Why Steelers, Cardinals, and Broncos Defenses Could Improve in 2021

Pittsburgh Steelers CB Steven Nelson
Pittsburgh Steelers CB Steven Nelson
Photo: USA Today Sports Images

The Steelers were the best pass defense in football in 2020, while the Cardinals were in the top 10 and the Broncos just missed. Each of those three teams has brought in plenty of new faces in the defensive backfield, but there's reason to believe that despite the changes, they'll be even better in 2021.

We have spent the past couple of weeks breaking down and analyzing offensive slot-wide splits, sorting through wide receivers, quarterbacks, and tight ends and running backs. Now, we can finish up by flipping the script and see how defenses coped with the ever-increasing amount of slot receivers putting up huge numbers.

This charting data once again comes from our friends at Sports Info Solutions. As a reminder, each player's position is based on where they lined up on the field, rather than relative to other wideouts. The outside receiver in a bunch formation is still considered a slot receiver as opposed to an outside receiver.

2020 Defense: Overall Slot v. Wide

Our first table looks at all targets each defense faced that were thrown to any player in the slot or out wide. We have listed each team's DVOA and number of passes against both slot and wide targets; the rate of passes to players in the slot as a share of passes to either slot or wide receivers (Slot%); and the difference in DVOA from wide to slot. As a reminder, negative DVOA means better defense.

Overall Defense, Slot vs. Wide, 2020
  Slot Wide        
Defense DVOA Rk Passes DVOA Rk Passes Slot% Rk Wide-Slot DVOA Dif Rk
PIT -16.6% 3 249 6.6% 22 151 62.3% 12 23.1% 1
ARI -5.9% 10 263 16.1% 29 146 64.3% 7 22.1% 2
DEN -12.0% 7 229 9.0% 26 152 60.1% 23 21.0% 3
SF -15.8% 4 232 4.7% 19 148 61.1% 17 20.5% 4
GB -20.4% 1 232 -0.2% 15 126 64.8% 3 20.2% 5
PHI 16.2% 28 214 30.3% 32 158 57.5% 31 14.1% 6
LV 13.5% 26 251 26.4% 31 131 65.7% 1 13.0% 7
CIN -3.1% 12 246 8.5% 25 157 61.0% 18 11.6% 8
KC -12.7% 6 226 -2.4% 14 122 64.9% 2 10.3% 9
DAL 11.9% 25 210 20.1% 30 145 59.2% 27 8.2% 10
LAC -6.9% 9 203 0.8% 16 141 59.0% 28 7.8% 11
NO -17.8% 2 255 -10.5% 9 141 64.4% 6 7.3% 12
CHI 1.0% 15 236 7.9% 23 136 63.4% 8 6.8% 13
BUF -14.1% 5 248 -8.0% 10 158 61.1% 16 6.0% 14
ATL 7.7% 23 252 11.8% 27 185 57.7% 30 4.1% 15
MIA -9.3% 8 236 -7.9% 11 147 61.6% 14 1.4% 16
JAX 7.4% 20 231 4.3% 18 139 62.4% 11 -3.1% 17
SEA 9.8% 24 302 4.8% 20 195 60.8% 19 -5.0% 18
DET 22.2% 31 255 15.7% 28 150 63.0% 9 -6.5% 19
CLE 3.9% 18 266 -6.1% 12 147 64.4% 5 -9.9% 20
CAR 7.4% 21 211 -4.7% 13 173 54.9% 32 -12.1% 21
TB -2.4% 14 239 -14.9% 5 164 59.3% 26 -12.5% 22
NYJ 18.3% 29 236 5.0% 21 161 59.4% 25 -13.4% 23
IND 2.0% 16 222 -13.2% 7 145 60.5% 20 -15.2% 24
NYG 18.9% 30 250 3.3% 17 136 64.8% 4 -15.6% 25
BAL -4.4% 11 274 -20.1% 3 171 61.6% 15 -15.7% 26
NE 2.1% 17 216 -14.4% 6 129 62.6% 10 -16.5% 27
MIN 7.5% 22 226 -11.2% 8 140 61.7% 13 -18.7% 28
WAS 6.0% 19 242 -15.1% 4 160 60.2% 22 -21.2% 29
HOU 33.1% 32 219 8.1% 24 154 58.7% 29 -25.0% 30
LAR -2.5% 13 225 -33.0% 1 150 60.0% 24 -30.4% 31
TEN 15.0% 27 278 -24.6% 2 183 60.3% 21 -39.6% 32

