The watchtower: lions vs. redskins

>> 9.22.2009

In last week's watchtower, we looked at the historical matchups of Gunther Cunningham defenses versus Brad Childress offenses, and Scott Linehan offenses versus Dungy-style Tampa 2 defenses. Here's what I came up with:

Given greater or equal talent, Scott Linehan’s balanced offense significantly outperforms its averages when facing a Dungy-style Tampa 2, especially with the run. Given lesser talent, Linehan’s offense meets or mildly outperforms expectations against a T2. However, a disproportionate amount of sacks and turnovers seem to be created by a Tampa 2 when facing a Linehan offense. It’s that pesky turnover thing that will make the difference; if Minnesota sacks Stafford five or six times and generates two or three turnovers, the Lions will have an extremely hard time keeping pace, even if the ground game is working well (and scheme or no, nobody runs on the Williams Wall!). However, Minnesota’s lack of a downfield passing game should allow Gunther Cunningham to turn up the defensive heat to an extreme level, which should have a disproportionately disruptive effect on the Vikings’ offense. Given how amazing Adrian Peterson looked in the Vikings’ first game, and how inept the Lions looked against the Saints, it’s tempting to say this will be a blowout—however, the Vikings didn’t blow out the Browns, and I don’t think this one will be a blowout either. It appears as though the Lions have a decided systemic advantage on both sides of the ball, assuming Gunther feels safe enough to crank up the heat. It remains to be seen if those advantages will be enough to overcome the gap in talent on both sides of the ball. Therefore, the most likely result of this game is a closely contested, medium-to-low scoring slugfest, with a lot of turnovers and penalties. It is slightly more likely that Minnesota’s talent overcomes Detroit’s systemic advantages, but this will be a volatile game in Detroit’s home opener.

This method of prediction--breaking down historial scheme vs. scheme data, and layering on subjective analysis--has its flaws. Indeed, the brighter analytical minds of the football universe have been privately critical of my work so far. While I take said critiques to heart, I think I may be on to something.

  • I found that Linehan's offense seems to be unusually successful against Dungy-style Tampa 2 defenses; Detroit was so successful on the ground that the Vikings abandoned their base defense, and put eight men in the box.
  • Linehan's teams seem to turn the ball over frequently when facing a Dungy-style defense.  I noted that avoiding that would be key to the Lions' chances for victory.  Instead, the Lions turned it over three times, and that was arguably the difference in the game.
  • I projected that Minnesota's lack of a consistent deep threat would allow Gunther to blitz, thereby disrupting the Childress offense. It did, he did, and it did; the Lions nearly carried a shutout into halftime. In the end, the turnovers--and resultant short fields for the Vikes--were too numerous to overcome.
  • It was indeed an ugly game marred by sacks, turnovers, and controversial calls--most notably the phantom Gosder Cherilus "chop block" that negated a long gainer by Megatron.

This analysis isn't anything more than that: analysis. But rather than pretend that Dominic Raiola had some sort of secret Pat Williams kyrptonite in his pocket, I prefer to believe that the interaction of schemes provided an opportunity for Kevin Smith to succeed . . . and I know of no other way to project or predict such interactions. I said as long as I find this analysis to have predictive value, I'd continue--and I do, and so I shall.

Zorn Gun Ornk PpG YpA YpC Drnk PpG PTS YpA INT YpC FumL Sack
SEA KCC 12th 21.1 6.47 3.45 13th 20.1 31 7.2 0 2.1 0-0 3-9
SEA KCC 19th 20.0 5.84 4.27 19th 22.1 17 5 1 5.5 4-1 5-37
SEA KCC 19th 20.0 5.84 4.27 19th 22.1 22 5.33 3 5.7 1-1 3-16
SEA KCC 14th 20.9 5.96 3.97 11th 19.7 28 6.6 2 2.94 0-0 2-8

Jim Zorn, former Lions quarterbacks coach, has only been an offensive coordinator for as long as he's been a head coach--this is his second season as anything more senior than a position coach. However, he's been a top assistant to Mike Holmgren in Seattle since 2001, and runs that flavor of Walsh-style offense in Washington, so I decided to extend my reach back to Seattle-era Holmgren offenses.

