Last Year’s Sacks don’t predict Next Year’s Sacks

But Hurries do

It’s been a few years since the analytics community taught us sacks shouldn’t be the “end all, be all” metric for pass rushers. Yet  many still fall for the trap. Even NFL teams, and their GMs, those who are paid (alot and solely) to select good players for their team, continue to “pay for sacks”. This stupidity offers us an opportunity to take advantage and profit.

Diving into this topic more in depth, in “Pass Rushing Slugging Percentage” we explained that the “last step in a chain/sequence” is never as consistent/predictive as the first/initial step(s). Imagine you were predicting home runs, and think through the “chain of events” a hitter must accomplish to complete the feat:

• Make contact with the ball (not whiff/k)
• Swing with enough power (usually measured by exit velo)
• Have the correct trajectory (measure of the swing angle) 
• Hope that the ballpark he’s hitting in, has small enough dimensions for the ball to get over

In other words, with an “extreme” or “absolute” stat, there are typically many steps that need to be completed beforehand. And given the very nature of variance in sports, the first steps are almost always going to be more consistent than the raw “absolute number”. The same goes for sacks.’ 

We’ve been preaching how much pressures should be used as a much better “success indicator” for a pass rusher, is a story for another day. Today, we are simply going to look at how we can profit on its predictive ability based on this very clear phenomenon. 

Take a look at the last 4 seasons, and how Year (n)’s sacks relate to Year (n-1)’s pass rushing numbers.

*Note data was limited to players with 24 pressures within one season.

As you can see total sacks, or really any form of sacks only predict about 20% of next season’s sack numbers for an individual pass rusher. However, Pressures, or better yet, pressure rate is more than twice as correlated. Not only are pressures a better performance metric, they are a hidden gem in regards to predicting how sack totals may rise or fall in the following season. 

Looking at how much a pass rusher turns his pressures into sacks is called the conversion rate. Historically, we’ve found the average pass rusher “converts” about 16% of pressures into sacks. With that said, over the last four years, looking at the upper and lower quartiles (those that converted >24%, and <8% conversion rates the season prior) we see some stark predictive value in which way sack totals move in the ensuing season.

Pass Rushers that exceeded 150% expected sack conversion (Sacks / Pressures), and what their net sack rate and totals were the following season (>24% Pressures converted into Sacks in n-1 year):

Player	SK	SK %	pSSN SK	pPRESS	pPRESS%	pconv	pssn sk %	Net Sk% Nxt Ssn	Net Sk Nxt Ssn
Julius Peppers	5	2%	11	27	8%	41%	3%	-2%	-6
Erik Walden	4	2%	11	30	10%	37%	4%	-2%	-7
Danielle Hunter	7	1%	13	43	11%	29%	3%	-2%	-6
Jordan Jenkins	2	1%	8	28	9%	29%	3%	-2%	-6
Vic Beasley Jr	5	2%	16	55	13%	28%	4%	-2%	-11
Von Miller	8	2%	15	53	12%	27%	3%	-1%	-7
Denico Autry	4	1%	9	33	10%	27%	3%	-2%	-6
Mario Addison	11	3%	10	35	12%	27%	3%	-1%	2
Jamal Adams	10	10%	7	24	30%	27%	8%	2%	3
Bruce Irvin	9	3%	7	24	10%	27%	3%	0%	2
Matt Ioannidis	9	2%	8	28	11%	27%	3%	-1%	1
Jason Pierre-Paul	9	2%	13	47	9%	27%	2%	0%	-4
Nick Perry	7	2%	11	42	12%	26%	3%	-1%	-4
Chandler Jones	19	3%	13	50	10%	26%	3%	1%	6
Julius Peppers	11	3%	8	29	8%	26%	2%	1%	4
Lorenzo Alexander	3	2%	13	49	16%	26%	4%	-2%	-10
T.J. Watt	15	3%	14	55	13%	25%	3%	0%	1
Ifeadi Odenigbo	4	1%	7	28	11%	25%	3%	-2%	-4
Justin Houston	8	2%	11	44	10%	25%	3%	0%	-3
DeForest Buckner	8	2%	12	48	10%	25%	2%	-1%	-5
Jonathan Allen	6	2%	8	32	7%	25%	2%	0%	-2
Jordan Jenkins	8	3%	7	28	10%	25%	3%	0%	1
Sam Hubbard	9	2%	6	24	7%	25%	2%	0%	3
Cameron Heyward	8	2%	12	48	11%	25%	3%	-1%	-4
Chris Jones	9	2%	16	63	13%	25%	3%	-1%	-7
Frank Clark	9	2%	10	41	10%	24%	3%	-1%	-1
Benson Mayowa	6	2%	7	29	13%	24%	3%	-1%	-1
AVG								-1%	-3

In other words, those that vastly out “convert” pressures into sacks in the previous year average about 3 less sacks the next year.