When reading this table, keep in mind that slot DVOA is significantly more consistent from year to year than wide DVOA. The correlation in slot DVOA from 2019 to 2020 was 0.35, compared to 0.21 for wide DVOA. And that 0.21 for wide DVOA is unusually high; it was at just 0.02 between 2018 and 2019. Teams that are good at defending the slot tend to remain so year after year, while teams good at shutting down receivers out wide tend to revert more towards the mean in future seasons.

There are a number of factors playing into that, though the most significant might be the sheer number of players involved. When a receiver splits out wide, defenses will generally match them up with one of their top two cornerbacks. SIS charting listed 71 different defenders as the primary coverage guy on at least 20 wide targets, or 2.2 per team. But a slot target? That might come from a boundary corner moving inside to follow a top receiver, or a nickel or dime corner against a shifty slot guy, or a safety or linebacker matched up against a tight end or running back. There were 111 different defenders covering at least 20 slot targets in 2020, or 3.5 per team. If 20 feels like too arbitrary a cutoff for you, there were 551 defenders with at least one target in the slot compared to 417 out wide. Interior coverage tends to have more moving parts than boundary coverage, so the effects of one guy having a good or bad season, or one injury or free-agent acquisition altering your starting lineup, tend to be less impactful inside.

So does that bode well for Pittsburgh, Arizona, and Denver, all of whom put up top-10 performances in the slot in 2020 but couldn't crack the top 20 out wide? If the pull of statistical regression is the most fundamental force in the universe, sure, but the fact that these three teams will have five new starting outside cornerbacks probably has more of an effect than any regression from 2020.

For the Steelers, their struggles out wide were less about the players involved than about first- and second-half splits. Over the first half of the season, they had a 22.3% DVOA against wide targets, third-worst in the league. That improved to a more respectable -8.2% over the last eight weeks, placing them firmly in the middle of the pack and mirroring the improvement of overall pass defense, which went from -14.5% to -25.4% as the year went along. There isn't an obvious explanation for this, no mid-season returns or sudden shifts in the schedule or anything; they simply gave up some big games to Darius Slayton, A.J. Brown, and the other wideouts they faced early in the year. More important than the splits, however, is the personnel. The Steelers' top-targeted outside corner was Steven Nelson, and he's gone, off to Philadelphia in a cap space move. Nelson ranked eighth in coverage success rate a year ago and allowed a perfectly adequate 6.6 yards per target; while he's not an irreplaceable piece, he does leave at least some moderately sized shoes for Cam Sutton to fill. And if Sutton does spend more time outside, that brings up questions for slot performance, as the two Steelers who spent the most time covering in the slot were Sutton and the departed Mike Hilton. We still have Pittsburgh projected with the top defense in 2021, but I think it's safe to say that there are questions in the secondary at the very least.

While the Steelers were shedding corners, the other two teams up there were adding them. Denver was on this list last year as well—strong in the slot, poor out wide—and we suggested that replacing Chris Harris with A.J. Bouye might bring success. Well, a shoulder injury, a concussion, and a suspension ended hopes there, and the duo of Michael Ojemudia and De'Vante Bausby ended up being less than an ideal pair of boundary corners in a division where you have to face Patrick Mahomes and Justin Herbert twice a year each. Phrased like that, drafting Patrick Surtain in the first round makes a ton of sense, as does bringing Kyle Fuller and Ronald Darby in in free agency.

Arizona, too, will have a new top pair of boundary corners, as Dre Kirkpatrick and Patrick Peterson were their top two defenders last season, and both struggled mightily. The Cardinals going old, with Malcolm Butler coming in in free agency and Robert Alford returning after missing two years with injury. That's … less than ideal, but it's hard to imagine them being much worse than the Kirkpatrick/Peterson duo was a year ago.