This opened up something I've been anxious to see: two meetings between the same schemes in the same year.  In 2000, Holmgren's Seahawks met Cunningham's Chiefs twice; this should give us an idea of how reliable our single-sample data has been.  I don't want to get ahead of myself though, so let's look at the prior meeting first.

In 1999, the Seahawks brought the 12th-best scoring offense (21.1 ppg) up against Gunther's 13th-best scoring defense (20.1 ppg). You'd expect the Seahawks to match their usual output, and they roughly did, scoring 24 points. While they outperformed expectations through the air (6.47 ypa avg., 7.20 actual), their already-weak running game was completely ineffective (3.45 ypc avg., 2.14 actual). They didn't simply abandon the run, either—that’s 60 yards on 28 carries! The three sacks might be interesting, but paired with zero turnovers, it doesn't point to anything systemic.  Let's get to the really interesting case study: 2000.

In 2000, the Seahawks were a below-median scoring offense, posting up exactly 20 points per game (19th). Meanwhile, Cunningham's defense was also ranked 19th, allowing 22.1 points per game. The offense scored very near expectations: 17 and 22 points in the two games.  In both meetings, the offense passed below their season averages: only 5 and 5.33 yards per attempt, versus the average 5.84. In a reverse of the 1999 data, though, they ran with success far above average: 5.47 and 5.72 yards per carry, over the average 4.27. The disruption numbers, however, are consistently high: 5 & 3 sacks for -37 and -16 yards, 4 & 1 fumbles forced (1 and 1 recovered), and 1 & 3 INTs.

What are we to make of this? I'm excited by the strong parallels in the stats between the two games. This is exactly what I was hoping to see; it buoys my assertions that the 16-game averages can roughly capture annual swings in talent and execution. There's a quirk, though: when the Seahawks had a strong passing game but weaker running game, the run was shut down--but the passing attack outperformed expectations. Then in 2000, when the Seahawks were much better on the ground than through the air (thanks Shaun Alexander), the Chiefs took away the pass, but were steamrolled with the run. Before drawing any conclusions, I'll go to the last data point.

In 2006, Seattle had a fairly balanced offense, scoring 20.9 points per game (14th), averaging 5.96 yards per pass attempt and 3.97 yards per carry. Cunningham's Chiefs were the 11th-ranked scoring defense, allowing 19.7 points per game. Though the Seahawks, at first blush, exceeded expecations, 7 of their 28 points came from a defensive TD--putting them right where you'd expect them, at 21 points. Now, we see a reversion to the 1999 pattern: the Seahawks passed for 6.60 YpA, 0.7 yards above average--and rushed for a measley 2.94 YpC, 1.03 yards below average.

I looked at the 1999 data, and the Seahawks were ranked 24th in the NFL in rushing attempts (25.5 per game), and 24th in yards per carry (3.45). In 2000, however, the Seahawks ran almost exactly as often (25.19 CpG, ranked 23rd), but were markedly more successful when doing so (4.27 YpC, ranked 9th). It seems as though Cunningham was almost "taking what the offense gives him"--simply blitzing both the run and pass, generating many turnovers and sacks, interrupting the scoring success of what a team does well, and denying what they don't. As this is the guiding principle of Cunningham's defense--extreme blitzing, and a focus on touchdown denial vs. yardage denial--I'm willing to say that given equal talent and execution, there is no systemic advantage for either Jim Zorn's WCO or Gunther Cunningham's hyperagressive 4-3.

If this week's game follows the above trends, Clinton Portis will have an above-average day on the ground, but the Lions' blizting will disrupt and confound Jason Campbell and the Redskins' ho-hum passing game. Point production by the Redskins should be right at expectations--which, for 2009 so far, means 13.0; they're ranked 31st.

