Predicting 2021 Sack Totals

Many of you know by now, everything we do at FPS is rooted in finding data that can help predict outcomes, THAT THE MARKET DOES NOT account for already. Frankly, if you’re wagering any other way you’re simply counting on luck.

With that said, one of our favorite themes is to look at “matchup based variables“, like NFL QB vs Coverage (how well a QB performs vs a certain coverage, and how likely his opponent is going to play said coverage), to help understand likely performance, NOT captured by traditional market makers. We use these types of information to make a killing in both fantasy and individual prop bets alike. One such season long prop we love investing in is “Sack Totals”. As we’ve noted previously that Last Year’s Sacks don’t predict Next Year’s Sacks, but guess what books primarily look at when setting a player’s Sack line? You guessed it, last year’s sacks.

If you want to learn more about the “easiest prediction to make in football betting that the books have not caught up to yet”, I advise refreshing yourself with that article. However, the gist is this: Pressures, or the “first step in a chain of events” (Think exit velocity for a baseball hitter), is much more predictive, or even, consistent, year over year. Hence, we use that information to set our own sack total lines.

That info alone is extremely valuable, however we can couple it with the “Matchup Based Variable” to strengthen it even more: D-Line (DL) Strength of Schedule (SOS).

By now, we’re all familiar with SOS, it’s modest (and idiotic) beginnings when handicappers used last years win totals as a base to determine a team’s upcoming strength of schedule, to more specific uses like how many points a defense gives up to WRs when setting a fantasy lineup. We can have a separate conversation on the validity of “different SOSes”, but used wisely, across different applications, they help us predict outcomes. Hence, today we are going to couple proper SOS information with our “Pressure based Sack predictions” to help strengthen our “sack total prediction” for 2021.

Here’s how we did it, using our friends at PFF’s final 2020 OL rankings, along with their 2021 OL projections, we can see how a defense, and better yet, a D-Line’s SOS will change (for the better or worse) this season.

Here is how the D-Line ACTUAL (vs. projected) SOS shaped out in 2020:

TEAM1234567891011121314151617MeanTough MUsEasy MUs
CLE16306277173024231922151631291720.62.05.0
KC2332164241025291824525288213220.33.06.0
NE28142411259102916231232328102918.91.06.0
IND22262920130131615215232423172218.62.03.0
LAC3011185822252428291042124251118.43.05.0
DEN1517529411322124288111810322418.13.04.0
WAS1912116331273113302717914181917.92.05.0
BUF292832415112941412329172542817.83.06.0
TB818253220224318183112621132117.64.04.0
PIT312523191151627302216610307117.44.05.0
CIN321192216711517631282717231617.44.04.0
JAX715283023133223217126151620717.24.04.0
NO524213321820592125211911261816.83.03.0
MIA410221492931232252930114241016.83.05.0
LV18841011513225112129732282516.75.06.0
GB261382152326922720191318152016.63.02.0
TEN252226102317302071671221322316.54.03.0
BAL12311630191774151727122313016.35.04.0
DAL32114131126191726616309193116.34.04.0
ATL14272021826131825824832511516.05.04.0
LAR27191031692028145912429141215.63.04.0
SEA2142728261291031219312963915.64.05.0
PHI633091716312731114281227615.06.05.0
CAR2453212212082111513262526814.96.03.0
ARI9613182927142810144331199314.84.04.0
CHI1331217518381526213232622214.76.03.0
DET2021282221726618232021552614.66.02.0
NYG1720932761956193014121162714.45.03.0
MIN27152314212132027182252081314.45.01.0
HOU11161726221522214137207301514.35.02.0
NYJ1097251228101143228241431413.94.04.0
SF1229311928341428310627121413.96.04.0

Note the numbers next to the corresponding 17 columns represent the Offensive Line (OL) Ranking of the opponent for the team in the first column. For example, In week 3 Cleveland played the 6th rated OL, Washington. The lower the number the better rated the OL opponent was in 2020. You can see the mean D-Line SOS for each team towards the right.

Cleveland finished with the easiest DL SOS, and the 49ers with the hardest. I also included two columns to tally “Tough Matchups” and “Easy Matchups”, for games vs “<9th ranked OLs” and “>24th ranked OLs” respectively. We did this to help hone in on the “absolutes”, like playing 4 games vs BAD OLs.

Now looking at 2021 and the PFF PROJECTIONS, here is how that same analysis shakes out for the coming season:

TEAM1234567891011121314151617MeanTough MUsEasy MUs
DET9151227262481729127262111231918.42.05.0
CIN2627292215101228125291892112718.22.06.0
DEN3222281229251166171871024251818.13.05.0
CLE72027261811212924310121225152918.12.05.0
NO153133216195231417136285303118.04.05.0
DAL518173132326212372541632161117.94.05.0
KC11218171316143215256212518292417.92.04.0
PHI23967315251018214322816321617.74.05.0
MIA313252522231320122831322841417.24.05.0
LV12293018272117327246167121216.95.04.0
BAL25710212182426302712911524816.85.05.0
SF10171519112271182226192423142016.82.02.0
BUF2930162071430222824353132316.76.05.0
LAC16672511231726292124327202116.75.04.0
TB6238330172741632223134312816.76.05.0
ATL1753216283031463225319101316.45.05.0
NYJ31321142332421330201743022516.45.03.0
CAR28420617263223311163023135416.35.04.0
NYG2116234683172551730186172716.36.04.0
JAX202111241430191329238142028316.23.02.0
WAS18321323471521531192561761716.25.03.0
NE30284520628183112314132132216.15.04.0
PIT1325241521191271018241226147116.13.03.0
GB41092924271611719268271212616.04.05.0
LAR2725111932102014915221119261215.92.03.0
IND198143012209142822135203112515.83.03.0
CHI824110251559291210111526193215.73.04.0
MIN241119110316121815910292781515.33.03.0
HOU2213113321183014282192218914.65.03.0
SEA2142698294221511169208271014.44.03.0
TEN111922822137284203222993014.36.03.0
ARI142622891201593119278102614.25.03.0