Butler leaves Tennessee, which had reverse splits from the teams we just talked about—near the top out wide, and near the bottom defending the slot. Butler ended up being the top-targeted Titans defender in both, so Tennessee's pass coverage will look different all over the field. In fact, the Titans will have to replace both of their most frequent slot defenders, with Desmond King off to Houston; it was something of a purge of a bland secondary this offseason. With the top three corners in snap count gone, it's safe to say that Tennessee's secondary is in flux. That's not the case in Los Angeles, where Jalen Ramsey and Darious Williams return after leading the Rams to the top DVOA both out wide and deep; if you wanted to have success passing against the Rams, you were limited to a box in the interior of the field. The Rams' top two defenders in the slot were Troy Hill and John Johnson; both are in Cleveland now as the Rams lost a lot of pieces from last year. They'll need young players such as Terrell Burgess and David Long to step up if they're going to maintain anything like last year's success.

And then there are the Houston Texans. They were a sieve out wide, ranking 24th with an 8.1% DVOA. And then they were a leaky sieve in the slot at 33.1%. And yet, the Texans are trotting out mostly the same secondary they had last season—Vernon Hargreaves, Eric Murray, Bradley Roby, Keion Crossen, and Lonnie Johnson are all back, with Philip Gaines being the only Texans defensive back with 20 slot/wide targets to not return in 2021. I think it's safe to say that the secondary wasn't the primary issue the Texans had last season, but it is odd, in this Houston era of "competition" as both philosophy and buzzword, to see so little change here.

2020 Defense vs. Wide Receivers

The next table features the same data as the previous table, but limited just to targets that went to wide receivers.

Defense vs. Wide Receivers, Slot vs. Wide, 2020
  Slot Wide        
Defense DVOA Rk Passes DVOA Rk Passes Slot% Rk Wide-Slot DVOA Dif Rk
DEN -17.2% 2 182 10.6% 25 147 55.3% 26 27.8% 1
ARI -5.4% 10 197 19.6% 29 131 60.1% 6 25.0% 2
PIT -14.3% 4 182 5.7% 22 137 57.1% 19 19.9% 3
SF -14.2% 5 184 5.3% 21 137 57.3% 18 19.4% 4
CIN -6.8% 9 180 9.7% 24 142 55.9% 23 16.5% 5
ATL -1.2% 17 193 14.6% 28 163 54.2% 27 15.9% 6
NO -14.5% 3 190 0.6% 16 127 59.9% 7 15.1% 7
BUF -18.2% 1 181 -4.4% 10 143 55.9% 25 13.8% 8
PHI 19.1% 30 157 29.9% 32 150 51.1% 31 10.8% 9
GB -12.7% 7 185 -2.4% 15 113 62.1% 3 10.2% 10
MIA -13.4% 6 189 -3.2% 14 136 58.2% 13 10.2% 11
CHI -4.9% 11 184 2.9% 17 120 60.5% 5 7.8% 12
LV 16.6% 27 198 24.1% 31 118 62.7% 2 7.5% 13
JAX -1.0% 18 186 4.1% 18 127 59.4% 10 5.1% 14
DAL 16.6% 28 170 19.7% 30 133 56.1% 21 3.0% 15
LAC -3.8% 13 160 -4.0% 13 126 55.9% 22 -0.2% 16
CAR -3.6% 14 156 -4.2% 11 158 49.7% 32 -0.5% 17
NYJ 14.6% 26 173 10.9% 26 150 53.6% 29 -3.7% 18
MIN 11.4% 24 176 4.6% 20 128 57.9% 15 -6.8% 19
KC -8.4% 8 181 -15.6% 6 106 63.1% 1 -7.2% 20
SEA 11.5% 25 233 4.2% 19 172 57.5% 16 -7.3% 21
DET 20.1% 31 207 12.0% 27 130 61.4% 4 -8.1% 22
CLE 2.1% 20 199 -7.4% 9 142 58.4% 11 -9.4% 23
NE 1.9% 19 167 -10.9% 8 114 59.4% 9 -12.8% 24
WAS -2.1% 15 186 -21.1% 5 134 58.1% 14 -19.1% 25
TB -2.0% 16 190 -23.2% 4 150 55.9% 24 -21.2% 26
NYG 17.7% 29 182 -4.1% 12 123 59.7% 8 -21.8% 27
IND 9.0% 22 184 -13.5% 7 136 57.5% 17 -22.4% 28
BAL -4.0% 12 192 -29.3% 2 147 56.6% 20 -25.3% 29
HOU 33.4% 32 169 8.0% 23 143 54.2% 28 -25.4% 30
TEN 9.2% 23 232 -27.2% 3 166 58.3% 12 -36.3% 31
LAR 6.9% 21 152 -33.0% 1 142 51.7% 30 -39.9% 32

