters scare you, then the biggest contiguous area with no critter fatalities stretched from Minnesota, to the west of Missouri and down to Louisiana. (To be honest I have difficulty believing no one died from a snake bite in the southern part of this range since 1900, and if this omission from my data here offends you I'd like to invite you to go dig for the stats yourself.) Northern Iowa/Southern Minnesota is the only place safe from all five, which no doubt is why the Mayo Clinic was placed there, as there's no other reason to want to be in Rochester Minnesota.

Ranked in order, the states with the most critter fatalities in absolute terms are:



However: the absolute number of critter-victims in each state is lacking as a risk indicator. Why? Look at California and Alaska, which are tied. There are a lot of people in California, and not so many in Alaska, and yet critters have managed to get the same number of people in each state. Concretely speaking, there are 38 million people in California, and not even a million in Alaska, so walking around in Alaska it's actually *76* times more likely that critters will get you! So if we weight for population (i.e. critter fatalities per person in the state) what does it look like? Now, Hawaii, which somehow manages to keep its shark attacks very quiet (I wonder why?) comes out on top, so I adjusted the rates to express the others in terms of how many Hawaiis-worth of critter risk they represent:



Interestingly, Hawaii has only ONE type of the five critters, but apparently in Hawaii that one type of critters eats well. In this analysis, California drops from #3 to #13. An also-ran! Even New Jersey is per capita more dangerous than California! (Granted, this is owing to the black-swan/white-shark event of the Matawan Maneater.)

People in Hawaii are in the water a lot, which reminds me that I once calculated statistics showing that people in Florida were 3 times more likely to get attacked by a shark, but people in California were 3 times more likely to be killed by one. (Which is to say, once you're attacked, you're 9 times more likely to get killed in California than in Florida. Because we have great whites and they have coot widdle tigers and bulls.) But again this statistic still understates the danger, because Florida has a lot more accessible coastline (California is mostly cliffs) and it's warm, so lots more people go in the water, therefore a LOT more exposure to shark attacks in FL. And still more fatalities in CA? That is to say, once you're in the water in CA, you're much more likely to be attacked and killed - but I don't have ocean bather-numbers to back that up.

And a good day to you.




Above: the head of the Hawaiian Tourism Development Board. He advises that all tourists bring Worcestershire sauce and perhaps tuck a sprig of parsley behind the ear before swimming. From animaltime.com.

Expected Utility, Expected Points

Cross-posted to Cognition and Evolution MDK10 Outside.

New analysis of football stats at advancednflstats.com, from an expected utility standpoint, rather than just "yards gained". One yard from the 50 to the 49 yard line is not the same as one yard from 1 to the end zone. The article points out that even points expected is limited and must be considered when time is not a factor, since a team up by a single point will sacrifice the opportunity for more points just to run time off the clock, and avoid the small chance of turnovers. Of course the departures from rational optimization are the most interesting, but they don't get into that (yet). A lot of those apparent deviations from rationality may in fact be maximizing something besides win-loss records. Professional football is a commercial endeavor and isn't played in a vacuum, and contra Lombardi, winning isn't the only thing. Profit is.

Below: from the linked post. Note the (unsurprisingly) slightly sigmoidal (importantly: non-linear) shape of the curve, which is why expected points is not the same as yards gained.

Geographic Analysis of Superbowl Winners: No Temperature, Distance or Time Zone Effects

In sports, even in championship games that are supposedly on neutral ground, some teams are thought to have home advantage if they play in or near their home state, for whatever reason; cultural similarity, less wear and tear from traveling, etc.. Do we see this in Superbowls?[1] (I didn't find any effects here so you can scroll up for a little more interesting one on ratings if you want.)

Not really. The winners, on average, have actually had to go a little farther from home than the losers (winners are on average 1487 miles from home, losers 1409, using Google maps driving distance to approximate.) And the home-state advantage, so far as it's happened - only four times so far, all with California teams playing in Cali - is a wash (2 wins, 2 losses). I also checked to see if the curve is U shaped, inverted or otherwise. For instance, if it's inverted U, maybe it's rough to take a 400 mile bus ride since further than that, you fly, and a full transcontinental flight is dehydrating and stiffening. But no dice. (If I cared, I would look at overall NFL records, not just the Superbowl, unless we're assuming the Superbowl is different.)


Above: distribution of Superbowl Games I through XLVII



Above: Superbowl wins and losses (#wins and losses proportional to radius of circle)

Seeing the maps above, you can tell that Superbowls and Superbowl teams are not distributed randomly. There's an obvious bias to the Northeast, as would be expected based on population density (and bigger markets). These maps show wins by CITY, not franchise (unless the franchises have remained stable and separate - so Jets and Giants are separate, but L.A. Raiders and Rams are not). But all things being equal, this should come out in the wash anyway.

