The NBA and NFL are currently in a similar spot. The Detroit Lions have played eight games and have a record of 2-6. The Detroit Pistons have played eight games and also have a 2-6 record. With the NBA, though, the sample is considered too small to reach a definitive conclusion (although I think I am pretty sure the Pistons are not going to be great this season). For the NFL, the sample of games played represents half of the eventual population. Yes, eight games is still not much of a sample. But we are also not going to get much more, so let’s start jumping to conclusions.
The Lions Are Immensely Improved
As noted, the Lions are only 2-6. Last year the Lions were 1-7 after eight games. And in 2008 the Lions were 0-8 at the midpoint. So the Lions have improved, but in terms of wins and losses the improvement appears modest.
In evaluating teams, though, we know that wins and losses aren’t the entire story. For example, the Pistons have been outscored by seven points per game. That suggests – as I indicated above – that the Pistons are not really that good this season.
The Lions, though, look much better when we consider margin of victory (MOV). So far the Lions have scored 203 points while allowing 188. In addition, Pro Football Reference tells us the Lions Strength of Schedule (SOS) ranks the highest in the NFC. Pro Football Reference also has a measure – the Simple Rating System (SRS) – that combines MOV and SOS. And according to SRS, the Lions – with a 4.5 mark — rank 3rd in the NFC. Yes, the 2-6 Lions – according to SRS – are the third best team in the NFC.
Last season the Lions had a -14.4 SRS. Not sure what the record is for improvement in SRS, but one suspects that a leap of 18.9 is pretty impressive.
And it is even more impressive when we consider the history of the Lions. The Lions have not had a positive SRS since 2000, or the year before Matt Millen came to Detroit with a mission to destroy the franchise. And the last time the Lions had an SRS that was as high as what we see in 2010, it was 1995 and the Lions were led by Barry Sanders; and before 1995 the last time the Lions bested the 4.5 mark was 1981 (when the team was led by Billy Sims). In sum, although the Lions are only 2-6, this is actually one of the better Detroit teams in my lifetime (and yes, that in itself is somewhat depressing).
The QB Story
When we think about why a team succeeds or fails, we often look at the quarterback position. For the Lions, quarterbacks historically struggle. And that is the same story this year.
Matthew Stafford – the number one pick in 2009 — has only appeared in three games this season, after missing six games in his rookie season. And when he has played, Stafford hasn’t posted amazing numbers. Last season Stafford posted a 0.151 WP100 [Wins Produced per 100 plays]. Average WP100 in 2009 was 0.497, so Stafford was well below average. This season he has improved this mark to 0.315. Yes, it is better. But yes, it is still below average.
Stafford – once again – has missed most of the season. The quarterback who has taken the majority of snaps for the Lions is Shaun Hill, and he has done more than Stafford. But as the following table indicates, Hill’s WP100 of 0.391 is still below average.
Although Hill is below average, he is still posting a higher WP100 than Brett Favre (the QB with the most experience) and Sam Bradford (the QB with the least experience). Of course, quarterbacks are inconsistent (a point made in The Wages of Wins and Stumbling on Wins). So these numbers can change (although they might).
For now, though, the top quarterbacks in the game – in terms of WP100 – are Michael Vick, Phillip Rivers, and Kyle Orton. And it appears Tony Romo does a bit more than Jon Kitna (as Wade Phillips discovered).
One last note on quaterbacks… if we look at the difference between the NFL’s metric (QB Rating) and WP100 we can see who is the most overrated and underrated. Topping the list of overrated quarterbacks is David Garrard. Eli Manning should also be seen as overrated. The list of underrated signal callers begins with Donovan McNabb and Mark Sanchez.
Again, this is football. So these numbers will probably change. Maybe that wouldn’t happen if NFL teams played 82 games, although I suspect the interaction effects are so large in football that even with a much longer schedule players would be relatively inconsistent (by the way, intereaction effects in the NBA are much, much smaller). So if you don’t like where a quarterback is rated, wait a few more game. By the end of the season this list might look quite different (then again, maybe it won’t).
– DJ
Russell
November 11, 2010
The only game they have lost in the 4th quarter was the first one of the season. In the other games, the Lions enter the 4th quarter loosing. Most of those games, the Lions make up ground in the 4th quarter. Football is not played in a vacuum. Coaches tend to play more conservatively to hang onto 4th quarter leads. If the Lions had played a little better, the opposing teams might have followed suite. I’m sure you know this which is why I find it confusing that you come to the conclusion that the “Lions are a top team” based largely on point differential.
In reality, there is only one measure of a football team’s success. That is their win-loss record. Teams do not play to have a good point differential. Being a “top team” should not be based on such. If the playoff seeding was based on point differential, I’m sure GMs would design different teams and coaches would use different strategies.
The strength of schedule makes for a better argument, but there are flaws in that as well.
