• OL – Starting offensive line played all 71 offensive snaps: Armstead, Peat, Unger, Warford, and Ramczyk all with 71 snaps each.
  • QB – Brees 71
  • RB – Kamara 46, Ingram 28, Line 10
  • WR – Thomas 56, Ginn 53, Coleman 52, Snead 14, Lewis 3
  • TE – Fleener 36, Hill 31, Hooman 24

Notes: Fleener with the most TE snaps, first time that has happened in a while.  Snead still with only 14 snaps.


  • DE – Jordan 69, Okafor 50, Hendrickson 28, Kikaha 1.
  • DT – Rankins 55, Davison 48, Onyemata 29, Hughes 5
  • FS – Bell 69, M. Williams 64
  • SS – Bush 21
  • CB – Crawley 67, PJ Williams 63, Harris 36, Lattimore 6
  • LB – Robertson 69, Te’o 49, Mauti 30

Notes: Kikaha with just 1 snap, Hendrickson with 28.  Expect those numbers to both increase with Okafor out for the season (achilles).  I’d like to see Muhammad, who had a great preseason, get his chance, too.

With Lattimore out, PJ and Crawley got the bulk of the snaps at CB.  It’s unclear when Lattimore (low ankle sprain) will return to the lineup, but obviously the Saints desperately need him out there.  Based on a tweet he made, I think Kenny Vaccaro will be back against the Rams, but that’s just a hunch.

Special Teams

Mauti 28, Harris 25, Hardee 22, Edmunds 22, Banjo 22, Bell 18, Bush 18, Hodges 16, Hill 16, Line 14, Lewis 14, Lutz 13, M Williams 12, Hooman 12, those are the ones in double figure snaps.

Snap counts data via PFREF.


The following table was put together based on stats from PFREF.

yards / play 11 18 1 21 13
Scoring % 13 9 4 26 26
over %
22 9 9 11 15
passer rating 6 7 4 17 12
Sack % 19 9 1 15 24
rush yds / att 24 29 4 17 8
rush yds / gm 23 15 3 13 18
punt yds / ret 31 27 30 22 -4
kick yds / ret 21 28 28 6 -29
points / drive 11 7 3 22 23

The numbers are the offensive, defensive (and special teams, where applicable) rankings (lower the number the higher the ranking).   For example, the top row is for yards / play where the Washington offense is ranked 11th, the Saints defense 18th and so on.  The NO NET ADV. column is found by adding the New Orleans offense and defense ranking, and then subtracting from that the Washington offense and defense ranking, and finally multiplying the result by -1.  Confused yet?  Good.

Here is the NO NET ADV. column in chart form.  Further the bar is to the right, the bigger the advantage to the Saints, further to the left, bigger advantage to the Redskins.  Only advantage favorable to Washington is on special teams kick/punt return/coverage units.

nfl was v no net adv

Sack % refers to percentage of time quarterback gets sacked while attempting to pass.  Turnover % refers to percentage of drives ending in a turnover (whether INT or fumble doesn’t matter).  Scoring % refers to percentage of drives ending in some type of score (whether FG or TD doesn’t matter).  For special teams stats, offense refers to the return team, while defense refers to the coverage team.

Here is another handy dandy chart for benefit of those not yet confuzzled.  These are the rankings, color-coded.  Shorter the bar, the better the ranking.

nfl was v no rank comp 2017

Saints *should* win this game, based on this statistical comparison.  But the Saints can’t afford to overlook Washington, which is a dangerous opponent, and look too far ahead with a big showdown coming up against the no-longer-hapless Rams.  I don’t expect this team to overlook anybody, not coming off 3x 7-9 seasons, and not starting off 0-2, but I’m sure that will be the narrative if the Redskins pull off the upset.

Jason La Confora is reporting the Saints have added special teams coach Mike Westhoff to the coaching staff.


This is a good move for the Saints because they’ve been struggling mightily on special teams.  Not so much in the kicking of field goal and extra points (need help there, too), but in the punting and kicking coverage units as well as the punting and kicking return units.

