clock menu more-arrow no yes mobile

Filed under:

Was Taylor Heinicke a Figment of Our Imaginations?

Investigating the phenomenon of winning with TH4 in the 2022 season

If you buy something from an SB Nation link, Vox Media may earn a commission. See our ethics statement.

Green Bay Packers v Washington Commanders Photo by Scott Taetsch/Getty Images

Taylor Heinicke is a Rorschach test for Commanders fans. The comment threads on Hogs Haven are full of strong opinions about his value and contributions to the team. Reading through them, I get the sense that most opinions tell us more about what the commenter wants to believe than they do about the Commanders’ backup quarterback.

The facts are that, for the third season in a row, the scrappy, undersized quarterback was pressed into service to replace Washington’s injured starter. When Heinicke took over from Carson Wentz, the Commanders were 2-4. Over the next six games, the Commanders went 5-1, including a shocking upset of the previously undefeated Philadelphia Eagles in Week 10.

The sudden reversal of fortune led many to conclude that something about Heinicke must have provided the spark to break the team out of its losing streak. The losing streak had actually ended the previous week in a 12-7 win over the Chicago Bears, in one of the worst games of football ever played. Nevertheless, as the wins started accumulating, more and more fans became convinced that Heinicke’s leadership and heads-up play were key factors.

The curious thing about the “winning with TH4” phenomenon is that, aside from intangible leadership qualities, which undoubtedly play a huge part in locker room dynamics, there is very little on the stat sheet to suggest that Heinicke’s play at QB was helping the team win games.

By most objective measures, Taylor Heinicke is not one of the NFL’s better starting quarterbacks. His passing production of 206.6 yards per game ranked 23rd in the league. He ranked 15th in net yards per pass attempt (6.18) and 26th in completion rate (62.2%). He had the 14th highest interception rate per pass attempt (2.3%), the 15th highest sack rate (6.8%), and the 4th highest rate of fumbles per game (0.78) among QBs with 5 or more starts. His QBR of 44.5 ranked 25th in the NFL.

In Week 11, following the Eagles upset, I got curious about what factors might be contributing to winning games when Heinicke was starting. As expected, the factors that correlated most strongly with winning were performance of the defense, in particular passing yards allowed and total yards allowed. The next most strongly correlated factors were turnovers and time of possession, which reflect contributions of the defense and the offense.

What came as more of a surprise were the results on offense. The factors which most strongly correlated with winning were run/pass ratio and rushing yards. Contrary to the growing belief that Heinicke was a key factor in the winning streak, the results showed that the more the team kept the ball out of his hands, the more likely it was to win.

In the next few weeks, the Commanders will need to make a decision on whether to bring back Heinicke, or seek another veteran as insurance in case Sam Howell doesn’t develop as hoped. As the discussion about the future of the backup QB position is beginning to heat up, it seems that the Legend of Taylor Heinicke is only growing. In an effort to sort out myth from reality, I decided to make another attempt to find something about Heinicke’s play at QB which is positively correlated with winning on the football field.

Winning with TH4 – Major Factors in 2022

My previous piece showed that the biggest factors driving success with TH4 under center were not QB-related, or only partly related to QB play. These included defense, rushing, turnovers, and time of possession. That analysis was based on three seasons, playing with very different teams. Before delving into QB-factors, I’d like to revise that analysis, looking just at the 2022 season, which now has a larger sample size, and adding another factor: Strength of Schedule.

Defense

In agreement with the pattern throughout Heinicke’s time in Washington, the Commanders tended to win games with Heinicke in 2022 when the defense limited opponents’ yardage. The following graph illustrates the correlation between total yards allowed on defense and point differential in games Heinicke started. A positive point differential means the Commanders won.

As in the previous analysis, there was a strong, negative correlation (r = -0.685) between yards allowed by the defense and point differential. In 2022, defensive performance, as measured by opponents’ yards, explained 47% of the variance of Commanders’ overall performance (point differential) in Heinicke’s starts.

Other factors on defense that also correlated strongly with point differential were opponents’ scoring (r = -0.80), which is not unexpected since opponent score is half of point differential, and takeaways (r = 0.63). Interestingly, point differential was not strongly correlated with opponents’ 3rd down conversion rate (r = 0.07), which could be interpreted to suggest that the offense was not taking full advantage of the opportunities that the defense gave them.

