≡ Menu

Week 10 Game Scripts and Average Field Position Data

Austin and the Rams blew out the Colts

Austin and the Rams blew out the Colts.

We have a new leader in the clubhouse for most lopsided game of the season. The St. Louis Rams pulled off one of the biggest blowouts by a heavy underdog in league history in week ten, defeating the Colts in Indianapolis, 38-8. In the process, the Rams also held an average lead of 23.2 points, the largest Game Script score of the season.

Indianapolis kept it close early, and the only first quarter score came via the St. Louis defense. On that play, Robert Quinn — who with 12 sacks through 10 games, is a legitimate Defensive Player of the Year candidate — stripsacked Andrew Luck, and Chris Long picked up the fumble and raced 45 yards for the touchdown. Incredibly, the Colts are lucky this game wasn’t even more one-sided. Late in the first quarter, Kellen Clemens and Zac Stacy botched the exchange on a handoff on the goal line with the Rams looking to go up 14-0, and Indianapolis recovered to end the scoring threat. That didn’t set back the Rams for long, however, as St. Louis scored 21 points in the third quarter to take a 28-0 lead into the locker room. Tavon Austin — who had a day for the ages — scored in the third quarter to give St. Louis a 35-0 lead early in the third quarter, effectively ending any hopes for another Luck comeback.

Three teams lost with positive Game Scripts in week 10, but unlike in week nine, there were no big comebacks, as all three games were back-and-forth affairs.  The Panthers won with the worst Game Script of the week, holding an average deficit of 2.6 points against the 49ers. San Francisco jumped out to a 9-0 early, but Carolina eventually won 10-9 on a late field goal. Since I wrote about how the 3-9 Panthers were about to turn things around, Carolina has gone 9-3. In an unrelated note, I recently injured my hand on my back.

The table below shows the Game Scripts data from week 10:

Winner
H/R
Loser
Boxscore
PF
PA
Margin
Game Script
Pass
Run
P/R Ratio
Op_P
Op_R
Opp_P/R Ratio
STL@INDBoxscore3883023.2183633.3%551479.7%
SEA@ATLBoxscore33102312.5274239.1%381670.4%
NORDALBoxscore49173211.5423852.5%271662.8%
BALCINBoxscore201739.5413057.7%563164.4%
PHI@GNBBoxscore2713149.3213736.2%443059.5%
JAX@TENBoxscore292729.2262947.3%432662.3%
DEN@SDGBoxscore282089382263.3%333350%
TAMMIABoxscore221936.4223836.7%441377.2%
PITBUFBoxscore2310135.9343350.7%422265.6%
ARIHOUBoxscore272431.8332854.1%462168.7%
DET@CHIBoxscore211921.1352558.3%512071.8%
NYGOAKBoxscore24204-0.9253839.7%302554.5%
MINWASBoxscore34277-1.2282453.8%413653.2%
CAR@SFOBoxscore1091-2.6362955.4%272452.9%

Only two of the teams with positive Game Scripts passed more frequently than their opponents. San Diego is normally one of the most pass-happy teams in the NFL, as Philip Rivers (who still leads the league in completion percentage) executes a high-percentage passing game that takes the place of many runs. But boldness apparently has its limits, and against Peyton Manning and the Broncos, the Chargers called 33 pass plays and 33 runs1 in an attempt to keep the Denver offense off the field. That helped San Diego win the all-important time of possession battle, 38:03 to 12:57, which enabled San Diego to still lose by 8 points.

The other game that went off Script was in New York, where the Giants remain terrified that Eli Manning will throw a pick six at any moment. He threw another one against the Raiders, but the story of the day was Tom Coughlin giving perhaps the most injury prone running back in the NFL, Andre Brown, 30 carries in his first game of the season. For what it’s worth, 16 of New York’s 38 rush attempts came with the team leading in the fourth quarter. Over the last four weeks, New York has passed on just 52% of all pass plays, but actually passed less frequently in the games with negative Game Scripts.

The Ravens had a very high pass/run ratio for a team that held a large lead most of the game, due to the continuing ineffectiveness of the team’s running backs. The passing game has not been good this season — Baltimore ranks 7th in pass attempts but only 17th in yards and 20th in touchdowns — but the running game isn’t getting the team anywhere. Against Cincinnati, Ray Rice and Bernard Pierce combined for just 61 yards on 26 carries.

As always, you can view the Game Scripts from every game this season at this page.