According to PFREF, the Saints are 28th in yards per kick return, 30th in yards per punt return, 28th in yards per kick return allowed, and 27th in yards per punt return allowed.  That’s not getting it done, not for a team with playoff / Superbowl aspirations.  The winning streak has masked some of the deficiencies on special teams, but not from Sean Payton.  Props to Coach Payton for recognizing the weakness and taking a pro-active step to correct it.

We’re 9 games into the season, and Drew Brees has 2398 passing yards, putting him on pace for his lowest output (4263) as a Saint.  This is a good thing, not a bad thing.  Read on to find out why.

We’re accustomed to seeing Brees lead the league in passing yards every year.  Brees led the league in passing yards in 2016, 2015, 2014, 2012, 2011, 2008, and 2006.  That’s 7 out of the last 11 seasons and each of the last 3 (and 5 of the last 6).  In case it got past you, the Saints made the playoffs in 2013, 2010, and 2009, all 3 of the most recent seasons in which Brees did *not* lead the league in passing yards.

This year Brees is at 2398 yards in 9 games, which puts him at about 266 per game.  The Saints are also running the football more effectively and more efficiently.  Saints are now averaging 142 yards per game rushing.  When the Saints pass for between 250 and 300 yards *and* rush for 100+ yards their record is 22-2-0 (.917).  This is winning football.  This is what you want.

You might say, okay, if this is so effective, why not do it every game?  It’s easier said than done.  When you fall behind on the scoreboard you have to try to catch up, and throwing the ball gives you a better chance to catch up than running the ball does at that point.  Yeah, you could be down 21 points and insist on getting 100+ yards rushing and probably get it, but it won’t help you win the game.  It’s no good to get the 100+ yards rushing and still lose by 14.

You need 2 things going for you in order to be able to hit that sweet spot (250-300 yards passing and 100+ yards rushing): 1) you need to play good defense so you don’t fall behind too far on the scoreboard, and 2) you need to be having success when you do run the ball.  There’s no point stubbornly running into a brick wall when it isn’t working (but that’s not to say you should just give up if the first few runs aren’t successful, either).

The difference this year with this team is they’re playing much better on defense.  Let’s face reality, the last few years the defense has been terrible.  Better defense means you’re not finding yourself down big on the scoreboard, which means you can stick to the game plan of maintaining balance on offense.  Saints are also having success running the ball.  The yards per attempt (4.6) is 4th best in the league (last game helped that average, but even before that it was pretty solid).  They’re getting stops, they’re running for a good average, they’re winning games, and the swagger is back in the WhoDat nation.


Snap counts courtesy of PFREF.


For the first time this year Brees didn’t take every offensive snap.  Brees had 72 of the 77 offensive snaps.  Daniel had the other 5 in mop up duty.

Peat, Unger, and Ramczyk had all 77 snaps.  Warford and Armstead, who have missed time with injuries, had 59 each.  The fact both had exactly 59 snaps is a good indication that they came out for rest rather than because of injury setbacks.  LeRibeus had 22 and Kelemete had 18 in relief.

RB’s: Kamara 29, Ingram 37, Edmunds 13, Zach Line 19 at fullback.

WR’s: Thomas 56, Ginn 50, Snead 36 (up from 18 last week, if memory serves), Coleman 29, Lewis 5.

TE’s: Hooman 47, Hill 43, Fleener 17.  I believe Fleener has earned more snaps on offense, and he might have gotten them if the Saints had not gone into running mode in the 2nd half.


There weren’t many defensive snaps to go around this week, and nobody played all 48 snaps.  Cam Jordan joked about running gassers after the game because he didn’t feel like there were enough defensive snaps in the game.

DT’s:  Rankins 28, Onyemata 26, Davison 20, Hughes 17.

DE’s: Okafor 41, Hendrickson 32, Jordan 31.

LB’s: Te’o 40, Robertson 35, Hodges 11, Klein 9 (Klein left the game with an injury, which is why Te’o got so many snaps, filling in admirably.), Mauti 4.