As expected, this is very similar to the previous table. Wide receivers get the majority of pass targets, and that's doubly true out wide. 77% of slot targets and 91% of wide targets went to wide receivers, so for most teams, differences here were slight at best.

Not for the Panthers, however! Carolina rises from 21st to 14th in slot DVOA when you only look at wide receivers; they failed to match up properly against running backs and tight ends working out of the slot. An 11% rise in DVOA in this split is nuts—and most of it comes from removing the efforts of Shaq Thompson from the equation. As we'll see in upcoming tables, Carolina ranked in the bottom six against both running backs and tight ends in the slot, and Thompson was the primary victim—he was charted as the primary defender on 11 of the 55 targets, allowing a success rate of 73%, and I suspect a healthy portion of the nine targets without a primary defender listed were Thompson's responsibility as well. Take him out, and the Panthers look … "good" would be an overstatement, but more competent. In a division where you have to face each of Jared Cook, Rob Gronkowski, and Alvin Kamara twice a year, not to mention Kyle Pitts joining Arthur Smith in Atlanta, this seems like the kind of problem Carolina needs to solve.

The Rams had the opposite problem, falling from 13th to 21st if you only look at their performance against receivers in the slot. They got destroyed by the combination of Gabriel Davis and Cole Beasley against Buffalo (combined nine targets in the slot for 172 yards), and Deebo Samuel caught 12 of 13 targets for 159 yards in his two appearances against Los Angeles. Take out those opponents, and the Rams rise to a much more respectable -9.1% DVOA, which would have ranked seventh. Of course, chalking Buffalo up to one bad game is all well and good, as the Bills are in the opposite conference. Waving away two games a year against the 49ers is a slightly harder argument to make. As mentioned, the Rams have an entirely new interior coverage unit coming into play this year; they'll have a tall task in a division that can trot out Tyler Lockett, DK Metcalf, Brandon Aiyuk, and DeAndre Hopkins, all of whom had over 100 DYAR in the slot last season.

And then you have the Vikings, who went from eighth to 20th in wide DVOA, even though we're only taking 12 passes out of the equation! Against non-wideouts split wide, Minnesota managed two interceptions (both inside their own 25-yard line), four incompletions, one completion for negative yardage, four failed completions for positive yardage, and a touchdown in garbage time against Detroit. That's a DVOA of -177.8%, and even over only 12 attempts, that adds up.

2020 Defense Versus Running Backs

These last two tables are here mostly for completeness. The average defense faced 8.3 pass attempts all season against running backs in the slot and just 7.8 split out wide. Nearly 20% of all running back slot/wide targets went to J.D. McKissic, Nyheim Hines, or Chase Edmonds. We have ranked defenses by DVOA here for you, but this is small sample size theatre at its finest.

