Extra Points
News and commentary from around the Web

The Manning Index (and The Brady Effect)

At the risk of starting another irrational Brady-Manning argument... This is an interesting post by Neil Paine over at Chase Stuart's site, looking at which quarterbacks have won more or less in the playoffs than what you would expect based on the Vegas line. It's great to call it the Manning Index because it has a Manning at the top and a Manning at the bottom. Also interesting is that Tom Brady isn't the only veteran quarterback who had a lot of playoff success early, and very little later in his career. He has this in common with Brett Favre, Troy Aikman, and a guy you probably don't think about much, Dave Krieg. Krieg took a 9-7 wild card team to the AFC Championship game in 1983 and went 1-1 in the playoffs in 1984, then was 0-4 as a playoff starter the rest of his career.

View Full Article


137 comments, Last at 30 Jan 2013, 10:55am

1 Re: The Manning Index (and The Brady Effect)

Any theories here on why Peyton has seemed to underperform in the playoffs? Is it something about his quarterbacking in particular, or is it just one of those things?

5 Re: The Manning Index (and The Brady Effect)

One theory: except for a few years when the Titans were good, the Manning Colts largely dominated a weak division. Thus, they had more wins heading into the playoffs than one would expect from the strength of their team.

That, and Manning has had some really unlucky playoff games. The Vanderjagt shank and Rahim Moore's matador coverage come to mind.

6 Re: The Manning Index (and The Brady Effect)

He didn't really underperform any more than any quarterback does. Playoffs are an extremely small sample size to judge any player or team on their ability or to use as an indicator of future/potential success. The Colts as a team suffered from two things in the 2000s. First, they were unlucky. This is not meant to discredit any other teams or say that it was a fluke, just that they happened to have bad games at the wrong time. Every team has them, the Colts just had their's in the playoffs. The second reason is Dungy's long-term philosophy of resting players to prevent injury. Momentum is huge in the playoffs. Not because it is a tangible thing or because in a moment during a game in the playoffs a player is thinking "oh gosh, we have momentum, we must be better than they are," but because it gives the team confidence through the downtime and during practice and creates a mind set of success, something you cannot win without. It is no coincidence that the year the Colts won the Superbowl, they played starters in week 17 and all four rounds in the playoffs.
Earlier in his career Manning played often on the road and had some bad games. But during the latter half of his career with Indy, he played quite well. Games were lost because of poor defense, poor game planning, and poor kicking. All three of these things were reasons the Broncos lost to the Ravens as well.

11 Re: The Manning Index (and The Brady Effect)

I agree with your sentiment regarding the limited utility of small sample sizes when making judgements as to quality of play, or making predictions. However, you then claim it is "no coincidence" that x preceded y, based upon a tiny sample size. Now, maybe your theory is right, and maybe it is wrong, but there isn't nearly enough data in the set known as "Peyton Manning's playoff starts as a Colt" to allow us to make any statements regarding the unlikely nature of coincidence.

(edit) If you have a large sample size of playoff games which indicates that resting starters is harmful, it'd be interesting to see.

13 Re: The Manning Index (and The Brady Effect)

People like saying that one game is "an extremely small sample size" but really, is it? It depends on what your unit of sampling is. If the unit is a game, perhaps. But I think it's far more natural to make the play the unit of sampling, just as an at-bat (or a pitch) is the unit of sampling in baseball.

14 Re: The Manning Index (and The Brady Effect)

Sure, but when people start using the final score of one game as an indicator, then, yes, one game is a miniscule sample size, especially when the one game takes place in a one and done playoff format. Even if we take passing and running attempts for qbs in a game, however, one game is tiny. Would you really want to judge the quality of a hitter based on 40 plate appearances? Even guys who play in an extremely large number of playoff games have barely more than a season's worth of passing and running attempts in the playoffs.

67 Re: The Manning Index (and The Brady Effect)

That would be interesting, i.e., getting a quantitative measure of the QBs performance on each play and see if he scored better/worse in playoffs. However, the question before the house is weather QB performance, as measured by win/loss of the *team* versus the Vegas line is any measure at all. Given a small number of games won or lost and the 'small' contribution of the QB to the total team (when among the def, special team, coaches, other offensive players) I vote, no.

