How NFL Coaching Changes Are Quietly Wrecking Your Fantasy Team
Our AI model crunched six years of performance data to reveal how sideline shake-ups are hitting your favorite players.
It’s nothing new… Or is it?
Everybody always talks about the risks of a new head coach coming into a system, and inevitably its consequence on fantasy output. It’s far from a groundbreaking concept. A head coach tends to completely rewrite a team: from routes to philosophy, no head coach would have been hired if there wasn’t going to be change. And, naturally, there must be time to adjust. Players must learn to integrate themselves with the new ethos, and over time this collective adjustment builds chemistry. Dan Campbell’s Lions in 2017 hardly had a great year. He installed grit in the team which has come to pay off. That seems to be the year of the turnaround coach: Kyle Shanahan was 0-9 before switching to quarterback Jimmy Garoppolo. There was a learning curve there, because Shanahan is no ordinary coach. And this is how it can be. Of course, it’s easy to be inflicted by the availability heuristic here. It would be dismissive and thoughtless to consider Andy Reid’s first season as a Chief a once-in-a-lifetime outlier. A new coach doesn’t always spell trouble. Obviously. But how do we know which way a new coach will turn? Well, it’s helpful to get an idea of the quantifiable impact, and that’s exactly what I’ve done – with the help of my AI model trained on the past six years of NFL and fantasy football.
There’s also the question of the effect of broader coaching team changes. This is another thing that’s easy to mention without really knowing the real impact that it can have on players. Fortunately for you, I’ve dived into this as well. Not to the granular level, which obviously would yield little of interest (or maybe it would…), but I’ve factored in changes to offensive and defensive play callers (I’ve labelled them as such, as opposed to OC and DC, to maintain accuracy when a head coach is also designating the plays – which is what really matters here). With over 1,000 players analysed across the period of interest, I’ve found some pretty interesting stuff that’s backing up, and maybe also disproving, some of the comments that you hear flying around.
So, what’s the damage here?
It’s not ideal. Coaching changes do have a statistically significant, negative impact on your fantasy performance. Using pooled data of all positions, a change in head coach led to 1.12 fewer fantasy points per player, per game. That means each player could leave 15 points out there before fantasy playoffs even hit. Doesn’t sound like much? That’s over 150 points across the same time span, or nearly 12 points a week. It’s not going to turn back the clock on the Steelers of your fantasy league, but it could absolutely be the difference between silver and gold in January. The p-value on this is 0.0006. This is a serious, noticeable phenomenon. We should start treating it as such.
Starting at the heart
A change in head coach is like having a brain transplant. It’s up to us to analyse how it’s affecting the other major organ – the quarterback. Spoiler: the diagnosis is not pretty. This is as we might expect. Having major changes in the coaching staff, especially to the centrepiece – the head coach – creates a whole new system. And while the coaches lay out the system in theory, the quarterback is the player primarily responsible for executing that vision. As I mentioned before, it takes time for that system to become natural to a quarterback. This effect is highlighted in preseason camp. There are two standouts of this in recent years, namely Jared Goff at the start of Sean McVay’s tenure (somehow, we’ve gone back to 2017) and Tua Tagovailoa with a fresh-faced Mike McDaniels. Yes, we’re talking about preseason here. But does this spill over into the regular season? Absolutely, and I have the data to back it up. Yet, there is a point in the season when this pattern fades. The new head coach is new no longer. Normally by late November, the system is in full flow, and the fantasy upside is extremely attractive, particularly because people have preconceptions on quarterbacks who faltered for the first half of the season. This is where opportunity lies. Just as money is made in the stock market by finding something impactful that is not considered by the consensus, the fantasy goldmine lies in the waiver wire players that are ignored by others. During the regular season I’ll be releasing a piece on this – looking at when exactly quarterbacks, and the rest of the team, tend to start gelling with the sideline. This should help you get the drop on the rest of your league. And here’s another really important point. Because these quarterbacks regress to the mean in roughly the second half of the season, and there is a noticeable effect on the season-wide performance, that means that their first half is dragging them down significantly. This puts all the more emphasis on acknowledging this trend for the draft. Otherwise, it would be very easy to pick a breakout star with a new head coach, promising to send your fantasy points to the moon, the excitement of which just fizzles out when play starts. And as you trade or waive a new quarterback, it’s only afterwards that the dreamy points are reaped. Let’s try and avoid this scenario and get into the data that’s behind all of this.
As I’ve alluded to, the largest effect lies within head coach changes. With a head coach debutant, quarterback fantasy points per game (PPG) drop by nearly two. A 95% confidence interval yields confidence bounds of -3.01 and -0.91 PPG – so it’s pretty safe to assume the true value lies somewhere within this range. This is an incredibly large impact, and the fact that it is yet to be quantified is unbelievable. Ignoring all other positions that are affected by the changes, this alone could tip you from the losers’ to the winners’ bracket in playoffs. This isn’t just some daydream either – the p-value is 0.003. In other words, the probability of this trend showing up in the data but not actually existing is 0.3%. Both offensive and defensive play callers have no significant effect on quarterback fantasy points, however. That said, the power of those tests was 0.09 and 0.27, respectively. Power refers to the probability of being able to find the effect of a change, if it were really there. That means that it’s very difficult to know whether there really is a trend here. But, there’s no point trying to pretend like something’s there when we really have no idea. For the sake of your fantasy draft, all you need to know is the relationship between head coach and quarterback. I’ll just reiterate as well – this is a sweeping trend, so you’re likely to see this on average. If you’re steadfast that one quarterback will do particularly well, then go all in on it. A head coach change could be the very thing that takes a quarterback from average to great this season.