Defense vs. Running Backs, Slot vs. Wide, 2020
  Slot Wide
Defense DVOA Rk Passes DVOA Rk Passes
PHI -62.9% 1 13 81.6% 25 7
HOU -61.7% 2 7 -50.1% 8 5
BAL -39.5% 3 14 55.1% 19 19
JAX -35.5% 4 6 16.8% 16 9
MIN -33.1% 5 11 -102.3% 3 6
CIN -28.0% 6 6 81.3% 24 7
TB -26.8% 7 6 -10.6% 13 4
SF -26.1% 8 14 92.7% 28 4
CLE -5.0% 9 8 82.8% 26 3
LAC -3.6% 10 5 -73.4% 5 6
KC 0.1% 11 7 113.8% 31 11
LAR 1.6% 12 9 -30.4% 10 7
MIA 6.4% 13 7 37.5% 18 4
PIT 12.9% 14 11 12.0% 14 11
CHI 15.4% 15 9 66.8% 22 7
NO 15.9% 16 9 -55.1% 7 5
NE 20.6% 17 10 -42.8% 9 9
TEN 26.1% 18 5 -19.1% 11 12
DEN 31.6% 19 9 -74.3% 4 4
WAS 31.9% 20 6 56.0% 20 15
ARI 32.6% 21 10 85.3% 27 5
BUF 33.0% 22 7 -134.7% 2 8
NYJ 44.4% 23 6 -228.6% 1 4
GB 47.1% 24 4 32.6% 17 7
NYG 52.1% 25 15 104.9% 30 8
SEA 55.0% 26 15 61.1% 21 13
ATL 62.0% 27 4 -72.1% 6 9
IND 63.5% 28 4 102.1% 29 4
DET 65.7% 29 13 78.3% 23 14
LV 80.1% 30 5 168.2% 32 6
DAL 90.9% 31 3 -13.1% 12 6
CAR 133.4% 32 6 15.3% 15 11

Small sample sizes produce disproportionate results, but Detroit putting up a 65.7% slot DVOA on 13 targets—or more than twice as many as anyone who ranked below them—is frighteningly impressive in its way. It's even more impressive when you realize that six of those 13 targets went to McKissic, who had a below-average DVOA against the Lions. Take McKissic's struggles out and Detroit's DVOA climbs to 120.8% as Alvin Kamara, Ameer Abdullah, Aaron Jones, and Chase Edmonds took turns embarrassing the Lions; every running back not named McKissic who received a slot target against the Lions caught the ball.

2020 Defense Versus Tight Ends

Tight ends split out wide are essentially not a thing; the clue is in the name. Only five tight ends—Darren Waller, Gerald Everett, Eric Ebron, Jimmy Graham, and Travis Kelce—hit double-digit targets split out wide in 2020. The slot column is a little more meaningful; it's still small sample sizes, but every defense had at least 30 targets to go on.

Defense vs. Tight Ends, Slot vs. Wide, 2020
  Slot Wide
Defense DVOA Rk Passes DVOA Rk Passes
GB -60.7% 1 43 3.3% 19 6
IND -40.1% 2 34 -75.1% 5 5
KC -37.2% 3 38 93.8% 31 5
NO -34.3% 4 56 -159.4% 2 9
PIT -29.3% 5 56 33.5% 23 3
LAR -26.1% 6 64 -48.6% 12 1
SF -21.9% 7 34 -61.8% 8 7
LAC -21.3% 8 38 99.8% 32 9
DAL -20.3% 9 37 62.3% 28 6
ARI -14.5% 10 56 -52.6% 10 10
SEA -6.7% 11 54 -52.1% 11 10
BUF -5.3% 12 60 44.6% 24 7
LV -4.6% 13 48 -24.0% 13 7
TB -1.2% 14 43 85.1% 30 10
NE -1.2% 15 39 -71.8% 7 5
MIN -0.3% 16 39 -239.4% 1 6
BAL 2.0% 17 68 60.2% 27 5
DEN 6.2% 18 38 27.1% 21 1
MIA 7.1% 19 40 -146.8% 3 7
CIN 10.3% 20 60 -73.8% 6 8
CLE 11.5% 21 59 -19.0% 14 2
NYG 15.6% 22 53 54.9% 26 5
DET 20.7% 23 35 -17.3% 16 6
PHI 22.4% 24 44 -100.5% 4 1
CHI 26.1% 25 43 30.5% 22 9
NYJ 28.9% 26 57 -17.4% 15 7
CAR 32.5% 27 49 -53.8% 9 4
ATL 35.1% 28 55 19.7% 20 13
WAS 35.8% 29 50 -13.1% 18 11
HOU 44.9% 30 43 51.5% 25 6
TEN 52.2% 31 41 63.1% 29 5
JAX 53.8% 32 39 -13.2% 17 3

Hello, Green Bay! A -60.7% DVOA against tight ends in the slot is the best we have seen since we started collecting this data. Seven different tight ends had at least four targets in the slot against the Packers, and six ended up with negative DYAR—Jonnu Smith, Cole Kmet, Trey Burton, Tyler Eifert, Dallas Goedert, and T.J. Hockenson. Oddly, the only tight end to have any degree of success against the Pack was Zach Ertz, the worst tight end in football last season by DYAR—he picked up two first downs on Philadelphia's only two scoring drives against the Packers in December. Hey, in a small enough sample size, anyone can have a good day, even the worst tight end in the league against the top tight end defense in football.