Lastly, see the chart below to find the change in (delta) of SOS:

TEAM21_Mean21_Tough MUs21_Easy MUs20_Mean20_Tough MUs20_Easy MUsDelta (d) MeandTough MUsd Easy MUs
DET18.42.05.014.66.02.03.9-4.03.0
SF16.82.02.013.96.04.02.9-4.0-2.0
PHI17.74.05.015.06.05.02.7-2.00.0
NYJ16.45.03.013.94.04.02.51.0-1.0
NYG16.36.04.014.45.03.01.91.01.0
DAL17.94.05.016.34.04.01.60.01.0
CAR16.35.04.014.96.03.01.4-1.01.0
NO18.04.05.016.83.03.01.21.02.0
CHI15.73.04.014.76.03.01.0-3.01.0
MIN15.33.03.014.45.01.00.9-2.02.0
CIN18.22.06.017.44.04.00.8-2.02.0
BAL16.85.05.016.35.04.00.40.01.0
MIA17.24.05.016.83.05.00.41.00.0
ATL16.45.05.016.05.04.00.40.01.0
HOU14.65.03.014.35.02.00.30.01.0
LAR15.92.03.015.63.04.00.3-1.0-1.0
LV16.95.04.016.75.06.00.20.0-2.0
DEN18.13.05.018.13.04.00.10.01.0
GB16.04.05.016.63.02.0-0.61.03.0
ARI14.25.03.014.84.04.0-0.61.0-1.0
TB16.76.05.017.64.04.0-0.92.01.0
JAX16.23.02.017.24.04.0-1.0-1.0-2.0
BUF16.76.05.017.83.06.0-1.13.0-1.0
SEA14.44.03.015.64.05.0-1.20.0-2.0
PIT16.13.03.017.44.05.0-1.4-1.0-2.0
LAC16.75.04.018.43.05.0-1.82.0-1.0
WAS16.25.03.017.92.05.0-1.83.0-2.0
TEN14.36.03.016.54.03.0-2.22.00.0
KC17.92.04.020.33.06.0-2.4-1.0-2.0
CLE18.12.05.020.62.05.0-2.50.00.0
IND15.83.03.018.62.03.0-2.81.00.0
NE16.15.04.018.91.06.0-2.84.0-2.0

The last 3 columns represents the (expected) CHANGE in D-Line SOS from 2020 to this season. The numbers may appear marginal, but that’s truly where the value is.

We see teams like the Lions, 49ers and Eagles get a much easier schedule, while the Colts, Patriots and Browns have a more difficult road to the QB.

Now let’s bring it all together:

As mentioned previously here are the predicted UNDER-sack performers given our “Conversion of Pressure” analysis in the Last Year’s Sacks don’t Predict Next Year’s Sacks. They are pass rushers we already had identified as likely underperformers (based on higher than usual pressure-to-sack conversions):

PlayerPass RushesSacksPressuresconvExpected 21dSOS_21
Devin White10692733%68
Haason Reddick38812.54627%9.520
Dre’Mont Jones3376.52427%3.518
Denico Autry3957.53025%4.528 (TEN)
Trey Hendrickson33013.55425%10.58
Myles Garrett467124924%930
Za’Darius Smith51512.55224%9.519

You can see the “Expected 21” column, which represents how far we predict the player’s sack total to fall by adjusting their conversion rate back towards their mean in 2021 AND the “dSOS_21”, representing the change in SOS for his DL this season. Bring it altogether and I would absolutely:

  • Target Denico Autry at 4.5 or under sacks, and drumroll…
  • Myles Garret, his high conversion rate AND much tougher OL schedule to drag his sack total UNDER his number (we predict in the 8-9 range)

On the other end, the underperformers, and those we expect higher sack totals from are:

PlayerPass RushesSacksPressuresconvExpected 21dSOS_21
Jordan Jenkins2682268%64
Shaq Lawson3214538%813
William Gholston3523408%721
Cameron Heyward4794547%825
John Franklin-Myers3323427%74
Clelin Ferrell2582297%617
Al-Quadin Muhammad3612307%632
Robert Quinn3242316%69
Sam Hubbard3452316%611
Tyus Bowser2282336%612
Marcus Davenport2271.5276%5.58
Jonathan Allen4492375%626
K’Lavon Chaisson3121264%522
Arden Key2740270%417

Specifically:

  • Target John Franklin-Myers to exceed expectations at 7+ sacks (with a 7% conversion rate in 2020 and the 4th best DL SOS increase)
  • Robert Quinn, Sam Hubbard and Marcus Davenport to all exceed their sack totals given 6-7% pressure-to-sack conversions, and all having top 10 increases in “ease” of DL SOS

These are all very solid bets. Obviously anything can happen in an NFL season, AND I have to add (again, this one is much more subjective), ANY preseason total (counting stat/volume-based )wager I am erroring on the side of over/more. Why? Well, we have an extra game this season. Yes, bookmakers KNOW this, but like all of us this is new to them and they are likely trying to understand exact pricing of all bets (and when you don’t know something, you “do the same as always”). All of this is to say, If you are going to place a future wager this offseason, pick a player from that last list, with a good SOS change number, on the OVER.

See you next time.

Leave a Reply