Team
Boxscore
# plays
Avg Yardline
New Orleans SaintsBoxscore8056.2
Washington RedskinsBoxscore7754.3
Minnesota VikingsBoxscore5253.6
Seattle SeahawksBoxscore6953.2
Buffalo BillsBoxscore6451.8
Oakland RaidersBoxscore5549.3
Philadelphia EaglesBoxscore5849.2
Pittsburgh SteelersBoxscore6748.6
Jacksonville JaguarsBoxscore5548.5
Green Bay PackersBoxscore7448.2
Chicago BearsBoxscore7147.9
San Diego ChargersBoxscore6647.8
Arizona CardinalsBoxscore6147.5
Baltimore RavensBoxscore7146.8
Detroit LionsBoxscore6046.8
Tampa Bay BuccaneersBoxscore6046.7
Indianapolis ColtsBoxscore6946.6
New York GiantsBoxscore6346.4
Cincinnati BengalsBoxscore8746.2
Tennessee TitansBoxscore6944.6
Houston TexansBoxscore6743.3
St. Louis RamsBoxscore5442.8
Carolina PanthersBoxscore6541.8
Atlanta FalconsBoxscore5441
Dallas CowboysBoxscore4339
Miami DolphinsBoxscore5738.7
Denver BroncosBoxscore6037.6
San Francisco 49ersBoxscore5134.2

Three things stand out to me:

  • The Saints set a record for first downs with 40 against the Cowboys. New Orleans’ nine drives (excluding the end of game drive) began, on average, at the Saints 22 yard line, which I suppose enabled Drew Brees to pick up all those first downs. But the fact that the team’s average snap took place so far down field despite the poor average starting field position is yet another sign of how dominant the offense was on Sunday night.
  • San Francisco was entirely ineffective on offense, and I’m not sure you can even blame the team’s average starting field position. The 49ers first possession began at the 20 yard line, and the team picked up 46 yards and kicked a long field goal. But after that, the Panthers defense completely shut down San Francisco. The most yards picked up by Colin Kaepernick and the 49ers on the team’s final 12 drives was just 17 yards. The two other field goals scored by San Francisco both began on drives in Panthers territory.
  • I poked fun at them before, but to the extent San Diego’s goal was to win the field position battle, well, they succeeded. It just didn’t seem to matter. Denver had 11 drives, and the best starting field position for the Broncos was the team’s own 27 yard line. The average Denver drive began on the Broncos 19, which kept the average field position data down, and San Diego held Denver to 22 or fewer yards on six of those 11 drives. The problem? Four of the other five went for 70+ yards and a touchdown.
  1. This excludes the fake punt run by Eric Weddle and the aborted play, charged as a fumble/run to Rivers. []

{ 19 comments… add one }

  • Arif Hasan November 14, 2013, 4:33 am

    SOS Adjustments. Attempting to wrap that pretty tag you used in last week’s article.

    Rank---Tm-----SRS
    1---------CAR---   6.84
    2---------SFO---   6.09
    3---------DEN---   4.66
    4---------NOR---   4.64
    5---------GNB---   3.91
    6---------SEA---   3.79
    7---------DAL---   2.81
    8---------KAN---   2.54
    9---------NWE---   2.13
    10--------CIN---   1.43
    11--------PHI---   1.37
    12--------ARI---   0.75
    13--------DET---   0.43
    14--------ATL---   0.24
    15--------TAM---   0.05
    16--------SDG---  -0.45
    17--------STL---  -0.60
    18--------CHI---  -0.64
    19--------CLE---  -1.19
    20--------BUF---  -1.27
    21--------MIA---  -1.51
    22--------BAL---  -1.60
    23--------NYG---  -1.70
    24--------IND---  -1.85
    25--------OAK---  -1.90
    26--------TEN---  -2.47
    27--------MIN---  -2.54
    28--------HOU---  -2.74
    29--------WAS---  -3.83
    30--------NYJ---  -3.86
    31--------PIT---  -4.71
    32--------JAX---  -9.66

    [Edited by Chase. Not sure what went wrong, but you just need to wrap them in < p r e > tags, without the spaces.]

    • Arif Hasan November 14, 2013, 4:58 am

      Green Bay had quite the drop (predictably), as did Cincinnati, in part due to a game that needed (failed) late-game heroics as much as due to strength of schedule (a poorly ranked Baltimore team) and the fact that their 40 point victory over the Jets is a) getting washed out with more data and b) looking less impressive with time (sorry, Chase) than anything else.

      Seattle was able to rise above tough games in the later weeks of their season and put their heels on Atlanta’s throat, which allowed them to rise four ranks.