Comments

18 comments, Last at 17 Aug 2021, 8:04pm

1 I don't really understand…

I don't really understand your point concerning the fact that more people cover the slot. That seems to line up with the fact that more passes are thrown to the slot than out wide, so this really seems like a logical fact anyway. If there are more passes to the slot, then it's no surprise that more defenders were assigned at least 20 times.

You probably would have to use different cutoff points for the different types of passes to make a better argument here. Does the point still stand when using 24 for slot and 16 for wide (going with the 60-40 split) as cutoff points?

4 True, outside is 99/100…

True, outside is 99/100 covered by a cornerback, the inside could be covered by 6 different guys.

Any Steelers fan would know that you can easily cover slot receivers with a run-stopping linebacker without running into any problems. 

14 More players Involved leads to greater year-on-year stability

Or, in other words, volatile performance by any one outside CB is much more likely to show up in team DVOA when looking at wide vs slot splits. A similar level of volatility by, say, a WILL LB will not have the same impact on team splits because he's not likely to have faced the same number/proportion of targets as an outside CB.

2 Worthless

The last chart is worthless.  Anything showing the Raider defense in the top half of any metric—other than blown tackles—is divorced from reality.

3 please use the following…

In reply to by Raiderfan

please use the following format for all complaints:

<Raider> is clearly ranked <too high> because <our defense sucks>. <Anything showing the Raider defense in the top half of any metric—other than blown tackles—is divorced from reality.>

6 Re: Zach Ertz vs GB

That is entirely believable, and parrots a bit from the Mark Twain story A Connecticut Yankee, where Twain points out that the best swordsman in the world has nothing to fear from the second best swordsman in the world.  Who he has to fear is the bozo who does not know what he is doing, because that guy will do things that don't have any bearing to fencing, and hence might trip up the best guy.  So GB really knew how to defend the TE, but they had no idea what it was that Ertz was doing last year.

7 I get your point, but if you…

I get your point, but if you or Twain honestly think that the #1 ranked swordsman would have a higher chance of losing to some random bozo than to the #2 ranked swordsman, well, I have a fence I'll let you paint, for a price... and I know a King and Duke who will give you half their fortunes, as long as you pay them a small fee to help them transfer them! 😉

8 Not So Farfetched

I've done martial arts for over 30 years, which besides making me old, has also taught me a bit. One thing I do know is that it can be very concerning to face novices in boxing, MMA, BJJ. Novices tend to freak out, flail, grab random digits, throw punches from ridiculous angles. And it only takes one lucky shot to put you down, followed by a lifetime of embarrassment when you get up.

In BJJ we always say that it's white belts you have to worry about because they will go full spaz. A few years ago I was passing some kid's guard. Not even sparring, just friendly drills. He panicked and jerked up off the mat while grabbing the back of my head. The result: his forehead crunched into my face, shattering my nose.

These days I approach the newbies very carefully.

 

9 I took fencing in college…

In reply to by RobotBoy

I took fencing in college and the instructor did in fact tell us that the best student in class had more to fear from the worst student in class than from the second-best student in class. 

10 I stand corrected! Thanks…

I stand corrected! Or rather, I lie on the floor corrected, with my nose busted by a roundhouse kick from a flailing n00b. Thanks for the perspectives, all.

I suppose this phenomenon exists for "one-hit KO" type sports. For more deliberate, protracted games, like golf or chess, not so much. I don't think a million spazzing monkeys in a million parallel universes in a million years would rack up a single victory over able-bodied Tiger Woods, or Magnus Carlsen... though they'd sure fling a lot of poop in the process.