CB’s: Crawley 47, Lattimore 37, PJ Williams 34, Harris 11.

SS: Bush 21, Banjo 11

FS: Bell 37, Marcus Williams 36.

Special Teams

I won’t go into special teams snaps too deeply.  Morstead joked that he put himself on the inactive list.  (Saints didn’t punt the entire game, not even once.)  But in reality Morstead did have 8 snaps, serving as holder for FG’s and XP’s.

Oh, and 1 ST snap for the naked, drunk guy, but I’m not sure if he should be listed on the Bills or Saints.  Good times.


All stats courtesy of PFREF.

Stat Offense Offense Rank Defense Defense Rank
Total yds/play 6.1 3 5.3 18
Passer Rating 105 3 79.2 7
Adj Net Yds/Att 7.9 4 5.1 6
INT % 1.50% 6 3.30% 6
Sack % 2.80% 1 7.90% 24
Rush yds/att 4.3 7 4.7 29
Rush yds/game 122.7 7 116.1 19
Punt yds/return 5.6 27 11.4 27
Kickoff yds/ret 19.4 26 27.1 30
Points/drive 2.27 6 1.66 10
Drive Score % 43.00% 6 32.20% 11
Turnover % 9.30% 8 13.80% 9

nfl halfway point rankings in 2017

Offense and defense generally playing very, very well for the Saints.  Offense is top 10 in every stat covered here and top 5 in 4 of them.  Defense top 10 in 5 of the stats and very close in another (drive score %).  Biggest struggles for offense: none, really, maybe turnover % a bit higher than we’d like, but still top 10 at 8th best.  Defense’s biggest struggles: rushing yards / attempt allowed (29th) and sack % (24th).  Also below average in rushing yards / game allowed (19th) and total yards / play allowed (18th).  (Keep in mind these numbers all include the first 2 games in which the defense was abysmal.)

Special teams have been terrible in both kick and punt returns and coverages.  Also are getting kicks blocked and muffing punts.  On the bright side, Saints did make a huge play on a blocked punt returned for a TD, thanks to Justin Hardee.

Failed to mention in episode 1B that all data courtesy of PFREF.  Go back and read episode 1 and episode 1B, or don’t, it’s up to you, but you probably should if you haven’t already.

The question at hand is whether the myth that we can “throw out the records in division games” because you just never know what’s going to happen in those rivalry games is confirmed, plausible, or the dreaded busted.

In this (probably last) look at this myth, we’ll be looking at win/loss records of the good teams, the bad teams, and the ugly teams (no teams are ugly, but how can you have good and bad without including ugly?)  In particular, we should expect the best teams’ records to be not quite as good in division games as in non-division games.  If the myth is completely accurate, we should expect the best teams to only be .500 inside the division even if they win all the non-division games and the worst teams to be .500 inside the division even if they lose all the non-division games.  In other words, the best teams should be not as good inside the division as outside the division and the worst teams should be better inside the division than outside the division.


Best 5 records Non-division division difference
New England Patriots 0.767 0.778 0.011
Indianapolis Colts 0.62 0.744 0.124
Pittsburgh Steelers 0.61 0.689 0.079
Denver Broncos 0.6 0.611 0.011
Green Bay Packers 0.58 0.706 0.126
Worst 5 records Non-division division difference
Jacksonville Jaguars 0.393 0.378 -0.015
Detroit Lions 0.373 0.322 -0.051
Cleveland Browns 0.367 0.233 -0.134
Oakland Raiders 0.367 0.344 -0.023
St. Louis Rams / Los Angeles Rams 0.36 0.372 0.012

The above table contains the winning percentages for the teams with the top 5 best non-division records and the worst non-division records, plus those same teams’ records inside their own divisions for comparison.  We should expect, if the myth is true, all these best teams should have poorer records in division games and all these worst teams should have better records in division games, but we have is the exact opposite.  Only one of these 10 teams has a division record versus non-division record that would tend to support the myth, and it’s only a .012 difference (the formerly hapless Rams).