      The biggest changes are to Indianapolis (dropping 11 ranks!) and St. Louis (the opposite!). Interestingly, it feels like that game alone had a lot to do with their changes as St. Louis had more previous opponents lose ranks (Atlanta, San Francisco, Houston, Tennessee) than gain ranks (Seattle and Carolina). The Jacksonville win had to help, though.

      I find it fascinating that Dallas’ rank remain unchanged.

      • Chase Stuart November 14, 2013, 12:09 pm

        http://www.footballperspective.com/2013-game-scripts/

        I wouldn’t have guessed it, but the Cowboys have only had negative Game Scripts twice this year (granted, the Chiefs/Broncos games were essentially ties). Even the Chargers game was very close, and you could argue that Dallas was four plays away from being 9-1. Not that you would, but after watching the Saints game, it would be reasonable to think the Cowboys had one of the worst defenses of all time and were a 2-8 type of team.

        • Shattenjager November 14, 2013, 1:53 pm

          I tried on my own several years ago to figure out how to discern the effects of a defensive coordinator using statistics, mostly DVOA. The only thing I could find that was at all consistent was that a great DC essentially never allows the defense to become horrible–their worst defenses were far from good but by no means were they near the bottom of the league. Monte Kiffin and Dick LeBeau are now coaching bottom-ten defenses.

          (Note: The research I did was not as interesting as it sounds and was lost a couple of computers ago. I would not be surprised if it did not even support my one firm conclusion. I just find it so strange that Kiffin and LeBeau are coaching such poor units.)

          • Chase Stuart November 14, 2013, 8:50 pm

            Yeah, that’s a tough one, although perhaps something we can look at in the offseason. Maybe consistently good ratings with lots of turnover in personnel?

            • Shattenjager November 14, 2013, 11:55 pm

              I tried to look mostly at teams who had little personnel change with more turnover at DC (Baltimore is the example that comes to mind) and teams who had the reverse (Philadelphia under Jim Johnson). Sample size was an issue but I couldn’t even find anything that I thought was a small sample fluke from comparing those two groups at the time. However, this was also when I was in undergrad, and I would have been quicker then to dismiss anything I saw that was not 100% clear than I would be now, since I’ve learned not to expect anything approaching scientific levels of significance out of sports statistics.

    • Chase Stuart November 14, 2013, 12:02 pm

      So I assume you are simply using the Game Script score instead of MOV, and then iterating in Excel using the process from that video Neil created?

      I dig that idea. I’m not 100% sure what the best use of it is, but it seems useful/interesting. I can’t quite figure out why the Cowboys rank so high, but I am on board with Carolina being a sleeper for the SB.

      • Arif Hasan November 14, 2013, 4:39 pm

        I’m sure I’ll figure out the coding soon. One thing I like about SBNation commenting is the preview function. Is that something you’d be able to do? I did try the “” tag, or at least I thought I did.

        Anyway, yes. It’s such a great process! It took me about 15 views to get it right, and then there’s a lot of troubleshooting I had to work out intuitively, as I don’t have Neil on-hand. The fact that things need to average 0 in order for the process to work is like 90% of where errors come from.

        I’m not sure what the use of it either is, although because I’m suspicious of “clutch” I think this is perhaps an even better way to determine team quality than point differential, although it likely falls significantly behind Burke’s efficiency numbers or FO’s DVOA. It’s REALLY aggressive about garbage time, which I like, although perhaps too aggressive. Still worth it.

        • Chase Stuart November 14, 2013, 8:57 pm

          Let me look into it. I just downloaded Live Comment Preview, but it doesn’t appear to be working (this may be because it’s not coded for the newest version of wordpress).

          I agree with your sentiments. I like to think of these systems as tools in a toolbox rather than the quest for the holy grail. To that end, I think Game Scripts offers something nice.

    • Chase Stuart November 16, 2013, 6:49 pm

      My fault. I did not realize that only admins had HTML privileges. Thanks to this site, it is now fixed. Let me know if you have any other issues.

