hero_pattern_miners.jpg

Blog

Musings on Logic, Analysis, Decision-Making, and Other Elements of Natural and Artificial Intelligence

The Super Bowl, the Iowa Caucuses… And Phil the Groundhog

SuperIowaGroundhog.jpg

This Sunday will be an event for which the country has been eagerly anticipating for an entire year.  And no, I’m not referring to the Super Bowl.  Nor am I referring to Monday's Iowa Caucuses.

I’m of course referring to Groundhog Day!  When Phil the Groundhog pops his head out of his hole in Punxsutawney, Pennsylvania on Sunday, he will hold, in his adorable little groundhog paws, the fate of all Americans - at least those who live in regions which experience the season known as winter.

GROUNDHOG DAY AND THE SUPER BOWL

One region which does not experience winter — at least not the snowy kind — is Miami, Florida.  That’s a good thing, since Miami’s hosting the Super Bowl on Sunday, mere hours after Phil makes his weather prediction.  Even if Phil sees his shadow, Miami is unlikely to feel the repercussions… and I doubt if “Inches of Snow at Hard Rock Stadium at Kickoff” is even an available prop bet.  If it is, take the under.

This will be only the 2nd Super Bowl to be played on Groundhog Day.  The first was the Seattle Seahawks' victory over the Denver Broncos in 2014.  Ironically, that was the only Super Bowl played outdoors in a “winter” state — New Jersey.  Unfortunately, Phil saw his shadow that morning, meaning he predicted a longer winter — however, not only did the game not feature the blizzard many feared, the temperature at kick-off was 49-degrees, which, from my perspective, is pretty cold, but to groundhog-watchers seemed to conflict with Phil’s prediction.

If we were to infer a pattern from our one data point for Super Bowls played on Groundhog Day, we might conclude that the Seahawks will beat the Broncos again this year as well.  Only problem (other than the poor logic) is that neither team is in this year’s Super Bowl.  Instead, the Seahawks’ rival, the San Francisco 49ers, will be playing the Broncos’s rival, the Kansas City Chiefs.  That’s right — California is playing Missouri.  Blue state vs. Red state.  The day before the Iowa Caucuses.  Which made me wonder about the history of Super Bowls in Iowa Caucus years… (Note: I refer to the color of various states by their votes in the 2016 Presidential general election…)

THE SUPER BOWL AND THE IOWA CAUCUSES

The 2016 Iowa Caucuses were held on February 1, the Monday before the Super Bowl.  Hillary Clinton won that Democratic Caucus with 23 delegates, while Ted Cruz won the Republican Caucus with 8 delegates.  Meanwhile, the Denver Broncos beat the Carolina Panthers the following Sunday in the Super Bowl 24-10.  That’s right — 2016 blue state Colorado beat 2016 red state North Carolina 24-10, closely mirroring the winning numbers in the respective parties’ caucuses.  Coincidence… or eerie prediction?!  Let’s check another one…

Ignoring 2012 (as Barack Obama ran pretty much unopposed in the primaries), we’ll go four years prior.  On February 3, 2008, Obama upset Hillary to win the Democratic Caucus with 16 delegates, while Mike Huckabee won the Republican Caucus with 17 delegates.  Those totals sort-of reflect the final Super Bowl score of 21-17, with the heavy underdog New York Giants upsetting the previously-unbeaten New England Patriots on February 5… only 2 days after underdog Obama upset heavily-favored Hillary in Iowa.  Those two events have to be connected, right?

It is also worth noting that the NFC’s New York Giants won the Super Bowl in both 2008 and 2012, while the AFC’s Broncos won in 2016… mirroring the Democrat Obama winning the general election in 2008 & 2012, and the Republican Trump winning in 2016.  (You’ll see where I’m going with all this soon…)  Another point to note is that both the Giants and the Broncos are from 2016 blue states, while Obama and Trump both are as well (Illinois & NY, respectively).  Wow, when you look hard enough, you can find predictions everywhere!

