The Cleveland Browns are still the most analytically advanced organization in the NFL, according to an ESPN poll of analytics staffers around the league.
The Browns took the mantle from the Baltimore Ravens in 2021 and kept it this season after receiving a plurality of votes, though their margin is smaller than a year ago. The Browns were voted as the most analytically advanced team by staffers from 10 of 21 teams this year after receiving 17 of 22 votes in 2021.
This is the third year we've surveyed NFL teams' analytics groups to better understand the landscape of quantitative analysis in the league and get their opinions on questions facing teams. Some questions have remained consistent, and some are new. (Check out the 2021 and 2020 versions.)
The survey was sent to a member of each team's analytics group, and 21 responded. Survey recipients were permitted to collaborate with other members of their analytics group as long as they submitted only one ballot per team. Some staffers left comments, and ESPN called others for contextual follow-ups. Participants were allowed to weigh in on information about their own team in some, but not all, of the questions (designated below). All were granted anonymity so they could speak freely, and although there were 21 responses, some abstained from some questions.
Survey responses took place between August and November.
Which NFL team is the most analytically advanced?
Voting for your own team was permitted for these first three questions.
Which team produces the highest level of analytics work?
1. Cleveland Browns (7)
2. Baltimore Ravens (4)
3. Philadelphia Eagles (3)
T-4. Atlanta Falcons (1)
T-4. Buffalo Bills (1)
T-4. Dallas Cowboys (1)
T-4. Houston Texans (1)
T-4. Jacksonville Jaguars (1)
T-4. Los Angeles Rams (1)
T-4. Miami Dolphins (1)
Which team most incorporates analytics into its decision-making?
1. Cleveland Browns (9)
2. Philadelphia Eagles (3)
T-3. Baltimore Ravens (2)
T-3. Indianapolis Colts (2)
T-5. Arizona Cardinals (1)
T-5. Atlanta Falcons (1)
T-5. Buffalo Bills (1)
T-5. Los Angeles Chargers (1)
T-5. Minnesota Vikings (1)
The Browns swept the three categories again. The most analytically advanced question is framed as a combination of the other two: the level of analytics work a team produces and how well it integrates its quantitative analysis into the team's decision-making.
"From what I understand, all their decision-making takes a scouting or coaching perspective and then marries that with some kind of quantitative analysis or quantitative investigation," an NFC analytics staffer said.
"[The Browns] have an analytically minded GM -- they have leadership all the way, director level and above, who is quantitatively inclined," an AFC staffer said, referencing Cleveland general manager Andrew Berry, who came up in scouting but is believed to embrace a data-centric approach.
Which teams are among the five most analytically inclined?
Baltimore Ravens (21), Cleveland Browns (19), Philadelphia Eagles (16), Buffalo Bills (8), Minnesota Vikings (7), Los Angeles Chargers (5), Los Angeles Rams (4), Denver Broncos (3), Indianapolis Colts (3), New York Giants (3), San Francisco 49ers (3), Dallas Cowboys (2), Green Bay Packers (2), Houston Texans (2), Miami Dolphins (2), Arizona Cardinals (1), Atlanta Falcons (1), Chicago Bears (1), Jacksonville Jaguars (1), Kansas City Chiefs (1), Las Vegas Raiders (1)
One voter selected six teams.
After the Browns, we see a clear tier of two right behind them -- the Ravens and Eagles, as has been the case the past few seasons. Both teams have large analytics staffs.
One analytics consultant highlighted the Ravens (along with the Colts) as leaders in the coaching side of analytics and cited offensive assistant Daniel Stern -- a former football research coach who assists with game management -- being part of the coaching staff as evidence.
The Ravens appear to be one of the most analytically inclined game-management teams under John Harbaugh.
"You continuously see Baltimore make analytically sound decisions," one survey taker wrote.
One respondent cited organizations or leadership that highlight their quantitative analysts, calling out the Ravens and Rams specifically.
"If the team is willing to give a public shoutout to their [analytics] team, they really do value their people," they said.
The same staffer said the Ravens are usually aligned with consensus opinion in the draft, another indicator of their quantitative-leaning mindset.
The Eagles are annually thought of as one of the most analytically inclined organizations. They received notoriety for their 2017 Super Bowl run, which was in part fueled by analytically driven game-management decisions, but the reputation has sustained since the Doug Pederson era.
That same staffer thought the promotions of Alec Halaby and Jon Ferrari to assistant GM were indicators of a data-driven approach in Philadelphia. Halaby is a longtime Eagles analytics staffer and was vice president of football operations and strategy before becoming assistant GM. While Ferrari doesn't have the same analytics titles, the staffer in question said, "When I think analytics, that's just evidence-based decision-making, right? [Ferrari's] not necessarily considered an 'analytics' guy, but he's very, very sharp."
The Bills received the fourth-most top-five votes like last year, but the Vikings and Chargers both moved up. The reasons are fairly obvious: Chargers coach Brandon Staley was far more inclined to go for it on fourth down than most coaches in 2021, though he has significantly slowed down in 2022. And the Vikings hired Kwesi Adofo-Mensah, the first GM to have experience as an analytics staffer.