Lin Bla. Ornk PpG YpA YpC Drnk PpG PTS YpA INT YpC FumL Sack
MIN CHI 8 24.4 6.60 5.3 25th 23.7 19 8.36 2 4.24 2-1 1-6
MIN CHI 8 24.4 6.60 5.3 25th 23.7 24 7.20 0 3.4 1-1 0-0
MIN CHI 6 25.3 8.18 4.71 22nd 21.6 25 8.23 0 5.18 3-1 3-16
MIN CHI 6 25.3 8.18 4.71 22nd 21.6 10 6.53 1 4.68 1-0 2-7
STL WAS 30 14.5 5.67 3.95 6th 18.5 19 5.23 0 2.92 1-1 2-12

On the other side of the ball, the analysis unfortunately gets murkier. 'Skins DC Greg Blache has been in Washington since 2004, but only as the defensive line coach until Zorn took over. Moreover, Blache was not a product of the Gregg Williams "tree"; he was a refugee from Dick Jauron's ouster in Chicago. While Blache allegdly had significant input into the Williams defenses of 2004-2007, Blache apparently removed many of the more exotic blitz packages, and called fewer blitzes in 2008. Further, the signing of Albert Haynesworth and drafting of Brian Orakpo bolster the notion that Blache's scheme and playcalling will more closely resemble the stout but staid defense Blache put in place during his time in the Windy City . . . therefore, I'll use Linehan's matchups against the pre-2004 Bears defenses, and the 2008 Redskins for my analysis.

In 2002, Linehan's Vikings met Blache's Bears twice, again giving us a good look at in-season trends. Linehan's Vikings were ranked 8th in the NFL in scoring offense, with 24.4 points per game. They passed for 6.6 yards per attempt, and rolled with an outstanding 5.3 yards per carry. The Bears scoring defense was subpar; they allowed 23.7 PpG, slotting them 25th in the league. You'd expect the Vikings to score well above their already-prodigous average, but don't; they only muster 19 and 24 points in two tries. While they were very successful in passing the ball (8.36 & 7.20 YpA, 6.60 avg.), the vaunted Vikings ground game ground to a halt (4.24 & 3.4 YpC, 5.3 avg.).

Disruption numbers were low (2 picks; 3 fumbles, 1 lost; 1 sack for -6 yards for both games COMBINED), but it's undeniable that the Blache defense smothered an excellent Linehan ground game with far inferior talent, and the success of the passing game wasn't enough to do anything but equalize the offensive output to normalcy.

In 2003, both teams were slightly bettter; Minnesota was the 6th-best scoring offense in the NFL with 25.3 PpG, and Chicago was the 22nd-ranked scoring defense with 21.6 PpG allowed.  The passing attack was extremely potent, averaging 8.18 YpA, and the running game was still solid (4.71 YpC), despite losing over half of a yard per carry from the previous season.  Expectations would be that the Vikings outperform their usual scoring average—and in the first game, they do, scoring 25 points.  However, in the second game, they took a pratfall, mustering only 10. 

Here we have the first discrepancy between two contests in the same season.  In the first matchup, the Vikings slightly outperformed their 16-game averages both through the air and on the ground (8.23 YpA & 5.18 YpC, vs 8.18 & 4.71 avgs.).  However, in the second contest, the Vikings managed a mere 10 points, passed for only 6.53 YpA, and ran for an average 4.68 YpC.  The only real variance between the two games that I can find is that the first game happened in mid-September in the Metrodome, and the second happened in mid-December at Soldier Field.  It’s entirely possible that the climate and crowd were decided factors in limiting the Vikings’ high-flying offense—though I admit, that’s nothing but conjecture.