92 Re: The Manning Index (and The Brady Effect)

I don't know why people like to say one game is a small sample size, but the reason you should believe it is that, whether they're based on a sample size of 50 or a sample size of 500, a player's stats are estimates of his "true" ability. For QBs, 500 dropbacks over the course of a season is simply going to produce a more reliable estimate of "true" ability than 50 dropbacks over the course of a game.

How large of a sample size do you need before various QB stats stabilize into reliable estimates of "true" ability? This blog post is required reading. You can find the exact numbers there, but the long and the short of it is that an accurate picture of a QB's ability to avoid sacks emerges pretty quickly, followed soon thereafter by his ability to make accurate throws. On the other hand, it takes a massive sample size of passes before his true ability to avoid throwing interceptions emerges.

126 Re: The Manning Index (and The Brady Effect)

I like a lot of what you say in this comment, but the problem you face is hidden deep in your assumptions. Does a quarterback have a 'true' ability? Is this 'true' ability an inter-temporal constant? I think not. Listen, it's really, really, really fun to debate this stuff, but it's generally a good idea to understand our limitations in the quest to dig deeper into what is 'truth.' People change. They are not the same exact person yesterday, today or tomorrow. Yes, at the extremes, it is glaringly obvious that Tom Brady has had much more ability over his adult years than, for instance, Tavaris Jackson, but when the differences become finer, it's not just that the statistics give you only a limited, obscured view into the 'truth,' it's that there isn't one truth that transcends the time period over which these statistics were gathered (even during one particular game in which minor injuries and variable amounts of energy may skew ability from play to play).

Not to put too fine a point on it, as you narrow down the time period to determine what is one's 'true' ability at that point in time, you are going to face larger and larger problems with inadequacy of sample size.

25 Re: The Manning Index (and The Brady Effect)

I was always struck by how often the Manning early 2000's Colts lost to the team that eventually went to or won the SB, and I am not talking about losing AFCC's. (This might explain the high 1-and-done numbers, as they didn't lose to chumps, but to teams clearly on a roll at that point.)

1999: Lose to Tenn, who wins 2 more playoff games and then loses in the SB
2003: Lose to NE, who wins the SB
2004: Lose to NE, who wins one more playoff game and then wins the SB
2005: Lose to Pitt, who wins one more playoff game and then wins the SB
2006: Colts win SB
2009: Colts lose SB

In that 11 years, they won only 1 SB, but lost to the eventual SB 4 other times. So, yes, Manning doesn't have a great playoff record, but he also didn't ALWAYS lose to bad teams in the playoffs. Of course, there was that little problem with losing to bad Chargers teams.

119 Re: The Manning Index (and The Brady Effect)

NE Since 2004:

2006: Lost to IND, who Lost SB
2007: Lost to NYG in SB
2011: Lost to NYG in SB
2012: Lost to Bal, who are going to SB

Any team that gets into the playoffs, especially with a bye, is going to have a significant percentage of their losses to the team that wins their conference. How much better in the "only getting defeated by the best" than IND would NE look had they lost the SB in 2001,03,04?

4 Re: The Manning Index (and The Brady Effect)

He hasn't underperformed. His poor W-L record is a combination of bad luck, poor defense, and bad FG kicking. Manning's own performance in the playoffs is actually quite strong.

23 Re: The Manning Index (and The Brady Effect)

Peyton had some real stinker playoff games early in his career. That's undeniable. But on the whole, if you look at his aggregate stats, they are fairly similar to Brady's. Better in some respects, worse in others.

Seems to me that both of them display stats that correlate positively with how good their respective defenses are.

8 Re: The Manning Index (and The Brady Effect)

I've always thought Manning has spent most of his career pulling mediocre teams to 13-3 records, and then losing in the playoffs because he's usually up against a vastly better team, at least on roster spots 2 through 53.

2005 is a case in point. That may have been the best Colts team Manning played for, but aside from the quarterbacks (and Ben Roethlisberger in 2005 was no slouch, for that matter) the Steelers were probably the better team.