So what about time zone changes? There's a known effect in sports called circadian advantage; it's assumed in medicine that we can adjust our circadian rates up or down 5%, so that we can adjust to one hour of time difference for each day. In fact in a peer-reviewed 1997 paper on Monday night football game advantage in the NFL, it was shown that West Coast teams do significantly better over East Coast teams than Vegas odds predict. (Note this is for night games, as would be expected.) It turns out that there is a slight but not significant difference overall: counting the Eastern time zone as 1, Central as 2, Mountain as 3 and Pacific as 4, Superbowl winners are from a little bit further east, with an average time zone of 1.8 vs the losers' average time zone of 1.9. There was no association between score margin and time zone either. I also checked to see whether there was any advantage to being closer to the time zone of the Superbowl location, or being from an earlier or later time zone, but there was not (the distribution was just about exactly coin-toss random).

While we're worrying about environmental factors, we might as well look at temperature - the home field's December average temperature, against the temperature on the field at the Superbowl. There is no clear curve for any relationship I checked (team's home December average temperature vs points scored, or difference from Superbowl city average temperature vs points scored.) There is a slight difference in terms of average "distance" (could be warmer or colder) between a team's home city, and the Superbowl city: winning teams are 20.9 degrees Fahrenheit off from the Superbowl temperature, and losers are 23.4 degrees off. In non-absolute value terms, the winners are on average 19.2 degrees colder at home vs the Superbowl, and the losers are 21.2 degrees colder at home.

What does this tell you about the Seahawks-Broncos matchup? Nothing! Who cares!

Footnotes
[1] Yeah, I said Superbowl. Not "big game". S-U-P-E-R-B-O-W-L, as in NFL. Listening to commercials you'd think saying Superbowl summons He Who Cannot Be Named or something.

[2] Smith RS, Guilleminault C, Efron B. Circadian rhythms and enhanced athletic performance in the National Football League. Sleep. 1997 May;20(5):362-5.

Advanced Statistics in Football

A New Republic piece about how statistics isn't just for baseball anymore.  One statement that I did find interesting:  “'Most of the analytical brain power in football is on the financial side,' according to Burke, but teams are 'waking up' to the on-field possibilities."  To the extent that franchises believe their profits are related to their performance and win-loss record, they will care about their performance and win-loss record.  A future project of mine will be to correlate scores and win-loss records with team profits and changes in profits over time.  I submit that the teams with independently measured higher fan loyalties will show less of a difference; sure, it's kind of just common sense, but it relates to the business decisions made by the owners. 


Meanwhile, March Madness is coming up, and I'm planning a more expansive crowdsourcing experiment than the interesting one last year.  Let's see how the wisdom of crowds stacks up against the pundits that pundittracker.com follows!

Are We Done Breaking Records?

World records for athletic events over time follow logarithmic curves (see here for the marathon record by decade, and an extrapolation that we'll see our first sub-2 hour marathon in 2038.)  Conversely, Peter Keating at ESPN points out the recent lack of broken records, and argues that just maybe, our top athletes are as good as humans can get.

Sub-Two-Hour Marathon in 2038

Inspired by Patrick Makau Musyoki's new marathon record in Berlin yesterday, I looked for trends in the marathon world records for each decade going back a century. I only included the fastest time in each decade. I expected a plateau like this, but I didn't expect it to be so neatly logarithmic. Explanation of the graph: the Y-axis times are in seconds, so 10,800 is a 3 hour race and 7,200 is a 2 hour race. The horizontal red line is a two-hour marathon, the vertical dark line is 2038, and the light gray prediction line is the trend for future marathons. (Note that the prediction line on the graph is for illustration only - the numbers come from the best-fit equation printed on the graph.)

Look at that R. That's a damn good fit. Marathon records are a very logarithmic phenomenon.

A whole crop of articles commented over the last year on statistically improbable sprinter Usain Bolt, who is ahead-of-trend by thirty years. In the same vein, looking at the marathon plot, we shouldn't expect a male human to break two hours in the marathon until 2038. And it's reasonably assumed that the incremental improvements we see in these times is a result of (decreasing marginal) improvements in training, nutrition, and running equipment. Therefore, if this record is broken significantly ahead of that (you can decide how to calculate standard deviations if you want to define exactly what "significantly" means) then I predict one or more of the following will have happened:

- A genuine new mutation and a recruitment sifting system that can deliver the talent to the field, as is presumably the case with Bolt in short distance (someone sequence this guy already);

- New technology allows undetectable doping, either to increase blood oxygen carrying capacity, muscular oxygen efficiency, or muscle strength;

- Deliberate breeding has occurred by nation states with long-term views and an overbearing need for international prestige (China, we're looking at you, and here's why);

- Most interesting possibility: a population of humans still mostly reproductively isolated and thusfar not competing in marathons, and that has gene variants which benefit them in this event. (Tarahumara, we're looking at you now).



Intrade doesn't have a market for this prediction. But they should.