  • Arif Hasan November 14, 2013, 4:34 am

    Pretty sure those tags didn’t work. Oh well. Passer IDs:

    Pass ID Rank Team PR ID
    1 DEN 134.64
    2 NOR 132.02
    3 DAL 131.39
    4 ATL 124.18
    5 KAN 115.57
    6 CLE 115.31
    7 NWE 111.52
    8 DET 111.20
    9 MIA 109.82
    10 GNB 109.34
    11 CIN 109.14
    12 CAR 106.26
    13 CHI 105.44
    14 SDG 101.86
    15 PIT 101.18
    16 BAL 100.95
    17 ARI 100.43
    18 IND 96.27
    19 STL 95.17
    20 NYG 94.16
    21 MIN 92.44
    22 SFO 91.97
    23 TAM 90.83
    24 PHI 89.40
    25 HOU 88.56
    26 SEA 87.58
    27 OAK 86.30
    28 TEN 86.19
    29 BUF 78.82
    30 WAS 75.19
    31 NYJ 63.62
    32 JAX 63.23

    • Shattenjager November 14, 2013, 1:45 pm

      I’m very happy the Broncos remain on top of this list. Perhaps irrationally over-happy about it. :)

      • Chase Stuart November 14, 2013, 9:03 pm

        Enjoy this. I suppose I’m do for an update on this, so maybe I’ll do a full post on pass identity next week, which would also give you a break!

        I do find the Jaguars so low on the list kind of interesting. Some of that is probably overstated, as in blowouts teams just stop passing because the game is over. The Jaguars were at 64-67% in the four games of -15.5 or worse, but they passed more often in two other games. The Jets, though, are definitely super run heavy, which is kind of crazy considering the presence of Mornhinweg. The Jets are 28th in pass attempts and 5th despite having a horrible Game Script.

  • Arif Hasan November 14, 2013, 4:46 am

    Passer ID against

    oPR ID Rk Team PR ID
    1 NYJ 136.76
    2 ARI 125.59
    3 MIN 119.45
    4 DET 119.33
    5 WAS 119.25
    6 NYG 115.70
    7 PHI 112.72
    8 JAX 112.21
    9 CLE 112.07
    10 TAM 107.60
    11 DAL 106.48
    12 BAL 104.61
    13 SDG 102.38
    14 OAK 100.64
    15 BUF 99.87
    16 MIA 99.61
    17 KAN 97.78
    18 CIN 96.95
    19 ATL 94.89
    20 STL 92.21
    21 CAR 91.93
    22 TEN 91.23
    23 DEN 90.17
    24 PIT 88.44
    25 IND 88.13
    26 NOR 85.44
    27 HOU 85.18
    28 GNB 84.55
    29 NWE 82.88
    30 CHI 82.24
    31 SEA 80.76
    32 SFO 72.94

    I mentioned on my website why I think some of the “Passer ID against” scores look odd. The first has more to do with the opponent’s identity than the character of the defenses, I think (coaches prefer to play to their strengths rather than their opponents’ weaknesses) and the other has to do with how coaches respond to extreme situations (low/high game scripts). The Jaguars have a low passer ID in part because their game script is so consistently low that ANY number of rushing attempts starts looking odd. They still rank 6th in pass/rush ratio, passing more than over 2/3rds of the league. But the fact that they run at all in the second quarter of most games…

    San Francisco, incidentally has played against the Colts (18th) Rams (19th), Seahawks (26th), Titans (28th) and Jaguars (32nd) but haven’t played anyone in the top ten of Passer ID and have only played Carolina (12th) of teams in the top sixteen.

    One last note: teams with “elite” quarterbacks are getting run on. This probably has something to do with “keeping good QBs off the field” (which is something I find to be stupid, but whatever).

    I find it interesting that Arizona is so often passed against given the reputations of Patrick Peterson and Tyrann Mathieu, but I suppose that front seven is as intimidating as anyone’s.

    • Chase Stuart November 14, 2013, 9:11 pm

      Very good stuff.

  • Arif Hasan November 14, 2013, 5:16 am

    I’ve been applying SRS adjustments to anything I can after doing it constantly and what I just typed inspired me to do it for Passer IDs. This isn’t “differential Passer ID” just Passer ID once you take into account the average Passer ID their opponents have faced. So if a team decided to pass a lot against a team with an “elite” QB, they would get a massive boost.

    These could “feel” more right given what I know about the team’s identities, but I’m sure I screwed up, given the Jaguars’ score. In fact, a lot of numbers seem very extreme. San Francisco is a passing team because of this (not many people throw against Seattle, Tennessee, Indianapolis or St. Louis, so they weren’t “below average enough”).

    The more I think about this, the more I feel like I did it wrong. Seattle seems ranked too highly as well.

