Explosiveness, Special Teams, and the Road to Understanding Clemson’s Football Struggles

As a Clemson alum, I am very much not thrilled by the Tigers’ 1-3 start to the season! I think it’s very easy to criticize Dabo (who must remember his generous pay package), but it’s important to use data to make sure that we don’t spread narratives that aren’t supported by the facts.

Dabo’s buyout clause is massive ($60 million in 2025), so I don’t think you can fire him over a bad season (or even maybe two or 3). I think there needs to be clear signs of decline/downtrends that could hurt the program in the long term to make that call. Ironically, I do think the magnitude, structure, and length of Dabo’s deal creates the potential for issues of alignment and accountability, given how expensive it is to fire him.

To myself and other fans I’ve talked to, it subjectively seems like Clemson’s Program is not developing players like it used to. I remember how exciting it used to be to watch each new Clemson team and see returning starters take big steps forward and the “next man up” filling the gaps from graduating stars. Casually, it seems like this has dried up. What happened to “Wide Receiver U?”

I think measuring player development is hard, and honestly I don’t have a measure for that right now (though hopefully it’s something I will cover on its own in a future article). I think looking at the number of big plays per game is a good proxy for this. While it doesn’t show anything about a player’s skill compared to the past, I think it’s reasonable to propose that a team that is developing players well should be able to create more big plays (20+ yards) per game. Clemson recruits well, so something is wrong if their players aren’t ultimately turning into dominant players able to burn their man in the secondary.

Although the average number of chunk yardage plays per game isn’t necessarily a clean measure of anything, it’s also downstream of play selection, which is another reason it interests me. If, between the explosiveness of the players and the play calling, Clemson is not generating many big plays, that puts pressure on the offense in multiple ways. Not only does it mean there are more opportunities for the drive to stall, but your opponent can also adapt to your narrowed playbook.

I remember occasional jokes when I was in school about Dabo basically being a glorified cheerleader (which always felt unfair but with a grain of truth), so I think it’s natural to look into how much of an impact Tony Elliott had as OC and Co-OC. I think it’s fair to say that Clemson’s offense has not been the same since he left. If part of Dabo’s genius is leaning on a great staff, then that might cause him to be more sensitive to changes in coordinators. And with massive job security, there’s also the possibility of long-term mismanagement of the staff compounding things.

After being shocked by Georgia Tech’s game-winning, running-clock field goal a couple weeks ago, of course we are also going to ask ourselves, “Could Clemson’s unit have made the same kick?”

Ultimately, we’ll never know, but historically, there have been plenty of examples of Clemson being bitten by special teams mistakes or clutch execution from the other team in close/winnable games (Pitt 2016, 2016 NCG, Louisville 2024).

Clemson has had some great special teams players under Dabo (like Catanzaro), but also lots of years of real struggles. I remember getting emails about emergency kicker tryouts when I was a student. That’s not the kind of thing that’s supposed to happen at an elite football program!

So, I have a couple hypotheses I want to investigate in this article: 1. Tony Elliot’s departure has led to a decline in big plays (25+ yards) 2. Clemson’s special teams have been below par for the ACC for the past decade or have been in decline (ok sorry that’s 2 hypotheses)

I don’t think either of these things is necessarily “the thing” wrong with Clemson’s current team, but they are important to consider over the long-run and could be relevant to understanding the current struggles.

To investigate these questions, I loaded play-by-play data from cfbfastR for the 2014-2025 seasons, as well as game-level data for Clemson games to get the win/loss outcome and final score (note: all data for 2025 is incomplete and is taken after Clemson’s loss to Syracuse). Let’s take a look at some charts and models!

Big Plays

First, I did find that the number of big plays Clemson generates on offense correlates with winning, but it is weaker than I expected at 0.1847572 for 20+ yards. Whether or not it’s statistically significant comes down to where you draw the line for a “big play.”

A Clemson team that isn’t frequently generating chunk yardage plays more likely to lose. It’s also just less fun for the fans and has game theory disadvantages that I noted before. But really, Clemson doesn’t feel like Clemson without big plays.

It isn’t clear to me that there’s a temporal effect here, but I think graphically we do see Elliot’s departure may have had an impact. In order to test this, I compared several linear regression models.

First, I looked at a simple model with Tony Elliot’s presence as OC/Co-OC as the only predictor of big plays per game.

Call:
lm(formula = big_plays ~ tony_elliott_season, data = big_play_summary)

Residuals:
    Min      1Q  Median      3Q     Max 
-4.7604 -1.5789  0.2396  1.4211  7.2396 

Coefficients:
                        Estimate Std. Error t value Pr(>|t|)    
(Intercept)               4.5789     0.2959  15.477  < 2e-16 ***
tony_elliott_seasonTRUE   1.1815     0.3735   3.163  0.00189 ** 
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 2.234 on 151 degrees of freedom
Multiple R-squared:  0.06215,   Adjusted R-squared:  0.05594 
F-statistic: 10.01 on 1 and 151 DF,  p-value: 0.001887

This suggests that having Tony Elliot as OC/Co-OC is associated with about 1.2 more 20+ yard plays per game, which is a pretty big deal. However, this model doesn’t account for other factors that could be influencing big plays. What if there’s a general downtrend over time?

