[CLICK HERE for my summary from last night, in the wake of the Monday Night Meltdown]
I mentioned this last night briefly, but I've been of the opinion for awhile now that the recent trend towards more aggression in decision making is the smart play. Such a philosophy clearly influenced the Eagles' brain trust in a couple of key moments of last night's nightmare, and I wanted to do a little digging to see if I could understand where the decisions came from.
Note: most of the research you're about to read is built on info from the incomparable stathead.com; if you like statistics at all you need to be investing in this site.
MOMENT 1: Nick passes on a chip shot FG in favor of a pass play from the Falcon's 9 yard line. We all remember this; the Eagles faced fourth and four and while most of us were a little surprised to see the offense stay out, we couldn't be too shocked. This is, after all, the MO of this franchise going back to the Dougie P era. The thinking is obvious: even with the chances of converting the fourth down being less than a successful FG, the obvious reward (7 points > 3 points, in case you're new to the sport) balances the scales.
Does it though? To properly assess this sort of decision, you need to multiply the points scored in each possible outcome by the percentage likelihood that this particular outcome will occur. A simple example: let's say Philly has fourth and goal from the one; they could choose to send Elliott and take the 2.999 Expected Points (slightly less than the full 3 because of the tiny chance of a miss), or they could run a Hurts sneak (which has about a 90% success rate). They would clearly run the Brotherly Shove, because 90% x 7 points = 6.3 Expected Points, and naturally 6.3 > 2.999.
This scenario from last night is more complex because of the fact that a successful fourth down play didn't necessarily guarantee a TD for Philly, but would have guaranteed a first and goal inside the five, a scenario that results in a TD about 74% of the time. 20% of the time such a team would still kick a FG and 6% of the time would turn the ball over. If you value a TD at 6.96 points (not the full 7 because of the remote possibility of the dreaded missed PAT) and value the chip shot FG at 2.83 points (which is Jake Elliott's career FG average for those length of kicks multiplied by three), gaining the first down yields 5.72 Expected Points. Now, you have to reduce that by the odds that you'll actually convert that fourth down; historically teams in this scenario (fourth and four from the opponent's nine) only do this successfully 27.3% of the time. About 18.2% of the time they'll actually score on the fourth down attempt. Multiply all this out (the formula looks like this: (.273 x 5.72 EP) + (.182 x 6.96 EP) ) yields a result of 2.83 Expected Points. If that number looks familiar, it should: that's basically a dead heat with the EP from the chip shot FG we wanted the Eagles to take. You also have to factor in the theoretical value of the Falcons starting inside their own ten, should the 4th down attempt fail. The difference of +/- 15 yards of field position has a value too (possibly as high as a full Expected Point). Edit: using a similar method to what's described above, I calculate the difference in field position to be worth 0.43 points. I guess what I'm saying is, if you squint you can find a justification for what the Eagles did here. The counterargument is simple: this was shaping up to be a low scoring game in which points would be at a premium. If that's so, take the easy points. Worth the debate.
One final point: if Sirianni knew that this was four down territory for the Eagles (which he clearly felt that it was), the obvious thing to do would be to call back to back running plays; Philly get the first down at least 80% of the time in that scenario. As we know, they instead opted to throw the ball twice and failed both times. With some hindsight, that mistake is likely the biggest coaching goof from this whole sequence.
MOMENT 2: the now infamous decision to throw on 3rd and three from Atlanta's ten yard line. This decision's rationale was also simple; get the FD, win the game. I don't think the decision to pass was so clear cut, though. For one thing, the Eagles had run the ball with some consistent success on that drive (five of seven running plays met the definition of "successful" in that they kept Philly ahead of the sticks). More importantly, we have data from thirty years' worth of playbooks to evaluate what was more likely to work. In this exact scenario (third and 3 from the opponent's ten), 97 teams chose to pass while 35 chose to run. 40 passers succeeded (41.2%) while 16 runners succeeded (45.7%). There's also the obvious clock management issue that we've all been hollering about for 24 hours; by passing and failing the Eagles gave Atlanta about 45 extra seconds to work with (they scored with 34 seconds left, for reference). Starting a drive inside your 40 with about a minute to play and needing a TD, the trailing team only scores about 8.9% of the time. With the extra 45 seconds, that success rate balloons to 23.9%.
If you're following me and you're mathematically inclined, here's how this breaks down:
Odds that Atlanta stops a Philly run (54.3%) x odds they score with 60 seconds remaining (8.9%) = 4.8% chance of success
Odds that Atlanta stops a Philly pass (58.8%) x odds they score with 100 seconds remaining (23.9%) = 14.1% chance of success
The Eagles' Big Brain playcalling gave Atlanta nearly 10% in additional win probability. You could argue the discrepancy was even higher than that as Philly's passing game was noticeably weaker without AJ Brown, and they've always been an excellent short yardage running team.
As we know, the coaching was NOT the only issue with the team last night. Here's a partial list of other failures:
too many penalties for the second week in a row
CJGJ blew a coverage leading directly to a score for the second week in a row
Devonta Smith dropped a TD pass in the third quarter, costing Philly four vital points
Hurts threw a game ending pick for at least the third time in two years (last year he did it against both the Jets and Seahawks)
the tackling and overall run defense was pretty poor
there's no pass rush at all, except seemingly by accident
the offense misses AJ Brown more than we hoped
The good news on this is that most of the above are at least theoretically fixable. There's too much talent for the defense to be this bad, and I do think they'll take another step forward once they're at full strength in the secondary (remember we've barely seen DeJean, Rodgers or Brown yet this year). One does wonder if the overall bend-don't-break philosophy that the Eagles have employed on defense can only work if the team has overwhelming talent to work with, which is an obvious concern if true
Looking ahead, I think we should aim for a 2-2 record going into the break, which isn't exciting to anyone (and is short of the 3-1 record I predicted three weeks ago), but both the Saints and Bucs look pretty good and we get both on the road. There's a good chance that we look back on this first month and realize we went 2-2 against several playoff teams. I don't think the Saints are quite as good as they look, and it's possible Philly pulls a big upset there (the Saints have opened as a slight favorite). Of course, we throttled a similar Tampa team in a similar spot last season.
Finally, other than the Saints and possibly Chargers, can we point to any team that's actually looking consistently good through two weeks? The Chiefs are two plays from being 0-2; the 49ers are 1-1 and dealing with lots of injuries; the Cowboys and Lions have also dropped games they were expected to win; Baltimore is 0-2 and unquestionably has more concerns than Philly right now; Buffalo nearly lost to the Cardinals; the Steelers and Seahawks haven't played anyone really good yet, etc. etc. etc. The final takeaway is that it's September and both triumphs and struggles are usually closer to a mirage than the truth. Let's see where we are in six weeks.
Commentaires