Sunday, November 15, 2026

QB Kneeldowns for Strat-O-Matic Football

 


SOM PRO FOOTBALL LINKS

QB Kneeldowns for Strat-O-Matic Football

A question came up on the Delphi Forum about QB Kneels.

Strat O Matic Football began accounting for kneel-down data in quarterback running ratings starting with the 2014 season set, which was released in the summer of 2015.

Rather than counting kneel-downs against a quarterback's rushing performance, Strat-O-Matic updated its formula to exclude kneel-down yardage from the card ratings.

(Other changes from that season forward included:

Split Defensive Ratings: Linebackers and safeties received separate pass and run defensive ratings (with the pass rating underlined).

Individually Customized Defensive Squares: Starters received personalized defensive squares showing player names, team names, positions, and ratings, plus a backside featuring secondary positions (with five blank reserve/rookie write-in squares per team sheet).

Yards After Catch (YAC): New YAC ratings and chart integration were added for receivers (starting with the 2013/2014 Windows seasons).

Kickoff Modifications: Default starting positions and touchback rules were updated with four distinct combinations per kicker along with -8/-9 return options. )

Key Details of the Change

Why It Changed. Previously, official NFL rushing statistics—which penalized quarterbacks for negative yards on victory-formation kneel-downs—were used directly to rate a quarterback's running ability. Excluding those plays created a more accurate reflection of a quarterback's true scrambling and rushing capability.

SOM Card Stats vs. Ratings. The printed rushing totals on the physical and digital cards still display the quarterback's official real-life NFL statistics (which include kneel-downs). However, the actual dice-roll results on the QB running card ignore those negative plays.

In-Game Kneel-Downs. Because kneel-downs are stripped from regular rushing ratings, the game handles end-of-game clock management separately through a dedicated "QB Takes a Knee" rule (usable in the final two minutes of a game).

I'm not going to argue that the changes weren't welcome, but I will point out (without going into this in too much detail) that prior to these changes, the QB rushing model had a bit of a "fudge factor" to account for these types of attempts.  The more QB attempts the less the model would subtract from the QBs rushuing baseline. Was it perfect?  No. But if you consider there are about 75 QBs in a full-size season, I would say having made a few hundred teams that this will account for 65-70 of them.  There are probably around 3-5 QBs per season or so that truly defied this model. And that's probably why they chose to look at this on a "go-forward" basis.

For seasons before 2014, that is, for seasons before where Victory Formation was as much a clock management tool as a game-ending playcall, the improvement/re-evaluation was going to be marginal in a lot of cases.  What would not be marginal is the effect on leagues in flight, and the effort needed to extend this code and approach backwards through time.  Would it drive additional revenue for seasons more than a dozen years old?  In some cases, the juice might not be worth the squeeze.    

Here are some examples from 2014: (the original model would fudge these UP 1.0-1.3 yards)  These QBs had the most kneels (22) in that timeframe:

Sam Darnold 2024

ATT 67 Yds 212 Overall average 3.16

Minus his 22 kneels: ATT 45 Yds 190  Carding average 4.22

Joe Burrow 2022

ATT 75 Yds 257 Overall average 3.43

Minus his 22 kneels: ATT 53 Yds 235  Carding average 4.43

Josh Allen 2020

ATT 102 Yds 421 Overall average 4.13

Minus his 22 kneels:  ATT 80 Yds 399  Carding average 4.99



Josh Allen

Of these QBs I would say 2020 Josh Allen stands up as a possible candidate for special handling.  The others are not likley to change by much.  But what helps to decide this is the data itself.  I know SOM has its sources; I don't know what these are.  But for our purposes, complete and structured play-by-play data for quarterback kneels in the NFL exists, and it’s quality varies depending on the public analytics source you choose:

Primary Data Sources and Ranges

nflfastR / NFL Play-by-Play: Public digital play-by-play data sets (such as nflfastR) track full play-by-play information, including specific binary flags like qb_kneel, going back to 1999. The only question I would have here would be the accuracy of that flag. If it were over or underapplied there might be problems.  But if the binary _qb_kneel_ is good this extends back the evaluation window to 1999-2013, or fifteen more seasons. 