On the flip side, the “unlucky” pass rushers that converted 50% of the expected sack conversion, saw the following changes to their sack total the next year (<8% conversion rate in n-1 year).

Player	Sk	Sk %	pSSN SK	pPRESS	pPRESS%	pconv	pssn sk %	Net Sk% Nxt Ssn	Net Sk Nxt Ssn
J.J. Watt	5	0.90%	4	50	17%	8%	1%	0%	1
Allen Bailey	6	1.14%	2	25	7%	8%	1%	1%	4
DeMarcus Lawrence	6.5	1.86%	5	63	15%	8%	1%	1%	2
Leonard Floyd	10.5	2.29%	3	38	10%	8%	1%	2%	8
Samson Ebukam	4.5	1.93%	3	38	14%	8%	1%	1%	2
Brandon Graham	8.5	1.75%	4	51	10%	8%	1%	1%	5
Maliek Collins	3	0.95%	2.5	34	8%	7%	1%	0%	1
Vinny Curry	3	0.88%	2.5	34	12%	7%	1%	0%	1
Tarell Basham	3.5	1.00%	2	28	10%	7%	1%	0%	2
Ndamukong Suh	6	1.15%	2.5	36	6%	7%	0%	1%	4
DeForest Buckner	12	2.38%	3	45	9%	7%	1%	2%	9
Ezekiel Ansah	12	4.05%	2	31	10%	6%	1%	3%	10
Tom Johnson	2	0.45%	2	31	9%	6%	1%	0%	0
Adrian Clayborn	4	1.33%	2.5	43	16%	6%	1%	0%	2
Sheldon Richardson	1	0.27%	1.5	26	6%	6%	0%	0%	-1
Sheldon Rankins	8	1.92%	2	35	7%	6%	0%	2%	6
Leonard Williams	5	1.00%	2	46	10%	4%	0%	1%	3
Romeo Okwara	10	2.51%	1.5	37	11%	4%	0%	2%	9
Sheldon Richardson	4.5	1.11%	1	27	7%	4%	0%	1%	4
Kerry Hyder	8.5	2.05%	1	30	11%	3%	0%	2%	8
Leonard Williams	11.5	2.40%	0.5	43	10%	1%	0%	2%	11
AVG								1%	4

The underperforming sack converters typically add 4 raw sacks the next year. To summarize, that’s a 7 sack swing between those that outperformed their conversion rates, and those that underperformed. 

Knowing this, here is a list of those >24% / <8% converters from the 2020 NFL season, and what we can expect for 2021 sack futures:

Player	Pass Rushes	Solo Sacks	Ast. Sacks	Comb. Sacks	Unblocked Sacks	Hurries	Hits	Knockdowns	Pressures	conv	Expected 21
Devin White	106	9	0	9	5	13	14	7	27	33%	6
Haason Reddick	388	12	1	12.5	2	24	22	6	46	27%	9.5
Dre’Mont Jones	337	6	1	6.5	0	13	8	4	24	27%	3.5
Denico Autry	395	7	1	7.5	1	22	8	2	30	25%	4.5
Trey Hendrickson	330	13	1	13.5	1	23	30	11	54	25%	10.5
Myles Garrett	467	11	2	12	0	25	22	7	49	24%	9
Za’Darius Smith	515	12	1	12.5	0	31	25	11	52	24%	9.5
Jordan Jenkins	268	1	2	2	0	19	12	3	26	8%	6
Shaq Lawson	321	4	0	4	0	27	39	15	53	8%	8
William Gholston	352	3	0	3	0	20	30	17	40	8%	7
Cameron Heyward	479	3	2	4	0	33	38	13	54	7%	8
John Franklin-Myers	332	3	0	3	0	28	23	11	42	7%	7
Clelin Ferrell	258	2	0	2	0	18	14	8	29	7%	6
Al-Quadin Muhammad	361	2	0	2	0	22	15	4	30	7%	6
Robert Quinn	324	2	0	2	0	19	16	6	31	6%	6
Sam Hubbard	345	2	0	2	0	23	21	9	31	6%	6
Tyus Bowser	228	2	0	2	0	22	22	12	33	6%	6
Marcus Davenport	227	1	1	1.5	1	21	15	7	27	6%	5.5
Jonathan Allen	449	1	2	2	0	16	27	11	37	5%	6
K’Lavon Chaisson	312	1	0	1	0	19	12	8	26	4%	5
Arden Key	274	0	0	0	0	16	21	11	27	0%	4

If any player in the green section’s sack total number is under, hit the over strong. Vice versa for those in red. As soon as individual sack futures are listed, we’re going to hammer them using logic.