Adam C Madison
November 11, 2010
I think football’s a mostly terrible sport and rarely watch it nowadays, but when I did I visited Football Outsiders.
http://footballoutsiders.com/dvoa-ratings/2010/week-9-dvoa-ratings
They say the Lions are 19th.
I’m calling shenanigans on this one.
Dan
November 12, 2010
I do agree with the numbers, but then how come they can’t beat other teams that are worse than them?
Daniel
November 12, 2010
Whatever formula you’re using needs to be updated, especially for quarterbacks. McNabb has been so bad in real world productivity that he was benched in favor of Rex Grossman. That really should speak for itself but it doesn’t. The Skins/Bears played one of the ugliest games in memory and only luck kept McNabb from throwing 6 interceptions in that game.
I would tell you to stick to basketball, but your rebound worshiping is no more accurate to real world basketball than this football “analysis”.
Tom
November 13, 2010
Donovan McNabb, in the game where he was benched, had roughly twice as many rushing yards as the rest of the Redskins combined. The lack of appreciation for that ties into what is written in Dave Berri’s books about the under-appreciation of QBs’ yards on the ground. It’s also worth noting that when you have no rushing offense, it makes things a bit more difficult for the quarterback. Under the circumstances, McNabb’s performance in that game was pretty good.
As for the win-loss record being the only measure of a team’s success, it’s the only one that counts, but if your goal is to predict future performance, it’s generally not very effective. There is a lot of luck in sports. Over the span of an 82-game basketball season it evens out, but over the span of a 16-game football season it doesn’t.
jbrett
November 13, 2010
McNabb’s benching starts a national firestorm of debate, and is variously explained as justified by performance, racist, a message/insult from the coach, over-coaching by ‘The Ultimate Leader’ (from his Denver bio) , a panicky blunder, and a ‘do something even if it’s wrong’ move. To Daniel, however, it’s proof that WP100 ‘needs to be updated.’
Always nice when the Flat-Earth contingent shows up, and puts in their two shillings.
Daniel
November 14, 2010
Apparently only the “flat earth contingent” still watches sports. I guess if you’re an enlightened one, you don’t need to watch sports and the numbers tell the whole story!
Get over yourself, sycophant. Just examine the last few games. When exactly was the last time he played well? It wasn’t the Lions game. It certainly wasn’t the Bears game (2nd worst performance by a QB all year, only “topped” by his opposite number). Wasn’t vs. Indy either. He had nice stats vs. the Packers but padded his stats by keeping his team behind until it was “pass every down” time.
See, in sports like football you actually have to WATCH THE GAMES to see how a player is playing. All incompletions are not created equal. Nor are all interceptions. As I mentioned above, only incredible luck kept him from having a truly horrendous day vs. the Bears. People who watch the games can tell that McNabb is not a good QB right now. People who look at stats think he’s “underrated”. I know who I put my money on.
O.Sagat
November 14, 2010
So after the Lions
a) put up 6 points in 59:30 of play
b) Lose to the Buffalo Bills
you want to re-evaluate that claim about the Lions relative power?
jbrett
November 14, 2010
“I know what I see.” That’s what you’ve got? Thanks for the insight. I feel so enlightened now. Just watch the games; why didn’t I think of that? all these years I’ve been wasting my time following the latest in statistical analysis when I could have been–
Oh, wait; all these years I HAVE been just watching the games, because until recently that and the boxscore were the only tools available. But thanks for the tip; I’m sure I still have my stone knives and bearskins around here somewhere. Hey, everybody, Daniel’s just saved us a mess of time and effort; let’s all just watch the games, and then we’ll all agree about everything–just like we always did, right?
Daniel
November 14, 2010
Any time you want to answer that question about the last time McNabb’s had a good game, that would be great.
jbrett
November 14, 2010
Don’t know; don’t care. Don’t think you know, either, but if you think you do, enjoy that feeling. My point–which has played out on this blog many, many times before, and will unfortunately continue long after you’ve lost interest–is that you started by lobbing opinions like hand-grenades, and followed up with ‘he fails the eyeball test.’ Old news. You have no idea how tired and hackneyed that argument is here–and useless as debate, because none of you can ever be swayed by ANYTHING. You saw it, you drew your conclusion, a finger of fire etched it in stone. Fine; you know all you need to know. Take the sermon somewhere else; a lot of us agnostics prefer formulae to faith.
There are hundreds of fan-mouthpieces on the web; why walk in the front door of this one and start with disparagements of the architecture and insults about the furniture? It’s tiresome, it’s rude–but most of all, it’s unproductive; it leads to nothing more than futile gainsaying of one another. No matter what I might think, I would never come to your NASCAR memorabilia auction and loudly proclaim it a waste of time and money; it would be inappropriate, pointless, and insulting to the sensibilities of those present. Yet, every day, people think nothing of popping on here and screaming, “Hey! Berri! All this information that derives from the work you do for a living? My eyeballs say it’s a pile of crap!” Not, “I don’t understand,” or even,”I disagree, for these reasons;” no, most go straight to, “You’re an idiot, because the work you do contradicts my fanaticism.”