11 It's more "amount of luck"…

It's more "amount of luck" rather than anything else. An amateur could often beat a pro on a golf course on a single hole. It's just the cumulative effects that change things. Not sure I'd go as high as a billion on golf, though. Billion's a big number.

Chess of course, but there it's even more obvious because there's no luck involved at all and in some sense there's "nothing new." However, the idea of "throw something at an expert they haven't seen before" does exist in chess (or at least, mostly did exist) - it's a common tactic at this point to throw wacko openings at simpler computer chess programs to get them out of their opening book. With more modern AI chess programs, obviously, this doesn't really work anymore. But it's the same idea.

 

12 Right- and I could beat…

Right- and I could beat Giannis in 1-on-1... if we played to 1, and I somehow managex to nab the ball and throw up a miracle jump shot!

It's hard to estimate odds when they get that remote... Tiger did shoot 81 in a major in 2002, when he was near his peak... but course conditions were comically bad (rain and gale-force winds), so your average 10-handicapper  might have struggled to shoot 121. The probability is definitely nonzero, though!

16 It's partly the "one-hit KO"…

It's partly the "one-hit KO" nature, but more about how it's difficult to counter an opponent when you have absolutely no idea what they're going to do, and maybe are going to try things they're not "supposed" to.

Here's an NFL-relevant analogy: Many years ago (as in, 15 years, maybe more), somebody went back and looked at Tom Brady's interceptions and found that on a lot of them, he had a wide open receiver somewhere else on the field. But none of those receivers were "supposed" to be open based on presnap alignment -- the defense had that receiver accounted for, then abandoned that coverage to focus somewhere else. 

In other words, the best way to intercept Brady was to do something so mind-blowingly stupid he would never see it coming, and hope that he followed through on his "proper" reads before realizing what a terrible mistake you had made.

17 Interesting! Though I'm…

Interesting! Though I'm guessing that gambit hit its expiration date right around September of 2007.

Could it be that throwing in occasional doses of randomness is actually an optimal strategy?

18 "an opponent when you have absolutely no idea"

This leads to my credence to my preference for just keeping James Morgan as Zach Wilsons backup instead of wasting new (guaranteed) money on Nick Mullens (I had this convo with mehlla, I believe it was, and was reminded of it because Mullens was terrible this past weekend). 

But I think the theory here runs out when quickly and relies essentially entirely on dumb luck which is a different type of difficulty. Sure the random can "try things they're not 'supposed' to." (I also question here if that random knows the rules and what they do is within them or not, if not, pretty easy no contest)

We see former (mind you) pro bball players wreck open gym full of randoms all the time for example. Like short Isaiah Thomas just dropped 81 at a pro am full of random people you've never heard of. Doubt IT was extensively studying them beforehand like the #1 fencer studying the #2 perhaps. 

Where #2 can beat #1 is not by doing dumb things necessarily but by switching things up as Rev implies. But generally, as I said, they just have different types of challenges. I'm not straight up picking Alabama to beat the Bucs over the Chiefs (or even Texans) under NFL rules. 

But I will pick James Morgan over Nick Mullens, not solely because of the randomness of his possibility to win (be good, beat a team, whatever) but because we live in a salary cap world. If the Bucs had all the time and money to focus on beating Bama, they would absolutely crush them anywhere. Bama still has a small chance but hardly one as good as the Chiefs (or even Texans) if we're being honest. 

13 The Tim Tebow experiment is…

The Tim Tebow experiment is starting to take on a whole new dimension for me.  Combine that with drafting an RB in the first round so he could play him as a WR and maybe the Meyers really is playing next-level chess.  I can't wait to see if he lines a DT up to take snaps in week one.

15 Those last two tables have some value

They tell me, for example, that the Chiefs are bad at covering RBs and TEs split wide, so I'd bet they likely faced these alignments mostly in high leverage situations. This probably ended up making the team's DVOA there even worse.

I'm sure that Chiefs opponents in 2021 are well aware of this weakness and will have some plays ready should the situation allow them. They'll face Darren Waller, Noah Fant, and Austin Ekeler twice each, so I believe we'll get to see whether Chiefs DC Steve Spagnuolo has figured out how to defend these plays.