12 Re: The Manning Index (and The Brady Effect)

I tend to be pro-Manning in these arguments, but it's a little ridiculous to say the 2005 Steelers were more talented. Manning was, by All-Pro/DYAR/whatever you want, easily the NFL's best QB. His RB was Pro Bowler Edgerrin James, who ran for 1500 yds and had one of the 100 best yds from scrimmage years ever. His WRs were Marvin Harrison and Reggie Wayne, two likely future HoFers (Harrison being a slam-dunk) in their primes. His TE was Dallas Clark, who would go on to be an All-Pro. His LT was a Pro Bowler; his center was 1st-team All-Pro. This offense, stacked with the above talent, scored 27.4 PPG -- 2nd-most in the league. Yet against Pittsburgh, they mustered just 18 points at home. I don't know how anyone can sit there and say that wasn't purely a blown opportunity by Manning and the Colts.

62 Re: The Manning Index (and The Brady Effect)

True- I remember that was a clear improvement in his game from that point on. Still, that o line performance was embarrassing beyond belief. Just terrible. A bit like the one the colts went through against the ravens in these playoffs. I want to add, I remember after Manning "threw his o line under the bus" Jaws was asked if he was being unfair. He responded with, "that was the worst o line performance I've seen in 10 years from a playoff team." Probably a bit hyperbolic, but even still, it was pretty awful.

127 Re: The Manning Index (and The Brady Effect)

I believe only one starter (Stephen Neal) left that game. On the Patriots final snap of the game, their o-lineman were Matt Light, Logan Mankins, Dan Koppen, Russ Hochstein and Nick Kaczur. '

That's four of their five starters, and Hochstein was often used in previous roles quite capably.

83 Re: The Manning Index (and The Brady Effect)

Revisionist history. At the time the games were played, Manning and the Colts were often favorites in the games they lost. And I don't recall Manning playing on bad teams he carried. Edge and Harrison were top NFL weapons, not JAGs. Now that Manning is gone, Indyfan is badmouthing the team all over the interwebs to prop up St Peyton. You'd think he was playing with the Little Sisters of the Poor team for ten years.

89 Re: The Manning Index (and The Brady Effect)

Mid career manning had some very exceptional skill position talent so you are right there.

Heres why the general impression(not just among colts fans) is that Manning seemed to carry the colts. For one, the only way they seemed to win was if the colts got a lead and let their pass rushers go to work. At no time did that defense ever feel trustworthy enough to stop the run(until ironically their sb run when their run defense that year was historically bad). As I mentioned below, Tom Brady has had poor games in the post season but his team has still been victorious. When manning had poor or even just run of the mill good games, they would often lose(again the lone exception being his sb run where he had a really bad day against baltimore and still won).

But it goes further than that. The colts have lost edge and harrison and the offense never really dropped off. Then one year the colts fielded a still pretty good passing game in 2010 despite a plethora of injuries to the wide receivers and an offensive line that was terrible. And finally, the year after he gets injured, the colts go 2-14 and are just plain terrible. All of this has built into the general impression(right or wrong) that Manning carries his otherwise mediocre teams. I don't always agree with that impression, but there is some evidence to make that suggestion.

Finally- notice what you are kind of doing. manning had great players and yet the playoff losses are all on him. Is that fair? Do we just assume harrison and wayne and edge and the o line and the defense all played well and it was just manning that sucked?

10 Re: The Manning Index (and The Brady Effect)

This is a good thread to also note again that if you take away a 46 yard field goal, a long field goal in a blizzard, and a db stupidly failing to fall down after an int at the end of a game, Tom Brady loses 7 playoff victories, and picks up two playoff losses, and his playoff record is talked about in entirely different terms. Yes, he also could have gained a few victories, and avoided some defeats, if he had some breaks go his way, breaks that have nothing to do with qb performance, but that just makes the point as well. A small number of playoff games, where there is a one and done format, is a really poor way to evaluate qb performance, especially when you are looking at the final score to provide insight.

18 Re: The Manning Index (and The Brady Effect)

And you can just as easily note that if a ball doesn't get glued to a helmet, a reliable Pro-bowl WR doesn't drop a ball while wide open, Brady would have two fewer losses, two more victories, and two more rings. If the NFL had end-zone cameras installed properly or instructed its referees correctly regarding "face guarding" rules (or the lack thereof), Brady could have another playoff victory or more.

The point is that the playoffs (usually) pit two teams that are both very good, and the smallest break either way can decide who wins, regardless of how awesome or elite the QB of either one is. Combine that with small sample size and the fact that one loss ends your chances to compete more in the same season, and it's pretty easy to conclude that playoff win-loss has very little to do with how good a QB is.