All the good teams have better division records than non-division records and all the bad teams (except the Rams) have worse division records than non-division records.  This myth is so busted.  Here’s a pretty chart:

nfl best worst comparison division non-division 2002-2016

The blue bars are the non-division winning percentages, the orange bars the division winning percentages, and the yellow bars (the real key part) are the differences.  If the yellow bar is in the positive above 0 that means the division records were better than the non-division records, while below 0 means the opposite.  Yellow bars should all be below 0 for the good teams on the left and all positive for the bad teams on the right if the myth were true.

Conclusion: BUSTED (yet again)

I recently posted episode 1 in this Mythbusting series, and maybe this should be episode 2, but I’m calling it episode 1B because it’s just revisiting the same question, but from another angle.  Go back and read episode 1 if you haven’t already.  No, really, go back now and read if if you haven’t already or else the rest of this won’t make as much sense.

It occurs to me that maybe the oddsmakers factor in the division game bias when setting the odds.  In other words, maybe Vegas realizes division games are always closer, so they adjust the spread accordingly, and thus using betting lines and upsets, maybe that’s not the best way to go as far as testing the myth.  (Or maybe it is.)

I’ve come up with another angle on this, another way to try to put it to the test, and that’s by looking at margins of victory in division versus non-division games.  (Note: in episode 1 I was comparing division games versus all games, whereas here I’ll be using division games versus non-division games, which is what I should have done before.)  The idea here is the superior team is more likely to blowout a non-division opponent if the myth is true.  There should be more close games inside the division than outside, all other things being equal, and presuming the superior team always wins.

The following table compares the margins of victory in division games versus non-division games from 2002-2016, inclusive.  (Note: 2002 is when the Texans came into the league and the divisions were separated into 8 divisions, 4 teams each.)

Victory Margins Non-division Non-division % Division Division %
7 or more 1531 31.90% 892 30.97%
10 or more 1175 24.48% 692 24.03%
14 or more 876 18.25% 494 17.15%
21 or more 457 9.52% 259 8.99%
28 or more 199 4.15% 114 3.96%
6 or less 865 18.02% 545 18.92%
3 or less 542 11.29% 326 11.32%

Victory Margins = the number of points by which the winner of the game won, doesn’t matter if the winner was an upset winner or not.

Non-division % and Division % = percentage of games out of total opportunities for the teams to win by that victory margin.  It’s not the percentage of games that were won by this victory margin out of the number of games played since there’s no way both teams could, for example, win by 21.  In other words, if there was only 1 game played and the winner won by 21 points, then the % would be 50% because 2 teams had a chance to win by 21, but only 1 of them did it.  Multiply the % by 2 if you want to know the percentage of games played that ended with that victory margin.  Have I confused you enough yet?  Good.

Maybe the following chart will make things clear as mud, nothing like a good picture:

nfl victory margins 2002-2016

The orange bars represent the non-division % and the green bars the division %.  The horizontal axis is the margin of victory.  The vertical axis is the %.  One interesting side note is there are more games that end in 7+ point victories than in 6- point victories, and it’s not that close.  In fact, there are roughly as many 14+ point wins as there are wins where the winner won by 6 or fewer points.  But keep in mind 7+ includes all the 10+, 14+, 21+, and 28+ wins, too, whereas the 6 or fewer games include only those games that were won by 6, 5, 4, 3, 2, or 1 point.

The division percentages are based on 32 teams playing 6 division games each for 15 years, so that’s 32 * 6 * 15 = 2880 division game opportunities for all teams combined.  The non-division percentages are based on 32 teams playing 10 non-division games each for 15 years (2002-2016  represents 15 seasons, I counted them on my fingers).  So, there were 32 * 10 * 15 = 4800 non-division game opportunities for all teams combined.   We get the percentages by simply dividing the number of games matching that criteria by the total opportunities (e.g. 7+ point wins in non-division games = 1531 / 4800 = about 31.90%).  Probably should have used 1440 and 2400, but I’m too lazy to back and do it all again, and the relative numbers are still the same either way, just not quite as close to each other.

