Just How Hard Is a Perfect Season?

Another week, another win.  Even though they kept it competitive, the Bucs never really felt that threatening today at Lambeau.  Nobody picked them to win, and, in the end, you could sense they didn't expect to themselves.  But underdogs do win sometimes.  In the Packers perfect season, it's easy to lose sight of that fact.

Today marked an interesting milestone: the 16th win in a row for the Packers dating back to last year -- or the equivalent of a full regular season.  It begs the question, just how hard is that to do?  What are the odds of winning 16 in a row?  What are the odds of going undefeated this season?  What would a mathematician put our odds at?  Why am I asking so many questions?

The reality, statistically speaking, is that the Packers will eventually lose a game this season.  I hope it doesn't happen.  I don't expect it to happen.  But from a strict, numerical, probabilistic, algorithmic standpoint, it is likely to happen.  Before proceeding any further in this article, I should warn the reader that I am about to attempt math.  You'd think with a last name like Trigg (o-nometry), this would be a forte of mine.  Not so much.  So be warned, someone could get hurt.

A US quarter: AKA the "Manitowoc Abacus"

Let's start with a baseline.  There are two teams in every NFL game.  If the contest is perfectly evenly matched, each team has a 50% chance of winning -- same as the flip of a coin, 50/50.  Hopefully, I'm not going too fast for anyone.  If you assume a 50% chance of winning in every game, then going undefeated over a 19-game season (16 regular season games + 3 playoff games) is 0.5 to the 19th power.  (That's that little superscript "mini-me" number that always screwed you up in algebra and I can't figure out how to do in HTML).

Now 0.5 to the 19th is a small number.  I know that because when you get the little "E" thing in Excel instead of actual numbers, it means it's a really small number.  How small?  (Hold on, Manitowoc, I'm about to blow your mind!)  It's a 0.0001907348% chance, or 1 in 524,288, that an average NFL team goes undefeated.  If, like most red-blooded Americans, you don't believe in math, try flipping a quarter 19 times and getting heads every time (then try getting tails 19 times, and call me in about 35 years once you've finished).

Now, you say, "yeah, but the Packers have better than a 50% chance of winning every week."  Of course.  That's what good teams do.  They improve their chances of winning slightly above the 50/50 baseline odds.  How much more likely are the Packers to win, on average?  Maybe 60%?  70%?  80%!?!  It's hard to imagine anyone other than the most aggressively optimistic or inebriated Packer fan asserting an 80% chance of the Packers winning every week, against every opponent, for an entire season.

But here's where that "19th power" thing really messes with your brain: even at 80% odds of winning each week, the odds of winning an entire season is only 1.441151881% or 1 in 69.  Don't believe me?  Check out my handy cheat sheet.  Excel doesn't lie, although it does drive you freakin' crazy when you're trying to get the IRR function to work at 2AM.  No wonder Vegas sports books are so good at taking everyone's money.  The actual likelihood of unlikely things is hard for our simple brains to get their cortexes around.

Occasionally Rodgers throws an interception.  Once in a while, opposing running backs break 7 tackles on the way to a 54-yard touchdown run.  Sometimes desperation onsides kicks are recovered.  In short, in the wise words of a baseball hat I once saw at a shop up in Door County with a plastic turd on the brim, "Shit Happens."  Once again this week, all you need to do is look around at the fate of some other top teams -- the Giants losing at home to the 3-6 Eagles, or Jay Cutler breaking his thumb and likely out for the season -- to remind yourself how fortunate the Packers are to be sitting at 10-0.

An undefeated season certainly won't be easy.  And if it happens, it would likely be a once-in-our-lifetimes event.  But although it's not probable, anything is possible.