108 Re: The Manning Index (and The Brady Effect)

Very well said....Manning has certainly choked at times in playoffs when he should have had things going for him (terry porter), but he's also been very unlucky....

Take away the missed FG that forces OT against Pitt, a Buckneresque mistake by Rahim Moore and he is in two more conference championship games.

16 Re: The Manning Index (and The Brady Effect)

If you just look at the names on the index I feel confident saying that while it is interesting, it has not gathered any information about these quarterbacks. Eli Manning, Joe Montana, Joe Flacco, and Trent Dilfer do not share some subtle quality that is key to winning superbowls. I do believe that the article comes to that exact conclusion, that these numbers tell you nothing.

20 Re: The Manning Index (and The Brady Effect)

Well this settles it.

Eli > Flacco > Dilfer > Sanchez > Delhomme > Brady > McNabb > Marino > Moon > Peyton.

Brought to you by the makers of Worthington's Law.

22 Re: The Manning Index (and The Brady Effect)

This notion that playoffs people choke is just bs. It really is. Think of what the narrative of Manning would have been if Denver had stopped that deep pass. By that point, Manning would have had 3tds and 1 int(which was the result of an uncalled defensive holding). Instead, the game extends and he throws a terrible int to effectively end it.

I think Brady is the shinning example of why judging people by playoff performance can be so damn meaningless. Again, anyone arguing brady of 01 is in the same timezone as brady of 2011 is just an idiot. Bill Simmons conveniently labeled brady's prime going from 2003 to 2008 and the rest his "decline". It makes me wonder if he's even watching the same patriots.

24 Re: The Manning Index (and The Brady Effect)

Bill Simmons knows next to nothing about football. He's admitted as such in print.

And you bring up the signature point. The Brady from 2000-2005 is clearly worse than the Brady from 2006-present. Yet according to the "rings" metric, the former Brady was better than the latter.

The correct conclusion is that the New England Patriots from 2000-2005 were better than the New England Patriots from 2006-present. Why it all gets put on the QB is beyond me. It's crappy analysis.

26 Re: The Manning Index (and The Brady Effect)

I agree, and Brady's career is the ultimate poster boy for this. He is so clearly a better QB in the second half of his career, and yet he won all his rings in the first half. If you rank QB's purely by SB wins, then you have to say that Brady has gone downhill, which he has not.

So often with these QB's, they highly influence the game, but just one 50-50 play (or unlucky play) going the other way, and who knows. For example, when you look at TB and PM, you get this list of events which "change the narrative" of their abilities based on events they had no influence over:

1) PM watches MV miss a 49-yard FG in OT which would have beaten Miami. Colts lose on ensuing possession.
2) PM watches Big Ben tackle, on a fumble return, a guy whose wife stabbed him in the leg the night before, leading to ...
3) PM watches MV shank a 46-yard FG in a dome, failing to tie the playoff game as time expires.
4) PM's teammate, a bench warming WR with a hot wife, can't hold onto the on-sides kick to start the second half against NO. Might not have changed the game completely, but was a huge play that PM had no control over.
5) PM watches his now Broncos give up a 70-yard TD bomb with 35 seconds left because a safety can't judge the ball.
6) TB watches AV hit a 45-yard FG (in the snow) to win playoffs leading to eventually ...
7) TB watches AV his a 48-yard FG as time expires to win SB against Rams.
8) TB watches AV hit a 41-yard FG as time expires to win SB against Carolina.
9) TB watches Baltimore's kicker shank a 31-yard FG to tie 2011 AFCC game in the last second.

Obviously, you could fine 10 others that flip the script to even better TB success and PM failure, but man, that is quite a list of 50-50 plays in close games that create a narrative.

27 Re: The Manning Index (and The Brady Effect)

Are the early era patriots really better than these late era pats? I'm not so sure. Its easy to say in hindsight that they are worse, but think about what the narrative was of the pats coming into their playoff losses. 07 was the greatest team I've still ever seen(going back to 1998). 2010 NE had one of the greatest offenses of all time and what was(at that point) a massively improved defense. 2012 ne had an average defense but again an excellent offense. The fact that they lost is really to me more of an indication that so little meaning can really be drawn from the playoffs with regards to how good a team really is.