As a strong quarterback is a crucial foundation to a fantasy team, here’s a list of the head coach changes that have happened heading into the 2025 season:
- Chicago Bears: Caleb Williams (Ben Johnson)
- Dallas Cowboys: Dak Prescott (Brian Schottenheimer)
- Jacksonville Jaguars: Trevor Lawrence (Liam Coen)
- Las Vegas Raiders: Geno Smith (Pete Carroll)
- New England Patriots: Drake Maye (Mike Vrabel)
- New Orleans Saints: Tyler Shough (Kellen Moore)
- New York Jets: Justin Fields (Aaron Glenn)
Moving downfield
This section is probably going to come as a surprise to you. It certainly did to me. With such a significant effect on the quarterback, it’s natural to think that running backs and wide receivers are at least partially affected. Right?
Wrong. Across the two positions and the three coaching groups, every p-value was over 0.3. That’s far from anything meaningful. That is, except for head coaches’ effect on wide receivers. Compared to WRs without a change, those with a new head coach scored 0.72 fewer points per game (PPR), at a p-value of 0.096. So, also not any indication of a significant impact, but it’s definitely something I’ll be keeping an eye on as I add more data to our model.
It's pretty joyful news if you had your eyes on the likes of Brian Thomas Jr., CeeDee Lamb and Garrett Wilson, and also for the strong collection of RB2s and beyond held in the teams facing disruptions to the command centre. I hope this gives you confidence to home in on finding really great wide receivers and running backs to fill your team, as opposed to thinking too much about what’s going on around them. But still, why are we seeing this happen when the quarterbacks are so affected? It’s inevitably true that these downfield positions are still feeling the ripple effects caused by slightly slower play and inconsistencies from the quarterback. However, there’s nothing significant in the statistics. In honesty, this could be due to the sample size not being large enough. I had 228 RB seasons and 432 WR seasons in this regression, so pretty good, but maybe not enough to reveal what’s really going on. But this could make a lot of sense – it aligns with what I found with wide receivers having a more noticeable impact from head coach changes. If the quarterback is struggling, it will affect the wide receivers, just not as much. Especially in PPR, the points will still get racked up for catches even if the play isn’t impactful enough to satisfy the head coach’s playsheet or reap passing yards for the quarterback. So, this nearly-negligible effect might be happening, and that’s really ok. It’s easier to explain for running backs. They’re affected by the same logic of receptions with checkdowns and the like, and then they can also hoover up points from their runs. Although routes and plays will change with a new head coach, there’s less of a learning curve compared to quarterbacks, and this feeds through to more consistent fantasy performance.
Products of the playbook
I’m writing this article in the same order that I did the data, and after seeing the lack of effect on wide receivers and running backs, I thought that tight ends would have fallen under the same category. If you’re thinking the same, don’t worry – there’s a twist. You can see below straight away that there’s a noticeable impact across the board of coaching changes. The mean difference is -1.36, -1.59 and -1.28 PPR fantasy points per game for changes in head coach, offensive play caller and defensive play caller, respectively. Similarly, the p-values for these are 0.022, 0.002 and 0.019. So, we can be surest that tight ends are affected by changes to offensive play callers, but they are impacted by other staffing changes as well, and in a pretty major way.
To keep things interesting, I’ve opted to show you the regression model this time. If you’re not familiar, that’s alright – I’ll explain what everything means. The predicted fantasy points are based on whether there were new coaching staff, and which in particular, and the season the data was taken from to act as a control variable.
A perfect model would have an exactly diagonal line (if the axis were to the same scale), and clearly we don’t have that. Unsurprisingly, there is more to tight end production than their coach and what year it is. However, the important number is the R-squared, shown by the trendline. It shows the percentage of the variation in fantasy points that happen because of the variables included in the model. Our model explains 12.4% of the variation in fantasy points for tight ends and given that most of this explanatory power is driven by coaching staff changes, you can picture the scale we’re working with here on an output level. Not life-changing, but definitely there, and definitely enough to make a difference for your fantasy league.
It's easier to infer the reason behind the effect of head coach and offensive play caller changes to tight end value. This is ultimately down to their dual capability of acting as either an offensive lineman or another wide receiver – different coaches will want to utilise this in different ways, and it can be quite complicated to get used to a new system for a player which has several different jobs at once. The really interesting part is the impact of new defensive play callers. It’s not possible to know exactly why they are affecting tight ends at all. One theory is that they’re a proxy for instability and change-ups within the team, that can lead to new team philosophies and strategies. Of course, new defensive play callers could signal poorer defences, and a different playing style when the ball is in possession. Another factor at play is that the sample size for tight ends was only 103 – so there’s not the same depth in the sample as there was for other players. Regardless of the reason, it’s important to keep it in the back of your mind when looking at tight end options ahead of the draft.
The AI era of fantasy football
I wanted to round up my first ever piece by saying thank you for reading, and I hope to have convinced you to rethink how you look at the fantasy season. This newsletter isn’t something that I hope to replace the vibes of a fantasy draft – I think it’s a shame to turn your fantasy league into something mechanical and emotionless – but I thought the findings of my AI model were pretty interesting and could steer you in the right direction. There are far more useful bits of information that it’s coming out with, and I’ll be writing about all of them as we approach the fantasy draft in late August.
Each of our longer thought pieces, like this one, will come out on a recurring basis. In between, I’ll be doing spotlights on players or teams that are particularly interesting based on the latest discussion. From this piece, you can expect coming analysis on the Jags, Raiders and more.
I look forward to working with your fantasy team.
See you next time,
Valentino