Notice how the orange bar is always taller for the big wins while the green bar is always taller for the close wins (6 or less and 3 or less).  It’s really, really close in the 3 or less columns, but it is a tiny bit less (11.29% versus 11.32%).  (Note to the grammar police, yeah, I know it should probably be 3 or fewer instead of 3 or less.)

So, what does it mean?  Is the myth that we can throw the records out in division games confirmed?  No, I’m not going that far, but it is interesting that even though the percentages are all very close, they *all* support the conclusion that division games *are* more likely to be closer games than non-division games.  That’s 7 different data points that all agree even though they’re all very close.  Even in the biggest category difference, 14+ point wins, the difference was only 1.1% (18.25% non-division versus 17.15% division).

I’m still going to call this one busted.  The data is there, make of it what you will if you disagree.  The numbers are all still very, very close, all within 1% or 2%.  If the myth that anything can happen in division games was true we should expect really big differences, way more close games and way fewer blowouts, which just isn’t the case.

Conclusion: BUSTED

Snap counts courtesy of @pfref.

There were 60 offensive snaps.

On the offensive line these players took all 60 snaps:  Peat, Kelemete, Ramczyk, and Unger.  Armstead had 24 snaps before leaving with an injury (chest).  LeRibeus came in and played 37 snaps.

Running backs: Kamara 25, Ingram 36.

WR’s: Thomas 49, Ginn 37, Coleman 30, Snead 18, Lewis 0.

TE’s: Hooman 38, Hill 28, Fleener 22.

On defense there were 59 snaps.

DE: Jordan 48, Okafor 44, Hendrickson 35, Kikaha 9.

DT: Rankins 45, Davison 28, Onyemata 28.

LB: Klein 59, Robertson 52, Te’o 16.

CB: Lattimore 59, Crawley 58, PJ Williams 23.

Safety: Marcus Williams 59, Bell 49, Vaccaro 28, Banjo 9.

These players had double-digit ST snaps: Mauti 19, Banjo 19, Edmunds 16, Lasco 16, Harris 15, Hardee 15, Bell 14, PJ Williams 12, Lutz 11, Hodges 10.

The turnaround on defense, in my view, came when they put Crawley into the starting lineup in place of Harris and PJ Williams.  Saints have been basically playing 3 safeties when they go to nickel, but when Vaccaro came out with an injury, PJ Williams got a few defensive snaps (and played well), along with Banjo getting a few snaps at safety (his retaliatory hit on Evans earns him a gold star regardless of what he did on the other plays).

Ingram didn’t start, as I recall, but he did actually end up playing more snaps than Kamara.  Both players had solid games, but Kamara’s was marred by a fumble.  (Maybe that’s why Ingram got more snaps.)

In this first episode (and possibly only ever episode) of this series called Mythbusting we will be looking at this myth, and putting it to the test:

“You can throw out the records in division games.”

In other words, upsets are more likely in these division rivalry games.  It makes sense on the face of it because these teams know each other very well, having played each other twice a year every year, and sometimes three times a year if they happen to meet in the playoffs.  But is it true?  And, also central to this undertaking, how can we test it?

PFREF has an excellent database of NFL stats, including what the point spread was for each game.  We can use that information to see how often the favorite wins the game in division games and in non-division games.

Using the team game finder tool, going back to 1970, and filtering the search to only include games where the team is the favorite (point spread < 0) the results include 9213 games in all.  The favorites were 6067-3126-20 (.660) in those games (straight up).  In other words, they won 66% of the time.  These are all games, whether division games or not.

If we filter to include *only* division games, we get: 4011 games, with the favorites going 2631-1370-10 (.657).  So, in division games the favorite only wins 65.7% of the time while the favorite wins 66.0% of the time in all games (whether division or non-division).  But this is only a 0.3% difference, which I am going to call negligible.  We can’t expect these 2 percentages to be *exactly* the same.  In reality, there is no appreciable difference in upset rates in divisional versus non-divisional games.  The oddsmakers get it right about 66% of the time in either case.

I’m going to call this one: BUSTED.



%d bloggers like this: