That’s fine with me.
The long snapper had everything to do with why they went for a 4th and 5 on the 30 instead of going for the field goal when the score was 14-9. 17-9 puts that game away, but they didn’t have anybody to snap to the holder, and that absence was painfully demonstrated when they had no choice but to punt and Harrison sent it out of the endzone.
You can hype “clutch play” all day long, but two scores down in the 4th the way the defense was playing? Forget about it.
When did 8 points become two scores?
Also, you may not have noticed this but the Giants scored 21 points, 19 without the safety. 19 > 17.
How silly of me. Of course they were going to get the 2-point conversion. They’d been doing so well up to that point at the goal line, it was a done deal.
Fine, so this miss the two-pointer, so they don’t get the PAT they wouldn’t have tried for, and they don’t get the safety. That’s STILL 18 points, which is still more than 17.
Which they might not have gotten without that lame-o free kick that resulted from the safety.
You are assuming that the touchdown that resulted from the great field position after the safety was a foregone conclusion. Had Pittsburgh had the snapper to kick the field goal they start 9 points down from their end of the field. Instead they started with good field position, pooched it anyway, trapped Pittsburgh deep, got a safety, and got dynamite field position for their game-winning touchdown.
Circumstances in totality are important. Nevertheless, the Giants won, and I’m happy for you. Just keep the Cowboys out of the playoffs and I’ll be happy, Pittsburgh is going to win their division anyway.
I happened to post on giants.com Eli’s late&close stats just yesterday, citing them at 87.1. This stat comprised approximately 197 attempts. Today he has 207 attempts and is listed at 87.4, so he’s continuing to climb by that metric.
In order to assess this going forward we need a starting point. So here are Eli’s current stats:
Career : 1124 of 2028, 86 TDs 68 Ints: 75.1
Late&Close: 129 of 207, 14 TDs 10 Ints: 87.4
Let’s just use those two to make bookkeeping easier. After each quarter season, I’ll revisit this and update his progress going forward by subtracting the above numbers out of the updated totals.
Please explain how punting from his own 9 is going to result in worse field position for the Giants than punting from his own thirty, because that’s what you’re arguing here.
If the safety didn’t happen, that means the punter caught the snap and punted. He was standing on (or around) his 9 when it sailed over his head. But because it did sail over his head for a safety, the very next play had him line up and punt a free kick from his own 30.
In short, if they hadn’t given up a safety, the Giants would have gotten better field position than they got from the free kick.
On that we’re agreed.
I’m arguing that had there been a long snapper the Steelers would never have had to punt, they would have made the field goal, gone up 9, and Jeff Reed would have kicked off. None of the other events would have happened.
Coulda woulda shoulda. It really doesn’t matter. I’m just saying that the outlook wasn’t as rosy for the Giants as you seem to think it was if Pittsburgh had someone to snap the ball.
This makes absolutely no sense. The Giants took over on their own 32 after the 4th down conversion that should have been a field goal attempt. The three times the Steelers kicked off the Giants started on their own 20, 35, and 26, for an average starting position of their own 27.
Are you really saying that because the Giants started on such great position as their own 32 instead of their own 27 things would have gone differently? Really? You’re hanging the outcome of the game on 5 yards?
What makes that even more ridiculous is that Giants ended that drive with a 24 yard field goal. I guess you don’t think Carney could have hit a 29 yard field goal?
A_D, are you saying they could have kicked a field goal from their own 20 instead of punting? Because that’s more or less where they were when they were forced to punt.
Oh, I meant it’s fine with me that you don’t have any doubts. I thought you wanted to hear the standard argument against clutch, so I gave it. You don’t really have to keep any books to prove anything to me – in fact I wouldn’t accept a few games’ worth of “close and late” QB rating as statistical proof of anything anyway, especially when the difference and the sample is so small – and I’m equally unworried about disproving it. I can save you the time and stipulate that he can have a QB rating of 95 over your sample, and I’m not going to be impressed.
A slightly different take on “clutch,” just as a supplement to Chitwood’s points, which I basically endorse:
Being an internet sports-stats geek, I am contractually obligated to express skepticism about “clutch” performance, and to mock those players who are most often described as clutch by giving them cute nicknames (e.g., Derek Jeter = “Clutchy McClutchenstein”). And I do, in fact, perform these duties.
However, I think there’s a certain level of agnosticism about the issue that is only healthy. There’s room enough for a small number of players to be effectively “clutch” to a small or moderate degree and have their data points be mistaken for statistical noise. Certainly, for example, my *subjective *impression of Eli Manning is that he performs better in high leverage situations. While I concede that I could never come close to proving this to be true, I think there’s a fair chance that it’s an impression with some predictive value.