Pro Football Reference: Full play-by-play data and play-level tracking on Pro-Football-Reference.com are complete back to 1977/1978 (following expansions covering the 1977 and 1978 16-game schedule eras).  You’d need a model to evaluate this via AI and you’d need to train and check the model. 

Box Scores and Rushing Statistics: Individual player rushing stats (where a kneel is traditionally recorded as a rushing attempt for -1 or -2 yards) go back much further—with season stats available back to 1920 and game stats back to 1933—though individual plays were not explicitly tagged as "kneels" in early historical ledgers. For the record 1947 (in both the NFL and AAFC) was the first Pro season where sacks were not counted as rushing attempts.  

But in some cases, it was a judgment call from the official scorer when there was a run for lost yardage versus when it was a sack.  There were plays where it looked like a QB was running with intent behind the line but moving forward and faking a pass and he got tackled.  Sometimes it was credited as a sack and other times it was not. There is a gray area in these judgments during this era.

NFL Playcalling Evolution

While exact play-level counts before 1999 are not cleanly aggregated in public ledgers, the frequency of the victory formation has evolved through three distinct eras driven by rules, play-clock standardization, and modern analytics.

1. Pre-1978: Near-Zero Usage

There were Rule barriers that prevented using this tool for incremental clock managment. Before 1987, NFL rules required a defender to physically touch a downed ballcarrier for the play to end. Simply dropping to a knee left the ball live and exposed the quarterback to hits.  There were also limitations in Conventional Playcalling. Coaches viewed taking a knee as passive or unsportsmanlike. Instead, teams relied on quarterback sneaks or fullback dives into the line to run out the clock.



2. 1978–1987: Rapid Adoption After the "Miracle at the Meadowlands"

The 1978 catalyst was The Miracle at The Meadowlands.  On November 19, 1978, the New York Giants attempted a late handoff instead of kneeling against the Philadelphia Eagles. A botched exchange resulted in a defensive scoop-and-score. Within weeks, coaches across the league began adopting tight formations with deep trailing protectors to eliminate handoff risk.  

Because quarterbacks still had to be touched to be considered down, defenses frequently dove over the pile to hit the kneeling passer. By 1987 the rules changed regarding play timing.  The NFL officially legalized "giving oneself up" by allowing a quarterback to take a knee and end the play immediately without defensive contact. 


3. 1988-1998: This 1987 rule change made the qb_kneel play standard procedure at the end of close games.  But the play was still not used in the volumes we see now.


4. 1999–Present: Modern Optimization and Steady Volume

Play clock standardization. Following the adoption of the modern 40-second play clock in the 1990s, offenses gained a predictable mathematical formula. Roughly two minutes of game time could be erased with three consecutive kneels when the defense had no timeouts.

Consistent modern frequency. Analytics tracking shows that starting quarterbacks average around 9 to 15 kneels per season, while top-performing quarterbacks on winning teams frequently log 15 to 20 or more. Over the five-season span from 2021 through 2025 (including playoffs), league leader Patrick Mahomes recorded 89 kneel-downs—an average of nearly 18 per year.

Expanded situational use. Beyond ending the fourth quarter, modern analytics-driven coaches now increasingly use the victory formation to close out the first half when pinned deep in their own territory, prioritizing zero turnover risk over low-probability scoring drives.

Analysis

There are certainly QBs in all of these eras that could use re-evalution; I would question, even if the data were available, if it would be ALL of them.  But I do believe a holistic approach could improve the modeling of QBs contributions on the ground; by "holistic" I mean the rules and CMs need to be evaluated as well as carding techniques.  It doesn't do any good to change cards if the rules and CMs won't pass the positive incremental changes through to game flow.  But do I think there are examples that could be re-evaluated ? Sure.

Click to magnify:


Fred Bobberts

Initial Date of Publication:  November 15, 2026

No comments:

Post a Comment