There's one more data point to look at, so let's do: last season, just two games after Linehan's ouster, his Rams faced Blache's Redskins. I realize that this won't perfectly capture Linehan's playcalling tendencies, etc., but it's the only data point we've got with Blache leading the Redskins' defense. The Rams were fairly well wretched that season--the 30th scoring offense, ekeing out just 14.5 points per game. The passing offense could only manage 5.67 YpA, and the rushing game was only good for 3.95 YpC. Meanwhile, Washington's defense was the 6th-best scoring defense in football, allowing only 18.5 PpG. One would expect a truly wimpy offensive output, and yet the Rams scored 19 points. How?

Well, their only touchdown came from a interception return, and the remainder of the scoring came from kicker Josh Brown's foot. YpA and YpC were both depressed (5.23 YpA vs. 5.67 avg.; 2.92 YpC vs. 3.95 YpC avg.). Disruption numbers again weren't great, but it hardly mattered; the defense did their job. Therefore, I'm concluding that regardless of talent or execution, Greg Blache's philosophy of a strong front four and committment to run-stopping disproportionately slows Scott Linehan's balanced offense. Where does this leave us? With the burden entirely in the hands of the respective quarterbacks. Given the Redskins' meager offensive output so far, and the lack of any decided systemic advantage for either the Redskins' offense or the Lions defense, the Redskins will move the ball only if Jason Campbell can be successful deep and force Gunther Cunnigham to call off the dogs. Meanwhile, the Redskins' impressive apparent advantage when matching up their defensive system against the Lions' offense means that Matt Stafford will have to minimize turnovers and connect with Calvin Johnson deep.

In my opinion, the most likely scenario is an absolutely brutal game, a physical brawl where both teams try but fail to control the ball with the running game, sacks and turnovers abound, penalty flags fall from the sky like rain . . . and the team whose quarterback performs the best wins.


Pacer,  September 22, 2009 at 9:04 PM  

Ty-here we go again. From Don Banks at Sports Illustrated". He is one of their NFL feature writers and this was in today's SI.

"I'm going out on a short limb here and predicting this is the week the Lions finally win. I say the 19-game losing streak ends Sunday, when Detroit is home against Washington. Obviously the Redskins are offensively challenged and defensively strong. But I just get the feeling the Lions are about to get a breakout type game on offense from the likes of quarterback Matthew Stafford, receiver Calvin Johnson, tight end Brandon Pettigrew and running back Kevin Smith."

IMO-if the Lions cut down on their mistakes they have a chance but one day the Redskins are going to break loose with that offense. I hope it is not this Sunday.

Nonetheless your synopsis mirrors others' thoughts as well. Should be interesting.

Nice Work.
PS-posted this on your previous article as well.

Neil,  September 22, 2009 at 9:40 PM  

I have the feeling that the one guy who is just going to kill us all day is Chris Cooley. I think we have a decent shot if we can reasonably shut everything else down, but, man, I wouldn't be surprised if Cooley hauls in, like, 10 passes for 110 yards or something like that.

I'm going to pick a Lions win on Armchair Linebacker in my preview post later this week, by the way. Which should tell you that I learned nothing from last week's hysteria induced bout of insanity.

Anyway, whatever problems these posts might have with sample size and all that, they have been so far, if nothing else, at least interesting. Keep posting them, keep refining them, and see where it goes. I don't see the harm in at least putting it out there.

Jeremy Reisman,  September 22, 2009 at 9:49 PM  

Just want to say this is probably my favorite Lions analysis segment on the internet. Great job with the research, and keep it up.

BigAlBundy,  September 23, 2009 at 8:19 AM  

This brings back memories from the days I was seriously handicapping the thoroughbred ponies.

Ended up designing a computer handicapping program based on results from the Daily Racing Form. What I did was enter into computer the results from the previous year. Then when the new year started entered those results also. Then some time past the racing season's halfway point began putting money down based on computer selections.

Lost money at first because after the race there always seemed to be something new that cropped up that wasn't considered in the original logic of the computer program.