"Incorporation of analytics lags behind work quality, so Minnesota has a big advantage just because they are so bought in," wrote one survey taker.
Which NFL team is the least analytically advanced?
One voter abstained.
In this category, staffers often vote based on what they don't hear.
"I don't know much about [the Commanders and Titans]," one staffer said. Which is sort of the point -- analytics staffers often know each other and talk, the same way coaches know coaches and scouts know scouts.
That Tennessee and Washington have small analytics groups -- just one staffer each, to the best of my knowledge -- and their work isn't well known to their peers is a negative indicator. It doesn't rule out heavier quantitative involvement that isn't known to the outside, but when asked which teams are further behind from a data-analysis standpoint, those two teams are consistently brought up.
Does momentum exist within an NFL game?
One voter abstained.
Analytically inclined people have often dismissed the idea of momentum, so it was surprising to see three-quarters of respondents believe in its existence.
"I am aware that I have no way to prove it," said a veteran analytics staffer who voted yes. "I've built the win probability models, when you look back at how you got to a spot it doesn't matter at all, I'm aware of all that. It's kind of like me taking off my statistician hat, putting on my football hat, I feel like you can feel it.
"Just because we can't measure it, or we aren't capable of measuring it, doesn't mean it doesn't exist."
Several people noted that they felt that while momentum might be real, the effect is ultimately small.
"Momentum and hot hands do exist, but you can make money by betting against them," wrote one survey taker. "So my answer isn't literally no, it can even affect certain players differently. But yes, momentum is still overrated and used retroactively as part of a narrative where it sounds good but ultimately is hand waving."
If momentum does exist, the question becomes: Does it or should it affect decision-making within a game?
"Yes, but my job is to not treat it as if it exists," said one staffer, who is on the coaching headset during games.
Another, who also felt momentum is real, felt differently. "Is it something tangibly usable? I don't think so."
A third felt that momentum exists and should affect decision-making but is often misused by coaches. They gave an example of a team moving the ball well and reaching field goal range, with the defense on its heels and getting tired.
"Take advantage of that edge!" the staffer said. "But it's more like they kill their own momentum by saying we're in field goal range, no negative yards, don't take any big sacks, throw the ball away. You end up with a midrange field goal, where you should have pressed the accelerator and scored."
Which position is the most difficult to evaluate quantitatively?
Offensive line (5)
One voter abstained.
Defensive backs dominated this question, with safety taking the top spot.
"A lot of what you're actually trying to measure with safeties are things that don't happen as a result of them," an AFC staffer said. "It's tough to measure a counterfactual.
"Scouts will say the same thing. There's so many plays at safety where nothing happens and they have minimal impact."
One survey taker who selected offensive line felt there was nothing truly objective to measure individual linemen.
"People are trying now to get better at the objective with player-tracking data, but I don't think that's there yet," the staffer said, noting that how one measures an O-lineman's performance using that tracking data was subjective.
Both survey takers who selected quarterback felt like the context around the position was crucial in its evaluation, making the process more difficult.
"The value that a quarterback contributes to a game is so heavily dependent on the other 21 players on the field," an NFC staffer said.
"You have the most information on quarterback, but it's hard to isolate the skill of the quarterback outside of the system he's operating in," the other quarterback voter added, also noting that the hit rate on QBs is lower than other positions, as well.
Which position is the easiest to evaluate quantitatively?
Running back (6)
Wide receiver (4)
One voter abstained.
Edge rushers and running backs led the way in terms of ease to evaluate. As one staffer put it regarding edge defenders -- their responsibilities are very clear.
"They're rushing the passer, so you can evaluate if they are creating pressure or beating blocks and in the run game, it's kind of the same thing: Are they getting off blocks, are they getting to the ball? For most edge players, that's it," they said.
An NFC staffer who selected wide receiver noted receivers' performance with the ball or when receiving targets is fairly easy to measure but added there is room for development when quantifying receivers' off-ball work.
Does your team have an analytics staffer on a coaching headset during games?
This is a slightly lower ratio than last year (when the split was 15-6), which could be a result of changing regimes but more likely is because different teams responded this year compared to last.
As a follow-up, I asked: What difference does it make having an analytics staffer on headset to assist with game management, as opposed to someone else?
"You want to get as much of the game-management stuff [prepared] pregame as possible," said one survey taker. "But every game there's going to be situations that come up that we didn't have a pregame plan for. And just having someone that can mentally think analytically at that moment has a lot of value."
"That discourse between coaches and analytics staffer is more efficient if your analytics staffer [is the one] communicating information they know most intimately," another added.
A third felt it doesn't have to be an analytics staffer on headset, but at least someone deeply familiar with the quantitative side of game management.
"Having a coach who has a full understanding of the analytics is just as good, but at that point I think we're just messing around with titles," they said.