    Rank Tm Score
    1 DEN 154.27
    2 CAR 145.65
    3 NOR 140.85
    4 KAN 139.91
    5 SFO 139.41
    6 SEA 136.91
    7 NWE 124.86
    8 GNB 121.94
    9 CIN 118.14
    10 DAL 111.85
    11 DET 106.61
    12 SDG 106.16
    13 PHI 105.02
    14 CHI 101.25
    15 IND 100.97
    16 TEN 95.87
    17 BAL 95.04
    18 STL 94.38
    19 ARI 92.65
    20 CLE 90.01
    21 MIA 89.81
    22 OAK 85.43
    23 BUF 81.24
    24 ATL 80.93
    25 TAM 79.87
    26 PIT 78.28
    27 MIN 76.42
    28 NYJ 70.88
    29 NYG 70.03
    30 WAS 69.95
    31 HOU 69.59
    32 JAX 20.55

    Naturally the process produces and SOS score, which if I understand the process correctly is the exact same as a “Passer ID against” score using the same process.

    oRank Tm oScore
    1 DAL 127.14
    2 SFO 122.83
    3 ARI 120.43
    4 TEN 119.96
    5 NYJ 119.71
    6 MIN 118.46
    7 CLE 112.43
    8 SDG 110.97
    9 NWE 109.07
    10 CIN 105.92
    11 DET 103.30
    12 PIT 103.08
    13 ATL 102.51
    14 SEA 100.62
    15 CHI 100.48
    16 HOU 97.94
    17 NOR 97.52
    18 NYG 97.19
    19 DEN 96.32
    20 GNB 95.01
    21 BAL 94.90
    22 BUF 92.71
    23 KAN 90.62
    24 CAR 90.47
    25 WAS 89.93
    26 TAM 89.14
    27 STL 88.21
    28 OAK 85.44
    29 MIA 85.01
    30 JAX 84.97
    31 PHI 84.79
    32 IND 60.28

    Which makes a lot of my analysis above invalid, I guess. At least it seems less extreme (which supports the theory—if I didn’t screw everything up, that is—that coaches would much rather fall on their own team’s perceived strengths than play to their opponents’ perceived weaknesses) but it seems odd to me that teams seem more concerned keeping Andrew Luck off the field than Manning, Brady, Rodgers or Brees. The Colts’ pass defense is 27th in net yards per attempt allowed while their run defense is 23rd in yards per attempt allowed.

    Oh well. This should conclude my Weekly Game Script spam.

  • Chase Stuart November 14, 2013, 10:35 am

    I very much enjoy the Game Script spam. Let me chew on some of this.

  • Chase Stuart November 14, 2013, 12:03 pm

    I find it interesting that NO and BAL had a difference of only two points in Game Scripts, but 29 in real life. Note that for Baltimore, this includes OT, so their regulation Game Script would have been even closer.

  • Richie November 14, 2013, 7:09 pm

    Since we’re data dumping our random analyses…

    Here’s the week 10 win probability added for QBs in the 4th quarter/overtime:

    Andy Dalton was the “clutchest” of the week, but that’s mainly due to the lucky hail mary to AJ Green. He actually had negative WPA aside from that play. At .485 WPA it was the most effective drive of the week. Next best was Matt Stafford (who is crushing the 4th quarter this year) with a .337 touchdown drive just before the 2 minute warning.

    The worst of the week was Colin Kaepernick with a horrendous -0.819 WPA. He led three “drives” in the 4th quarter. Two ending in punts and one interception. The 49ers began their final drive at 1:02 with a 0.378 probability of winning the game. But, whether it was his fault or not, Kaepernick couldn’t do anything on offense.

    Ryan Fitzpatrick’s sack/fumble with 2:39 to play was the single worst drive of the week. The Titans had a .551 probability of winning the game when they began that drive with 3:31 to play. But his sack dropped their chances to .005.

    Here’s everybody:
    Dalton 0.315
    Cassel 0.191
    Flacco 0.130
    Glennon 0.125
    Stafford 0.124
    Manning, E 0.115
    Palmer 0.113
    McCown 0.104
    Manning, P 0.017
    Roethlisberger 0.006
    Luck 0.000
    Romo 0.000
    Wilson 0.000
    Ryan 0.000
    Brees 0.000
    Clemens 0.000
    Manuel 0.000
    Hasselbeck 0.000
    Rivers -0.028
    Foles -0.036
    Tolzein -0.046
    Griffin -0.123
    Keenum -0.128
    Tannehill -0.252
    Newton -0.285
    Pryor -0.287
    Fitzpatrick -0.450
    Cutler -0.462
    Henne -0.543
    Kaepernick -0.819

    For the year, Stafford has been worth 1.347 WPA in the 4th quarter/overtime. (I’m still working on getting full-season scores for everybody else.)

Leave a Comment