Let’s look at a model with both season and Tony Elliot’s presence:

Call:
lm(formula = big_plays ~ season + tony_elliott_season, data = big_play_summary)

Residuals:
    Min      1Q  Median      3Q     Max 
-4.6363 -1.7150  0.1278  1.3244  7.2850 

Coefficients:
                         Estimate Std. Error t value Pr(>|t|)  
(Intercept)              84.04649  129.88898   0.647   0.5186  
season                   -0.03932    0.06426  -0.612   0.5416  
tony_elliott_seasonTRUE   1.04772    0.43344   2.417   0.0168 *
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 2.238 on 150 degrees of freedom
Multiple R-squared:  0.06448,   Adjusted R-squared:  0.05201 
F-statistic: 5.169 on 2 and 150 DF,  p-value: 0.006744

While it looked like there might have been a downtrend over time in the charts, it looks like that effect isn’t statistically significant when accounting for Tony Elliot’s presence. However, Tony Elliot’s presence is still statistically significant, and the effect size is similar to the previous model.

Using season alone to model this did yield a statistically significant coefficient, but the R-squared was much lower than the other 2 models (which if we’re being honest, aren’t great predictive models either).

Therefore, we can say that with a high degree of confidence that Clemson’s offense was more explosive with Tony Elliott. What I can’t say is how much of that is from play calling versus player development or confounders. However, with a healthy player development pipeline, I would expect that our superior players would be able to beat their man frequently for chunk yardage.

If you’re a Clemson fan (even during the Tony Elliott years), you’ve probably seen a game where it felt like the Tigers were trying to win the game with just bubble screens and RPO slants, and draws. It’s pretty nerve wracking because a lot has to go right!

I think the first hypothesis, that Tony Elliott’s departure has led to a decline in big plays (20+ yards), is supported by the data. But controlling for that effect, we don’t see a clear downtrend over time. Hopefully, this means that the Tigers coaching staff can adapt instead of it being a situation where Dabo has gradually lost some of his touch.

Special Teams

With special teams data, there is inherently a lot of noise in the data. There’s a lot of confounders for many of the stats. But an elite coach getting paid like Dabo should be able to maintain at least average special teams for the ACC in the long run and hopefully improve the kicking game over time, right?

If defense wins championships, special teams steal points and possessions in close games on your way to the championship. It’s a weapon that can get you points and field position, but there also isn’t a ton of room before it becomes a liability.

First, let’s look at field goal accuracy vs. the rest of the ACC and the national average (with trend lines):

This is always going to be more variable for a team than for a conference or the whole country, but this chart doesn’t inspire confidence because of the downtrend. And it seems like ACC kickers have improved over the same time span. But naturally, it matters a heck of a lot where you kick from and where you’re not willing to kick from.

When I first started started watching Clemson football, Dabo was actually very willing to kick long field goals. Using Spencer Benton as a long-range field goal specialist was innovative and successful! And that was at the same time that Catanzaro was on the team, probably the greatest kicker in Clemson history. But my dataset sadly doesn’t go back quite that far, and in recent years, it seems like Dabo has been reluctant to kick long field goals.

Next, let’s look at a chart that compares us to Florida State in longer FG attempts. I think this is important because if a team is NEVER willing to attempt field goals beyond a certain distance, we have to assume the teams options are limited in those situations.

It’s pretty jarring honestly how Clemson didn’t attempt a 50+ yard field goal for years during the start of this time period. It is encouraging though that he has been attempting more of those.

Reminder: Georgia Tech won a 55 yard field goal. And their playcalling demonstrated total faith in the kicker too. They ran the ball on 3rd down knowing that they would be able to get the kick off with around 4 seconds left. And yet, Clemson went years without attempting a field goal this long.

What if we look at Clemson’s makes and misses on the same style chart?

Looks like we have a bit of a block problem here?

I think the next natural question was, “is Clemson punting in position for a 45-60 yard field goal?”

This is not a ton of punts, and there does appear to be a downtrend (depending on how this season turns out), possibly. However, every single punt here is still an example of points that the Tigers did not compete for that would be in play with a good enough kicker and solid FG unit.

Of course, going back to the previous chart, can you fault Dabo for punting instead of risking another block? But as a head coach, you are responsible for the kickers and long snappers you recruit and develop.

Let’s compare some of the punters Clemson has had over these years:

I still haven’t decided if Teasdall is redeemed for that play in the ACC title game yet. Also, keep in mind that Bradley Pinion’s data here is biased because I only had one year. But I think overall, nothing to complain about here. Jack Smith has been doing a good job this year, and in general, it is hard to blame games on a few yards here and there on punts.

In general, I think Clemson’s kicking game needs to step up for the program to be competitive in coming years. It’s also not there’s some smoking gun that proves that special teams are bad enough that we need to be sharply critical of Dabo in this area.

I think the evidence is kind of mixed on my second hypothesis, so I’ll let you call it how you see it. I guess I am kind of failing to reject the null hypothesis by default.

Conclusions and Further Questions

Losing to Syracuse at home after a close loss to Georgia Tech was really shocking, and to me, this year’s team seems unimpressive. I think it’s right for Clemson fans to be concerned and angry.

Although correlation does not equal causation, I have tried to tie my discuss of the analysis with causal mechanisms. And it seems like Tony Elliott’s departure has had a significant impact on Clemson’s offense.

Dabo is a great coach, and I hope he understands that the alumni and fan base expect him to put a better product on the field. I think the data shows that Tony Elliott’s departure has had a significant impact on Clemson’s offense, and I hope that Dabo can find a way to get back to being “Wide Receiver U” again. (Ugh, maybe now I have do do a whole article proving that we aren’t Wide Receiver U anymore?)

I will probably do another article at some point attempting to look into QB development. Additionally, I think it’s worth studying the “Tony Elliot effect” more broadly too. And can we quantify the impact of other Clemson assistant coaches?

What do you think about this? What other questions came up as you were reading this that you think would be cool to investigate? Let me know in the comments or on Bluesky.

I hope you enjoyed reading!