When your team makes a decision you disagree with, which area is that most likely to occur in?
Positional value (7)
Trade value (3)
Game strategy (e.g., playcalling, scheme) (3)
Pro player evaluation (3)
College player evaluation (1)
Game management (e.g. fourth down, 2-point) (0)
Four voters abstained.
One voter who selected positional value was not surprised to hear it was the most common response. But they couldn't quite put their finger on why, and said that was kind of the point. If they understood why they weren't able to align with decision-makers on positional value, it probably wouldn't be a subject of frequent disagreement.
Running backs are the most obvious example of a position-value disconnect.
"The whole idea of the first-round running back, the big contract to the running back. People see the production coming from the running backs and give it to the back when he might not deserve the credit," they said.
Though running backs' market values have dropped across the league, quantitative analysts still see the position as overvalued.
One survey taker who selected game strategy said it was hard making inroads with the coaching staff.
"We'll get more buy-in from our personnel department and our front office as far as being open-minded to research that we can provide," they said.
Another staffer who chose pro personnel evaluation said players at the bottom of the roster were harder to evaluate.
"I might not agree [with the decision], but I have enough respect to realize I'm wrong a lot when it comes to pro players, especially down on the roster -- 45th through 53rd man I don't have any data on those players playing in the NFL."
Name a player you believe to be generally underrated, based on your quantitatively informed opinion.
Excluding players from your own team.
Vikings QB Kirk Cousins (2), Seahawks WR Tyler Lockett (2), Titans DT Jeffery Simmons (2), Cowboys P Bryan Anger, Patriots WR Kendrick Bourne, Chargers CB Bryce Callahan, Jets TE Tyler Conklin, Bills OT Dion Dawkins, Broncos WR Tim Patrick, Vikings DE Harrison Phillips, Lions WR Amon-Ra St. Brown, Falcons CB A.J. Terrell, Seahawks CB Tariq Woolen
Five voters abstained.
Despite the plethora of voting options, it was notable that three players received more than one selection. In this section more than any, it's important to note that the responses came in over a large range of time -- from before the season to early November.
"Cousins probably isn't the most undervalued controlling for his position, but he probably generates the most value compared to how much people think he does. He's easily a well-above-average QB," one survey taker wrote before the season.
"Jeffery Simmons is proving it now too, he's proving all those guys right. That's a good one," said one voter upon hearing the overall results. "Tyler Lockett, he gets so much credit from the analytics community that I feel like he's where he should be, but I get it from the non-analytics community; that makes sense."
Simmons and Lockett are putting up serious numbers this season. Simmons has 6.5 sacks on the year and ranks third in run stop win rate. Lockett, entering Week 13, ranked fourth among all wide receivers and tight ends in Receiver Tracking Metrics' overall score.
Multiple survey takers mentioned how surprised they were that Callahan didn't generate more interest in the offseason, when he signed a one-year, $1.3 million deal with the Chargers.
"The last couple of years his coverage grades, his separation, his tackling have all been really good. I think the nickelback is still a bit undervalued," said the staffer that voted for him.
"No one seems interested in him even though he seems to always play well? He also gets hurt a lot," said another.
Name a player you believe to be generally overrated, based on your quantitatively informed opinion.
Excluding players from your own team.
Cowboys CB Trevon Diggs (2), Broncos QB Russell Wilson (2), Raiders QB Derek Carr, Cardinals RB James Conner, Cowboys RB Ezekiel Elliott, Steelers S Minkah Fitzpatrick, Titans RB Derrick Henry, Chargers S Derwin James Jr., Texans QB Davis Mills, Colts G Quenton Nelson, Steelers OLB T.J. Watt, Commanders DE Chase Young
Six voters abstained.
Perhaps it's unsurprising to see Diggs receive multiple votes in this category, considering his value has been the subject of significant debate after his notable 2021 campaign in which he recorded 11 picks but the rest of his game was questioned. That was the crux of the response from one survey taker, who said picks can certainly make up for allowed yards but that those interceptions last year inflated perception of his performance.
As for Wilson, he received one vote before the season and one well into his disappointing 2022 campaign. Wilson has a QBR of 35.0 this season, 28th out of 31 qualifying quarterbacks.
The staffer who selected Fitzpatrick compared him to Seahawks safety Jamal Adams.
"He's better than Adams but I feel like Fitzpatrick is just very chaotic. He has a lot of great flash plays, but I think he's out of position and making mistakes too much. I think his grades are always ... a lot lower than what the scouts are saying about him," a staffer said, referencing their team's in-house quantitative grades.
The voter who selected Young referenced Browns edge rusher Jadeveon Clowney as a comparison.
"Incredible athlete, he just hasn't produced to match his ability yet. I've always had a problem with people getting credit [and being] considered great players before they actually produce and prove they are great players," they said.
Upon hearing some of the other results, one staffer was stunned at one name.
"Wow, Derwin James?! My goodness," they said. "I like Derwin James. The rest make sense to me, Derwin James is a surprise."