PAUL THE OCTOPUS AND OTHER ORACLES

Paul the German Octopus gained international fame when he correctly predicted all 7 of Germany’s matches in the 2010 World Cup, then went on to correctly predict the Spain’s victory in the finals as well.  Of course, prior to the World Cup, Paul only got 4 of 6 picks correct in the UEFA Euro 2008 -- but maybe he was just honing his skills.  While Paul’s selection method — choosing one of two boxes with a tasty mussel or oyster in it, with the flag on the box denoting his selection — may seem random, we’ll never be able to test his psychic talents under closer scrutiny, since, sadly, Paul passed away only 4 months after the World Cup (reportedly of natural causes, however it’s possible that all that divination really takes its toll).  

Mani the Singaporean Parakeet was also a star predictor in the 2010 World Cup, correctly predicting all 4 quarter-final games, then got 2 out of 3 wrong in the semi-finals and finals.  Other famous psychic sports-predicting animals, who were unable to achieve Paul’s success streak, have included Phoenix the Cat, Flopsy the Kangaroo, Mystic Marcus the Pig, and several of my friends’ dogs, fish and lizards.  My son’s leopard gecko did correctly predict that the 2019 World Series would go 7 games (by eating 7 meal worms), but incorrectly picked the Astros to beat the Nationals (by retreating into his left cave, rather than his right) — though maybe I was just confused as to which cave was for which team.

Meanwhile, another sports oracle that captured people’s attention was the “Redskins Rule” which dictated that if the Washington Redskins won their final home game of the season, then the party that won the previous presidential election would win the next election, and if the Redskins lost that final home game, the other party would win.  This rule “predicted" the 17 elections through 2000 successfully, then was wrong in 2004, so the rule was bent a little to fit the new data… then broke altogether in 2012.  So, the rule was taped back together by replacing “will” with “usually will".  What good is a rule if it can’t be amended to fit the data, right?!

One common theme of Paul the Octopus, the Redskin Rule, and other such oracles is “hindsight bias” — that is, the psychological tendency to attribute causation (and predictability) to past events because of a chance correlation.  We notice Paul’s prediction success because his choices matched the results… while we conveniently forgot Mani, and the thousands of other oracle animals who didn’t quite get their picks quite correct.  Some animal was bound to randomly make correct selections — it just happened to be a German mollusk.

"HEY, DON’T FORGET ABOUT ME!” SAYS PHIL

But Punxsutawney Phil the Groundhog doesn’t want to hear about hindsight bias.  He wants to predict the weather.  And he wants to predict other things… like football and politics!  He’d like us to look at the last 3 presidential election years, and see how he did...

What did Phil have to say on February 2 in 2008, 2012 & 2016?  Well, in 2008 & 2012 he saw his shadow, but in 2016 he did not.  And, as noted above, an NFC team won the Super Bowl in 2008 & 2012, and an AFC team won in 2016.  And we already saw how those results led to the eventual winners of the presidential elections.  Did Phil predict these results?  Or cause the results?!  I would look farther back than 2008, but then this whole pattern thing falls apart, and what fun is that?

SO, LET’S MAKE SOME PREDICTIONS!

After this extensively-researched study, which, as you will no doubt agree, adhered to strict scientific guidelines, here’s the clear conclusion:

If, on Sunday morning, Punxsutawney Phil sees his shadow (which he’s done 83.6% of the recorded years since 1887 - yes, “Phil" is quite old), then not only will winter extend another 6 weeks, but the NFC’s 49ers are likely to win the Super Bowl later in the day, one of the current poll-leaders (Sanders, Warren or Biden) will win the Iowa Caucuses on Monday, and the eventual Democratic nominee will win the general election in November.  If, however, Phil does NOT see his shadow, then we’ll have an early spring, we’ll have an upset victory in Iowa (perhaps Buttigieg or Klobuchar?), but the eventual Democratic nominee will lose to whoever the Republican candidate is in November.

You’ll be so amazed at Phil’s psychic abilities when this prediction comes true!  But, if any result does not go as predicted, then let's start a new theory to retroactively fit the updated data.  Since, if we can rely on Phil the Groundhog to predict the weather, then we sure should be able to have him predict the Super Bowl and Presidential elections as well… even if we have to keep changing the rules to make it so!

Thanks for reading! Feel free to email me your thoughts.

David Chariton