General Team Factors

Turnovers. I previously showed that overall turnover differential, reflecting contributions from offense and defense, was strongly correlated with game outcomes throughout Heinicke’s time in Washington. That remained the case in 2022. In fact, of all the factors I examined, it had the strongest positive correlation with point differential (r = 0.84). The Commanders were 3-0 in Heinicke’s starts when they won the turnover battle, 2-1-1 in games with a turnover differential of 0, and 0-2 in games when they lost the turnover battle. As we will see below, Heinicke was a significant contributor to the team’s turnovers.

Time of Possession. Starting with the Week 6 victory over the Bears, the Commanders began leaning heavily on the run, resulting in long drives which ate up the clock. The apparent switch in offensive philosophy coincided closely with Heinicke taking over from Wentz, as well as the end of a four-game losing streak and the start of a seven-game stretch with six wins and one loss. A narrative soon developed that dominating time of possession, was a key to winning, despite the Commanders fielding a relatively inefficient offense.

Various explanations were proposed regarding Heinicke’s role in this ball-control offense, most notably “keeping drives alive”. I will address Heinicke’s possible role below, but the first question is whether the ball-control narrative holds water. To address that question, I calculated the correlation between time of possession as proportion of total (to account for the Giants’ tie in overtime having 10 minutes more clock time than the other games) and point differential. There was, in fact a strong, positive correlation between time of possession and point differential (r = 0.52), which is consistent with the view that controlling the clock was a key success factor during Heinicke’s starts.

Strength of Schedule.

Commanders’ fans have a history of attributing winning streaks to mediocre starting quarterbacks. Most recently, Alex Smith was widely credited with the team’s 11-5 record despite averaging only 209 passing yards per game (identical to Heinicke) and throwing 16 TDs to 13 interceptions during his time in DC. One fact that Smith fans seem to have overlooked was that he benefitted from an unusually soft schedule of opponents, facing only five teams with winning records in 16 starts. Might history have repeated itself with Heinicke?

In nine starts in 2022, Heinicke faced four opponents that ended the season with winning records (NYG x2) and four that finished with losing records. During Heinicke’s starts, the Commanders were 5-3-1 overall. Against losing teams, they were 4-0, while against winning teams they were 1-3-1. On that basis it seems plausible that much of the team’s success, and failure, during Heinicke’s starts had to do with strength of opponents.

Consistent with this view, opponents’ win total (r = -0.42) and season point differential (r = -0.44) both had moderately strong correlations with point differential in Heinicke’s starts.

Offense

Rushing & Turnovers. On offense, there was a strong positive correlation between rushing yards and point differential (r = 0.57) and a strong negative correlation with giveaways (r = -0.57). In other words, the more the Commanders gained on the ground and the less they coughed up the ball, the more they tended to outscore their opponents. Offensive scoring was also strongly correlated with point differential (r = 0.58). However, the correlation was not nearly as strong as with points allowed on defense (r = -0.80), suggesting that defensive performance had more to do with wins and losses than offense in Heinicke’s starts.

QB Play. This is where things get interesting. The third-strongest correlation with point differential amongst all the factors I examined was with run/pass ratio (r = 0.74). The more the Commanders ran instead of passed, the more they outscored their opponents. That seems to suggest that the more they put the ball in Heinicke’s hands, the more likely they were to lose.

To test that out, let’s see how factors directly in Heinicke’s control relate to point differential. Following on from the observation about run/pass ratio, the most obvious place to start was pass attempts. There was a modest, negative correlation between pass attempts and point differential (r = -0.37), consistent with the suggestion that the more Scott Turner put the ball in Heinicke’s hands, the less likely the Commanders were to win. The most surprising finding was that the same thing turned up when I looked at passing yards, as shown in the next graph.

The graph illustrates the relationship between rushing yards (burgundy) and passing yards (gold) in Heinicke’s 2022 starts. As discussed above, the more yards the Commanders gained on the ground, the more likely they were to win. The correlation with passing yardage, on the other hand, was negative (r = -0.40). This means that the more the Commanders gained through the air, the more likely they were to lose. Even more remarkably than that, passing touchdowns had an even stronger negative correlation with point differential (r = -0.58).

