Let’s keep things simple here.
You scored a touchdown to cut New York’s lead to four points. Kick the extra point and make it a three point game. Now, you’ve ensured that you can tie the ballgame with a field goal during an ensuing possession.
But no, Doug Pederson decided to go for two, and Carson Wentz got himself sacked, which meant the Eagles’ deficit remained at four and resulted in them needing a touchdown instead of a field goal at some point later in the game.
It seems to defy logic, and Sunday afternoon Doug explained it this way:
“The decision making there was to obviously trust my guys upstairs. We had just converted one before and felt good with the play call. You go for it and make it when the field goal can win the game for you later on in the fourth quarter. At least it gives you an opportunity to know what you need to do to win the game.”
And here’s his response Monday morning:
“So we have a two-point chart based on points and win probability. Down four, to go down two in that situation, your win probability goes way up and that’s why I made that decision to go.”
Well yeah, no shit! If you go down two, you can kick a field goal to WIN the game. But converting the two is HARDER than kicking the extra point, which is EASY. And when your offense is lumbering up and down the field and nothing is guaranteed, Doug, please take the points however you can get them.
More than anything, we need to take a look at this two-point chart the Eagles’ analytics department is using, because ESPN’s data is saying something completely different:
ESPN's Win Probability had the Eagles at 37.7% had they attempted the XP and 37.3% by going for 2 after scoring to make it 21-17. So pretty much a wash numbers-wise, anyway. — @ESPNStatsInfo
— Tim McManus (@Tim_McManus) November 15, 2020
Let’s get the ESPN dorks in the same room with the Eagles dorks and have them figure this out, because two models are apparently telling us two different things.
We’re not anti-analytics at Crossing Broad. We really aren’t. The problem is that you can’t become a slave to the numbers at the expense of using common sense and trying to read the flow and momentum and emotion of a game. No data set can properly parse the non-linear human factors I just listed.
What if you’re just a shitty conversion team? Or you run out of plays, as Doug said a few weeks ago? Put that on the chart.
At this juncture in the game, Doug should have said to himself, “look, we just had a good drive, we’re back in it, let’s kick the extra point, make it a one-score deficit, and try to get a defensive stop here.” They ultimately DID get that stop, but then found themselves on their own five yard line at the start of the fourth quarter, needing to go 95 yards for a touchdown instead of a much shorter distance to set up a game-tying field goal attempt.
Analytics are fine, and we’re always looking for new ways to approach sport and help decision making and roster building. But sometimes you just need to keep it simple, do the obvious thing, and use basic instinct, like Sharon Stone, to feel your way through the macro-level shifts that just cannot be packaged nicely into a data sheet. There must be a balance between data and common sense gut feel for the game.
No idea why they go for two in that situation.
— chris long (@JOEL9ONE) November 15, 2020