Take the 49ers against the rams this year. The first game ended in a tie with both team's posting excellent offensive dvoas but poor defensive dvoas. In their rematch(again going into OT), the script was switched, with both offenses posting negative dvoas and the defenses doing well. Or think of the 07 giants when they faced the pats. First game is an offensive explosion. The next game is a grind it out defensive struggle.

The point? Game to game performance even among rematches is so incredibly volatile that it makes you realize how much the game of football itself is so random. One has to wonder, is 16 games even a large enough sample size to accurately gauge a nfl team? Considering that all the other major sports have 82 and 162 games, 16 games might honestly tell you nothing.

In short, imo, clutch doesn't exist. Choking doesn't really exist. guts glory whatever you want to call it. SB wins are nice, but even they don't tell you much if you really dig deep. Sometimes, you just need the dice to roll your way.

30 Re: The Manning Index (and The Brady Effect)

Clutch and choking undoubtedly do exist (at least as concepts), its just that they are so often and so obviously overused and misused, particularly in team sports where any individual player can only exert so much influence on proceedings.

But I've definitely witnessed, for example, tennis players who are able to consistently elevate their game at crucial moments, and others who consistently fall short at these moments.

33 Re: The Manning Index (and The Brady Effect)

I'm very suspicious regarding the existence of "clutch performance", if that means someone whose performance consistently gets better when the stakes are biggest. I know for aa fact that choking exists, if "choking" means someone whose performance declines when experiencing the anxiety of playing for the biggest stakes. I doubt that anyone is clutch, but I know everyone chokes, some more reliably than others.

42 Re: The Manning Index (and The Brady Effect)

I don't see why clutch can't exist. Some people don't truly focus unless the stakes are high, at which point they stop wasting talent and actually perform. Some people just really like the limelight.

Granted, these people are usually disasters in their personal lives, but there's no reason it cannot exist.

43 Re: The Manning Index (and The Brady Effect)

Well, I didn't say it cannot exists. I just don't see much evidence of it among elite athletes. In contrast, I see choking all the time. I think the phenomena is most interesting among the elite of the elite, as they age. They begin to lose the absolute confidence they once had in their physical abilities, and then, when the anxiety builds, they completely fall apart, and look like a beer league weekend bumbler.

45 Re: The Manning Index (and The Brady Effect)

I agree with BJR that both of them exist. Take chess, which is purely mental and therefore and excellent example. I also used to play at a relatively high level, so I can tell you from first hand experience that many players elevated their performance when under great pressure (I mean when they were running out of time to make plays). That, in turn, made them intimidating, which had the effect of causing them to appear as if they had a "winner" aura -which they felt themselves, turning it into a "vicious" circle. They also had ability and preparation, of course, but they point is that not all talented players were able to do the same. And some, like me, certainly choked.

I've given the matter some thought and I have come to the conclusion that, under heightened pressure, some people disconnect their minds and enter a particular state of being where their abilities flow effortlessly -you might equate it to meditation. Others are unable to disconnect their minds under similar circumstances. Meanwhile, a third group experience doubts -possibly as a result of intimidation or some sense of unworthiness- which cause them to fail (choke).

The funny thing, in my case, is I suffered from philosophy at crucial moments. See here: Trolls and philosophy

FO posters are a peacock. You got to let us fly!

47 Re: The Manning Index (and The Brady Effect)

Well, I can't speak to chess, but in two types of competition in which there are decades of detailed information to examine, among the elite of elite competitors, baseball and golf, there is extremely little evidence to supoort the proposition that clutch performance exists. I doubt that the phenomena could be absent in those two types of competition, and exist in other types of competition.

48 Re: The Manning Index (and The Brady Effect)

I should have noted that mostly my observations were subjective and anecdotal. But, again using tennis as an example (this is one of the sports I am most familiar with) it is known that Roger Federer wins more end of set tie-breaks than he ought to given his ordinary point win %, over a fairly large sample size. Pete Sampras was also well known for his recurrent trick of cruising through a set, just making sure he held serve, then stepping up his game when it really mattered. Statistically, he broke serve more often than he should have done at the end of sets.

Perhaps these players are able to simply maintain their normal high level of play, whilst their opponents suffer from 'choking'. Is this enough to define an athlete as 'clutch'? I don't know. It's a pretty inexact term.