For example you're using what rank an offense or defense is based on the entire league, what's missing from that ranking is "strengh of schedule." And I would suggest that might have something to do with overall ranking.

In horse racing that's termed class of horse. What I mean is in horse racing two horses might run identical races, but a stakes horse will beat an allowance horse which in turn will beat a claiming horse every time.

Keep going with your analysis, because after 3 or 4 years of entering results into computer from Ladbroke DRC I could make bets with confidence based upon just about any situation that occured, and with a tidy profit.

Lets say theres a cheap $3500 claiming race in which there are 2 horses that tend to try to get an early lead and then hope to outlast the late chargers. Also in the same race there are 2 late charging horses that fall back early but come on strong down the stretch. Also theres one horse in there who tends to stay close to the lead and has a little bit of kick left at the end. What horse to bet? What horse will win that race 80% of the time barring unforeseen circumstances such as a horse breaking down? It will be that horse that stays near the lead that has a little kick at the end. Handicap it, bet it, collect it, but it will take time to have enough data to make accurate analysis consistently and I didn't mention whether the track was fast or muddy which comes into consideration also.

Also like in horse racing are there any scratches from the race? What I mean is, did the teams in your ranking of offense or defense have any injuries, and in the game being analyzed will any of those injured guys be back thereby upgrading the offense or defense? Tons of aspects to consider, not all can be considered when first coming up with the idea, but keep going with it.

Footnote: Bob Raymond whos currently in the Lions P/R department was the head of Ladbroke's DRC's P/R department when I bet the ponies.

Big Al - the Bundy experience

Isphet,  September 23, 2009 at 8:24 AM  

I think the game goes as expected, but a few big plays by Portis on missed tackles makes the difference this week. Whoever can break the short, simple stuff for longer plays is going to win I think.

I think the Redskins are going to be fired up after getting hammered in the media all week and having the Lions actually picked by a few people to win. They HAVE to know that if the Skins lose this game; Jim Zorn is officially on the short list to getting canned.

Anonymous,  September 23, 2009 at 9:08 AM  

All I can say is wow. Great stuff and please continue. I really enjoyed.

Ty,  September 23, 2009 at 9:51 AM  


Yeah, that's great to read. It seems like if the Lions' D that showed up last Sunday shows up against a Redskins squd that could only put 9 on the terrible Rams, all the Lions' talent *should* be able to outscore them.



Ty,  September 23, 2009 at 10:06 AM  


Thanks! I AM going to keep refining it, because it really does seem as though, even if there's no objective predictive value, it's a unique and interesting way of breaking down the game.

You make a good point about Cooley; I'm not sure how he'll be covered. Peterson? Dizon/Levy (Sims will be out)? Man him up w/Delmas? Mix it up? Very very interesting . . .


Ty,  September 23, 2009 at 10:16 AM  


Wow! Thank you. That's really high praise, and I deeply appreciate it. I hope it continues to improve. Thanks for reading--and thanks again for taking time out to let me know you like it!


Sanman20,  September 23, 2009 at 11:28 AM  

Wow what an amazing piece! Keep it up Ty. Love your blog.


Ty,  September 23, 2009 at 1:29 PM  


1) that's a brilliant idea w/the ponies!

2) I understand that the per-year ranks and averages are themselves subject to small sample size issues (only 16 games) . . . you're right, though--the relative strength of schedule will play a role as well. I think I'm capturing the major trends (e.g., the 2003 Minnesota Vikings were really good at running the ball and scoring points) but there's a lot of grayscale that's not getting factored in with all these averages.

I really don't want to build a database and resort to data visualization.

3) as far as the scratches, mud vs. fast track, etc., I'm spot-checking for that stuff, especially when I see unusual results--like when the terrible Rams offense faced the pretty good Redskins defense and put up 19 points. I went "huh?" and saw there was a defensive TD; stuff like that.