At first glance, that doesn’t seem to make any sense. How is it possible that passing yardage and touchdowns could be negatively related to game outcomes? The answer might have to do with the other things that tended to happen when Heinicke threw the ball.

As we saw above, one of the strongest correlations with winning and losing was turnover differential. Let’s see how much of that was attributable to Heinicke. Heinicke had slightly lower than league average interception and sack rates, but was the fourth-most fumble-prone QB in 2022. As expected, his sack totals per game had a moderate, negative correlation with point differential (r = -0.34), while QB-turnovers (fumbles lost + interceptions) was more strongly correlated (r = -0.57).

Putting these findings together leads to an obvious explanation for the negative correlations between passing yardage and touchdowns with point differential. Heinicke never passed for more than 270 yards in 2022. It appears that the gains associated with his modest aerial production were not enough to make up for his tendency to take sacks and turn over the ball when he was involved in the passing game.

The Elusive It-Factor: Keeping Drives Alive

To summarize the findings thus far, it appears that the Commanders’ success during Heinicke’s starts in 2022 were strongly related to performance on defense, turnovers, rushing production, high run/pass ratio, time of possession and strength of opponents. Remarkably, the factors most directly related to Heinicke’s play, including passing production, sacks and QB turnovers were all negatively related to the team’s success during his starts.

This is consistent with Heinicke’s -0.08 Expected Points Added per dropback (EPA/dropback) value, which ranks 28th among NFL QBs in 2022 (Carson Wentz ranked 30th at -0.11), and indicates that, on average he detracted from the team’s chance of scoring when he played.

It has been suggested by myself and others that Heinicke’s contribution to the Commanders’ winning formula at midseason might be too subtle to show up using conventional statistics. It is often suggested that his high football IQ and heads-up play allows him to keep drives alive without necessarily registering big gains on the stat sheet. To get at these more elusive qualities, earlier in the season I compared Heinicke to Wentz in terms making “clutch plays” to keep drives alive, and avoiding drive-killing errors. It wasn’t hard to demonstrate that Heinicke is better than Wentz in both areas, but the real question is whether Heinicke is any better than potential replacements in free agency.

To determine whether Heinicke really does possess magical abilities to keep drives alive, I revisited my previous analysis. This time, rather than comparing two quarterbacks’ abilities to pull off clutch plays and avoid drive-killing errors, I did a within QB comparison, asking whether Heinicke keeps more drives alive than he kills.

To answer that question, I went drive-by drive through each of Heinicke’s 2022 starts and counted Drive Ending Plays and Drive Extending Plays.

A Drive Ending Play was defined as any pass play or QB run which resulted in the end of an offensive drive without a score. These included QB turnovers and any failure to convert on third or fourth down, including incompletions and completions short of the line to gain. The latter might be placing blame on Heinicke for Scott Turner’s failings, since Turner was fond of calling screens and short passes on third and long. However, it is ultimately the QB’s responsibility to execute the play for a first down, and it is often difficult to tell what the primary play call was and what other receivers were open.

A Drive Extending Play was defined as a passing completion or QB run resulting in a completion on 3rd or 4th down to keep a drive alive, excluding scoring plays, which have been dealt with above. Some readers may wonder why I’m not counting first down conversions on earlier downs since they also extend drives; I will come back to first downs in a few paragraphs. For now, the intention is to provide a metric that is directly comparable to Drive Ending Plays. The intention was to focus on clutch situations, which seems to be what this all about, so I chose to focus on situations where the drive is on the line. However, I didn’t think it was appropriate to ignore turnovers on earlier downs under Drive-Ending Plays, which is why they are included.

The results of this analysis are shown in the following table:

Once again, Taylor Heinicke’s special ability has proven elusive. While he did make plenty of clutch plays to keep drives alive during his nine game run as the starter in 2022, he failed to execute and committed errors on more than twice as many plays, leading to the end of offensive drives. On balance, he would seem to be more of a drive killer than a drive extender.

QB 3rd Down Conversion Rate. That sounds pretty bad, but we really want to know is how Heinicke compares to other QBs. The analysis of Drive Ending and Drive Extending plays was labor intensive, and I’m not about to repeat it for 31 other QBs. For a quicker, if somewhat less comprehensive answer, I compared all 33 QBs with 60 or more passing attempts in terms of QB 3rd down conversion rates defined as Passing 1st downs + Rushing 1st downs divided by Passing Attempts + Rushing Attempts on 3rd downs.