Speaking only for myself, thanks for clueing me in; could you find the next thing to insult now? Surely there’s a referee somewhere who would enjoy obscenities hurled at him, or a department store cashier whose supervisor you can demand to speak to about his attitude? Wouldn’t that be fun? Worth a try, right? Please? Come back later, and tell us all about how you put someone in his place. Or don’t.
Daniel
November 15, 2010
So let me get this straight.
You know nothing about the topic at hand (McNabb’s recent play). Yet you feel the need to chime in about the correctness of the formula (a formula cousin to the one that says that the now 2- 7 Lions are one of the 3 best teams in the NFC) in spite of the evidence provided by real gameplay.
You also have not contradicted the point that in a sport with a sample size as small as football, the luck factor is extraordinarily high. Going back to that Chicago/Washington game again (which you should see at some point if you want to continue this argument), McNabb was fortunate to not throw 6 interceptions instead of the 2 he was credited with.
I mean really all you need to say is “I don’t understand the games I’m watching and need a formula to tell me what is what” and we can end this. It’s clear enough to me that this is what you’re saying. Until then, keep waiting for McNabb to “turn it around” and keep waiting for the Lions to make the playoffs. Reality is that a way.
DKH
November 15, 2010
Hey, Daniel, let’s try translating your points into something that maybe resembles constructive criticism instead of uneducated insults:
1. Does the formula for rating teams accurately evaluate low-win-percentage teams with low margins-of-loss to good teams? Perhaps the Lions recent loss to the Bills (which may be within the home-field advantage point value) suggests otherwise, or maybe the Lions are a solid team with bad injury luck.
2. “with a sample size as small as football, the luck factor is extraordinarily high.” I think that’s a good point, and probably other posters here wouldn’t disagree (although they may call it “variability” or something other than “luck”). At any rate, is it possible to see a variability measure for the WP100 statistic?
3. “McNabb was fortunate to not throw 6 interceptions instead of the 2 he was credited with. ” I don’t know what you’re referring to, but is it dropped interceptions by a defensive back? That isn’t that uncommon, really, and it’s probably part of the variability in results. Sometimes, 2 out of 2 bad passes are intercepted; sometimes it’s only 2 out of 6.
4. As a general point, it’s not unreasonable to watch some games and then rely on statistics to keep informed of other games/teams. Assists can be subjective stats in basketball. Are you ok with looking at the box score the next morning to see how many assists a player had, or do you insist on seeing every play yourself?
============
As a final request, can we maybe see WP100 numbers for Colt McCoy and Max Hall? I understand their sample sizes will be smaller than everyone else, but they were or are starters for a period for their teams and I’d like to have the numbers (maybe also Jake Delhomme and/or Seneca Wallace for comparison).
jbrett
November 15, 2010
Daniel,
Sorry to keep you waiting; we all know astrology is a pile of crap, so I’ve been going on astrology websites all day telling them what idiots they are. Oddly, they seem unconvinced, and not terribly interested in debate.
You raised some actual questions this time, so let me respond:
1. You call Pro Football Reference’s Simple Rating System “cousin” to Berri and Schmidt’s WP100. Why? They’re derived completely separately; they measure different things. If you believe we’re all descended from Adam, or Noah, then maybe we’re all cousins–but we’re not exactly closely related. Maybe you should go back and read the post more slowly.
2. I would never oppose an argument that, in football, “the luck factor is extraordinarily high.” I am puzzled by your apparent position that McNabb is only affected by GOOD luck (dropped INT and such). I’m also confused as to how you can also hold the position that YOU can see through all the lucky bounces to who can play and who can’t. Congratulations! Is that a marketable skill?
3. My own opinion is that whether it’s you, me, or Mike Shanahan, we’re all applying the most inexact science in the history of the world; how else do we explain a benching one week and a contract extension the next? Obviously he won’t really get the guaranteed 40 million; but the guy you applauded for benching him just gave him more money! What do you know that The Ultimate Leader doesn’t? What does HE know now that he didn’t last week? Perhaps you should go on his website and tell him he’s an idiot; I’m sure that will be helpful and constructive, and cause him to reevaluate his decisions–possibly his entire life.
4. I don’t have a rooting interest in McNabb, the Lions, any other subject of the article, or anyone else you have mentioned; I’m certain I never suggested I did, and I’m completely in the dark as to where you got the idea I had a dog in this fight. My issue, right from the beginning, was your fan-centric, obnoxious dismisssal of a metric I now realize you had zero comprehension of–and we’re right back where we started. You can use your eyes; there’s obviously no curve to the landscape or the horizon; to suggest there is would clearly be ludicrous, so any measure that suggests the Earth isn’t flat “needs to be updated.” Certainly can’t argue with you there.