I'm investigating going to Football Outsiders' DVOA, which actually uses teams of volunteers who chart every game, discount what should be discounted and calculate performance on a per-play, relative-to-the-league-average level:

Basically, I've been trying to sand the corners off of square rocks to make wheels, and these guys have a showroom full of Corvettes. The problem is, these are partially premium stats, so I haven't yet explored exactly how much is available for free, and how much I can re-use for my site, even if I did pay for it.

Either way, I'm going to keep adding and I'm going to keep refining . . . in a few weeks' time, when the averages settle down a bit, I'll start including '09 data, at least as an interesting-to-look-at sidebar.


Ty,  September 23, 2009 at 2:10 PM  


That's a good take. I do question that Portis will succeed where All Day failed, but they'll almost certainly keep grinding with Portis because they won't have another effective option. I really hope the Lions' offense can find a spark, get a couple TDs early, and put the Skins back on their heels.


Ty,  September 23, 2009 at 2:11 PM  


Thanks, man! I really appreciate you taking the time to read and comment. It keeps me going.


Matt,  September 23, 2009 at 4:01 PM  

On Cooley: He's definitely going to be the one the Lions have to watch out for. The Vikings TEs were largely absent (aside from the TD to Shiancoe), but was this good Lions D or just the Vikings' gameplan? New Orleans used their TEs effectively (Shockey and Dave Thomas), but they used EVERYONE effectively.

Bottomline: The Lions have given up 3 TD passes to TEs and haven't yet faced one of Cooley's caliber. Much like with Harvin/Berrian, the corners should be able to (mostly) handle Moss/Randle El. But what do you do about Cooley (Sims' absence, at least, takes away some of your options)?

On Portis: He's kinda' like a significantly down-graded version of Peterson. You DO have to start by stopping/slowing him, but it should be much easier. Portis just isn't on AP's level (at least not anymore) and hasn't looked like himself so far this season. Washington's line also is weaker than Minnesota's and just lost Randy Thomas. Like with Minnesota, the Lions' best bet vs. Washington is to put the game in the hands of the QB. They can use the same general blueprint and it should prove more effective. On top of that, this QB isn't nearly as good as the last one.

All this being said, I like the Lions' chances this week. I'm not ready to predict the victory, but they've got a better chance than they did against Minnesota. Whichever way it goes, Ty, you're absolutely right that it'll be ugly.

Swive,  September 26, 2009 at 8:17 AM  

Ty -

Your analysis is well done, from one of my favorite blogs.

Swive,  September 26, 2009 at 8:22 AM  

See what happens when you click "post comment" by accident? I wanted to add this thought about those "experts" who have been critical of your analytical techniques. Galileo was taken to a court and threatened with torture and excommunication if he continued publicizing his reasoned analysis that the earth moves around the sun, not the opposite (as the church put forth). He recanted, but on his way out of court he was heard to mutter "nonetheless, it still moves".
The proof is in the pudding - and I will be reading your analyses with great interest as we move forward farther and farther away from the Millenium (well, it seemed that long anyway).

Ty,  September 29, 2009 at 9:45 AM  


Sorry I'm so late coming to these comments! I'd be interested to hear your take on the 'Skins gameplan, since Cooley seemed to be a nonfactor--they pretty much tried to beat us with the run and the deep pass instead.

I can tell you this: Zorn's playcalling is not highly regarded amongst Skins fans.


Ty,  September 29, 2009 at 9:54 AM  


Your analogy was HIGHLY appreciated, just so you know. For what it's worth, when I take criticism from the folks behind Football Outsiders and Chris at Smart Football, I assure you I'm the one who has learning to do.

HOWEVAH, I think I'm really on to something here. I'm more of a a "fan beat writer" blogger than a "fan Quality Control coach" blogger--so I lack the technical background to really pull this off. But, that won't stop me from trying!


Post a Comment

  © Blogger template Simple n' Sweet by 2009

Find us on Google+

Back to TOP