Heinicke had a QB 3rd down conversion rate of 30.5%, which ranked 32nd out of 33 QBs. The only QB who was worse at converting on 3rd down was Baker Mayfield at 29.2%. Surprisingly, Carson Wentz ranked 17th, with a conversion rate of 42%.

QB 1st Downs Per Game. Of course, 3rd down conversion only tells us about clutch situations where drives are on the line, and ignores drive ending turnovers on earlier downs. Another way to look at a QB’s ability to keep drives going is to simply look at first down conversions regardless of down.

In my final attempt to determine whether Taylor Heinicke has a special ability to extend offensive drives, I calculated total QB first downs, defined as first downs on passing plays and QB runs, for the 39 NFL QBs with five or more starts in 2022. To adjust for differences in numbers of games played, I divided by number of games to give QB First Downs/Game. The top five NFL QBs by this metric were Patrick Mahomes (17.5 1st Downs/G), Josh Allen (16.8), Joe Burrow (15.6), Jalen Hurts (15.5), and Justin Herbert (14.4). It appears to be a pretty solid measure of QB ability.

The good news for Taylor Heinicke fans is that he looks a bit better by this metric, ranking 23rd in the NFL at 11.0 First Downs/Game and was only one place behind teammate Carson Wentz at 11.1 First Downs/Game. Of course, 23rd out of 39 is not really that good.

If Taylor Heinicke does have some special gift for keeping drives alive, it appears to be well hidden.

NFL-Dallas Cowboys at Washington Football Team Photo by Jonathan Newton/The Washington Post via Getty Images

Conclusions

Like most great works of mythology, there is some truth to the Legend of Taylor Heinicke. We know for a fact that he was sleeping on his sister’s couch, studying for Master’s degree in Mathematics when he got the call from Coach Rivera. The pylon dive in 2020 Wild Card game, in the closest playoff contest faced by the soon-to-be Super Bowl Champion Tampa Bay Buccaneers, is recorded for posterity on All-22 film. Heinicke’s teammates rallied behind his leadership throughout the 2022 season, and credit him for his role in their wins.

The question I have attempted to answer is whether any of that had to do with his play on the field, or whether it was simply a case of being in the right place at the right time. All of the factors that strongly correlated with winning in Heinicke’s starts were most directly related to other players, particularly the defense and running backs and factors outside of Heinicke’s control, such as strength of schedule.

The team appeared to do better the more they got the ball to other players and kept it out of Heinicke’s hands. His contribution in the passing game seemed to have a net detrimental effect on the team’s ability to win games. Finally, my search for the elusive knack for extending drives came up empty.

In conclusion, it appears that Heinicke benefitted early in the season from playing for a team with a strong defense, a strong enough rushing game to hide his deficiencies as a passer and a weak schedule of opponents. When the Commanders started facing stronger opponents around Week 13, and the defensive play began to fall off due to key injuries, the team’s weakness on offense was exposed, bringing the midseason winning streak to an end. Heinicke was fortunate to ride the rollercoaster while it lasted, but I can find no evidence that he was the driver.

I have no doubt that the Legend of Taylor Heinicke will continue to grow over time, as we remember the pylon dives and the Jordans and forget about all the hospital passes, sacks, and completions short of the sticks on 3rd downs. On the other hand, I can see no reason for that particular mythology to influence Eric Bieniemy’s search for a veteran backup for Sam Howell.

Acknowledgement: Edited by James Dorsett


Poll

Which vet QB should the Commanders get to back up Sam Howell?

This poll is closed

  • 46%
    Taylor Heinicke
    (507 votes)
  • 11%
    Jacoby Brissett
    (121 votes)
  • 5%
    Teddy Bridgewater
    (56 votes)
  • 6%
    Andy Dalton
    (70 votes)
  • 9%
    Gardner Minshew
    (103 votes)
  • 3%
    Baker Mayfield
    (40 votes)
  • 8%
    No vet, draft a rookie
    (89 votes)
  • 5%
    Lamar Jackson
    (64 votes)
  • 4%
    Someone else
    (49 votes)
1099 votes total Vote Now