Throwback Thursday: The Patriots won Super Bowl XLIX in improbable fashion. How did their chances compare to perhaps the most famous victory ever snatched from the jaws of defeat?
We’ve had four days now to digest what happened in Glendale, Arizona on Sunday. We all know what happened; down 28-24 with 26 seconds left on the (still-running) clock, Russell Wilson threw a pass to receiver Ricardo Lockette. Cornerback Malcolm Butler – an undrafted rookie – jumped the route, intercepted the ball, and just like that, the New England Patriots won Super Bowl XLIX. What looked like yet another Super Bowl defeat, in Arizona, on the heels of an improbable catch, instantly turned into celebration and euphoria from the Patriots sideline all the way across the country and into my living room. To call it “improbable” doesn’t do it justice.
On this Throwback Thursday, I’ll investigate just how improbable it was, compared to one of the most improbable endings in championship history, the 1986 World Series.
First, a disclaimer. Calculating win expectancy based on the situation at hand is far from an exact science. Typically, when we talk win expectancy, we’re doing so in a vacuum. We look at past scenarios where teams faced exact or similar circumstances, figure out how often those teams won and lost, and boom, there’s your win expectancy. The situation, the magnitude, the specific players involved – nearly all of that is kept out of the equation. As such, you’d get the same expected win percentage for the Patriots and Seahawks as you would for the Patriots and a group of toddlers. If other teams did it in the past, logic dictates that’s how the situation will play out, regardless of the players involved. So, with all of that in mind, just remember; these figures aren’t perfect. Still, it’s fairly close, and can at least show you the ballpark of how unlikely the outcome was. Let’s move on.
Let’s set the scene. Seattle trailed by four, 28-24, with the ball on New England’s one-yard line. With three downs to work with, Seattle needed to gain a single yard to take the lead and, almost certainly, win the Super Bowl. When Russell Wilson snapped the ball on second down, there were 26 seconds left on the clock.
I used the win probability calculator at Advanced Football Analytics, because it was the first one to show up on Google. After entering the score, time left, down, and distance, the calculator says Seattle had an 88 percent chance at winning the game. Now, this site certainly isn’t perfect, and it appears situations where a field goal was kicked were included, which obviously wouldn’t have happened. Moving over to the calculator at Pro Football Reference, the same situation produced a 93 percent win probability for Seattle. Let’s split the difference, and say Seattle had a 90 percent chance to win the game, beginning the moment the ball was snapped.
To me, this sounds about right. Nine times out of ten, you would expect Seattle to score a touchdown and not allow the Patriots to come back to win the game in regulation or overtime. So, Tom Brady and Co. won a game they statistically had about a 10 percent chance of winning. Unlikely? Yes. Impossible? No.
In fact, when we look at today’s Throwback Thursday, New England’s chances looked downright excellent.
Flashback to Saturday, October 26, 1986. The Boston Red Sox lead the New York Mets, 3-2, heading into Game Six of the World Series. A win in either game – both in New York – would give the Red Sox their first title in 68 years.
We’ll skip the first nine innings, which ended in a 3-3 tie. In the top of the tenth, Boston took a 5-3 lead following a Dave Henderson home run, a Wade Boggs double, and a Marty Barrett single. Heading into the bottom of the tenth, the Red Sox led by two, and were three outs from their first World Series in a nearly seven decades.
Using the win expectancy calculator here, the Red Sox were looking at about a 93-94 percent chance to win the game heading into the bottom of the inning, leading by two. This site uses data from all MLB games from 1957-2013 (excluding 1999 for reasons which are unclear). Keeping in mind that this data includes the game in question and games which have occurred in the 29 years since, we learn there have been 642 examples of teams trailing by two runs to begin the bottom of the tenth. Those teams have won 43 of those games, or 0.06697. Things looked bleak for the Mets.
Boston’s Calvin Schiraldi, in a rare third inning of relief, retired the first two Mets he faced, meaning the Red Sox now held a two-run lead with two outs and nobody on. Our calculator now gives the Mets a 1.9 percent chance to win. According to the data, through 2013, 362 teams faced such an uphill climb. Seven of those teams won, and one of those seven was these Mets.
As someone who grew up a Red Sox fan with a Mets fan father, I can pretty much recite the next string of events by memory, despite being born five years after the events. Gary Carter singled to left (Mets win % now up to close to 4 percent). Pinch-hitter Kevin Mitchell was next, after (rumor has it) being dragged from the clubhouse where he was booking a flight home. Mitchell singled as well. With runners on first and second, the Mets now had the tying runs on base. Win probability: five percent.
The next batter, Ray Knight, was down to his final strike before hitting a single of his own. Carter scored to make it 5-4, Mitchell advanced to third, and Knight stood at first. The Mets now have about a 20 percent chance of victory. Schiraldi was then removed for Bob Stanley. Things did not go well.
The first batter Stanley faced was Mookie Wilson. Things went south quickly. Wilson and Stanley battled, eventually leading to a 2-2 count after seven pitches. Stanley’s eighth pitch nearly hit Wilson, instead bouncing at his feet, past catcher Rich Gedman, allowing Mitchell to score the game’s tying run and moving Knight to second. Not shockingly, the Mets were now favorites with a 60 percent chance to win.
We all know what happened next:
Wilson hit a dribbler to first, the ball went through Bill Buckner’s legs, Knight scored, my dad went crazy. The Mets went on to win Game Seven and the series, and the Red Sox made Buckner public enemy number one until they finally won a championship, 18 years later.
Obviously, the situation for the Patriots and Mets were different. The Patriots led the game, despite the very, very low win probability. The Mets, obviously, trailed. In both situations though, teams snatched victory from the jaws of defeat in the most improbable – and, for one fan-base, gut-wrenching – way possible. Both games included some second-guessing of coaches (leaving Buckner in for defense/throwing a pass from the one-yard line). Both included fans going through the first few stages of grief, only to be brought back to life. Both games involved someone in my family having a heart attack, probably. Very similar.
While sitting in my living room, watching my favorite team give up another soul-crushing miracle catch in the last minute of a Super Bowl, I was nearly ready to admit defeat. When Seattle had the ball at the one-yard line, with the league’s top running back, I thought there was almost no chance the Patriots would win. And there was; they had about a 7-10 percent chance. All you need to do is ask the 1986 New York Mets though; as long as the chances aren’t zero, the game isn’t over. The Patriots win in Super Bowl XLIX was improbable; the Mets win in the 1986 World Series was nearly impossible.