On the other hand, there are some factors that may make an NFL QB *appear *clutch, both on paper and on the field. For example, a disproportionate number of offensive snaps in “close and late” situations will occur form the Shotgun formation, and a large majority of offenses perform better from the Shotgun than they do under center (and some QBs have *huge *splits between their shotgun performance and under center performance, which alone is likely responsible for many a “clutch” reputation). Similarly, a disproportionately large number of offensive snaps in these situations will occur while the offense is obviously in “4-down” mode, and a team’s passing statistics will be elevated if the defense has to worry about making 4 stops rather than 3, and if they’re willing to concede a lot of underneath stuff to defend against the long passes.
These last problems are something DVOAcan help with, as it only rates performance in comparison to what’s expected in a given situation (point differential, time remaining, field position, etc.). Unfortunately, what relevant information I can gather from their database does nothing to support my impression that Eli is better in high leverage situations.
They don’t yet provide splits for player stats at FO, but for paying members you can get a lot of splits for offenses or defenses, including in close & late situations (defined as all 2nd Half snaps with the point differential at 8 points or less). Since 2005, the Giants offense has performed almost exactly as well in “close & late” plays as it has overall:
. . . Clutch – Overall
'05 – 8th . . . 9th
'06 – 9th . . . 9th
'07 – 18th . . . 19th
'08 – 2nd . . . 1st
Obviously there are objections one could make: the running game is included in these stats, and starting “close & late” at the beginning of the 3rd Quarter isn’t what most people have in mind when they think of clutch situations. Still, this is evidence against Clutch Eli – in the absence of more or better information, how much weight one should give this evidence is a judgement call.
Rather his own 20.
Same difference; either is better position than his own 9.
So you don’t believe in clutch because you think it has no predictive value, and you will not accept any evidence to the contrary regardless of what it is. Gotcha.
If this logic had merit, one would expect everyone to have a higher “close&late” passer rating, right? In reality, it ain’t just McNabb who tails off in the category. How do you explain that?
Better information? What would that be? Any information at all? What you’ve described is wholly irrelevant to the question.
Oh no, wait, Eli hands off like a motherfucker in close&late situations, which of course mean up by 8 points at the start of the 3rd. Seriously, even with your disclaimer, you should be ashamed to bring that data to the table.
You really are one of a kind, you know that? I had included a preemptively sarcastic version of almost this exact response in my earlier post, but I decided, nah, no need to jump the gun.
But whatever, yeah; “regardless of what it is” is exactly what I said. You want a lecture on statistics, take a class.
Save your high and mighty act. What you said was far more ridiculous than how I paraphrased it. To quote:
I’m positive there is evidence he would accept – he’s saying that there’s hardly any way that the numbers for one player could be distributed that would constitute proof for clutch ability. People – by which I mean serious people who understand the math – have gone looking for evidence of clutch ability (mostly in baseball) and come up empty.
No, you wouldn’t expect *everyone *to have higher numbers in those situations, just most players. To my mild surprise, however, it would seem most QBs perform worse in “close & late” situations (as defined by Yahoo) than they do overall. Looking at 17 current QBs with at least a a little bit of starting experience, only 5 of them had higher C&L ratings than overall: Eli, Romo, Trent Edwards, Delhomme (barely), and Kerry Collins (also barely). Among the QBs with worse C&L ratings: Tom Brady, Peyton Manning. Among the QBs with *much *worse C&L ratings: Brett Favre, Kurt Warner, Chad Pennington.
I kinda doubt the utility of this particular metric for judging clutch ability. (Or is Kerry Collins preferable to Brett Favre with the game on the line?) Do we even know what Yahoo defines as “Close and Late”?
Nonetheless, like I said, I am surprised by the results of my informal survey. I’m willing to bet, however, that there’s something systemic about the way football is played that’s depressing those numbers (as opposed to most QBs suffering from the yips in the clutch).
Of course there are problems with the data as far as relevance to our specific question goes, but to say that it’s wholly irrelevant is an exaggeration. For those ratings to be identical despite Eli’s superior passing in the 4th Quarter of close games, then the Giants’ running game in the 2nd Half and passing game in the 3rd quarter has to be correspondingly inferior – do we have any reason to believe that’s the case? I’m not saying it’s gospel, I’m just saying it’s the data I happened to look up and it points the other way.
A more general point: you shouldn’t get prickly with me about this. I’m a lot more sympathetic to your point of view than other fans of my persuasion, and, if you hadn’t noticed, I said I agree with you about Eli.