In just a few seconds, it was over.
Seattle Seahawks quarterback Sam Darnold collected the shotgun snap, and waited … waited … for tight end Eric Saubert to chip his defender before shedding him and leaking wide open across the middle of the end zone. Darnold fired the ball to Saubert; Los Angeles Rams defensive tackle Kobie Turner nearly got a hand on it. He didn't. Saubert caught the pass as the stadium exploded, throwing his arms out in celebration with the ball in one hand. Game.
The walk-off 2-point conversion -- in Week 16, in overtime, against a division rival -- lasted only a moment, but it was in many ways a product of months of conversations and situational and schematic investment that reflected Seattle head coach Mike Macdonald's mantra: "chasing edges."
"The nerdy definition of it," he said, laughing, while speaking to The Athletic recently, "we call it 'the relentless pursuit of the hidden advantages that bring us closer to being the team and people that we want to become.'"
Macdonald, the third-year head coach who led Seattle to a Super Bowl win last season, has a defensive-oriented mind in a sport where much is dictated to defense, by offense. But Macdonald has a knack for inverting that norm. His defensive scheme is built to neutralize offensive personnel changes and the corresponding spatial manipulation that typically forces defenses out of their preferred structure.
Outside of scoring on defensive takeaways, Macdonald can't necessarily dictate how many points his team gets. So he obsesses over other margins: how aggressive Seattle's special teams unit will be (No. 3 in EPA last season); how he hires offensive minds out of established, competing systems (both his current and former offensive coordinator previously worked for the rival 49ers); how the Seahawks structure practices to prepare for close games; the technologies into which the team invests and how they're applied; the way the coaching staff discusses analytics and situational management in the same breath as their players' fundamentally sound, physically punishing style. Macdonald likes to say that Seattle is all about combining old school principles with new school methods.
Every NFL team is so good, Macdonald said, that "what it takes to win is the accumulation of these small advantages, small things that you can find to set your team apart."
In Seattle, "chasing edges" the way Macdonald intends requires investment into a unique combination of people, processes and technology.
Brian Eayrs' desk at the Seahawks' practice facility is crowded with four screens. They all feature something different; cutups of plays, graphs he's building or code for the analytical models he builds. Behind those screens is a larger one on which he can project any of this to whoever stops into his office. Macdonald does, often. Eayrs also always has open a digital dashboard that keeps him updated on league happenings -- he calls it "watching radar," a nod to the movie "Spaceballs" -- and an Excel sheet he calls his "workbook" where he logs ideas and notes to himself about upcoming opponents as he studies them. The "Hot 100" playlist Eayrs builds and tweaks throughout the year is probably playing in the background. He has a rule: If he puts a song on the list, he must remove one, too, so it's always exactly at 100.
Eayrs is Seattle's director of football analysis and special situations -- a role that has evolved over his 14 seasons with the franchise to focus on team tactics, opponent film and tendency studies, leaguewide schematic trend studies and coaching strategies. (The Seahawks have multiple analyst groups across their strength and conditioning, health and player performance, scouting and talent acquisition and salary cap departments.) In short: Eayrs' job is to literally find and apply these "edges" Macdonald seeks.
"It's a little bit coaching, it's a little bit analytics, it's a little bit dealing with the players (and) it's a lot dealing with the coaches," Eayrs told The Athletic. "(And) a key function of my job is overviewing the NFL to everybody, all the time."
A significant area of focus for Eayrs is situational analysis and in-game decision-making. Macdonald made this a priority when he became the Seahawks' head coach in 2024.
"(He) really created a lot of time and space for this concept, because we're working with the understanding that there's gonna be a lot of close games," Eayrs said. Seattle's coaches game plan as though they will face 10 to 13 one-score games in the fourth quarter in a season (a little more than half of all regular-season games were decided by one score in 2025).
"We kind of work backwards. We say, 'well, if that's gonna be close, how soon do we need to start talking about how we're gonna finish the end of the first half and finish the end of the game?" Eayrs said. "For our head coach, that's probably earlier than anybody in the NFL as part of the weekly process."
NFL teams typically build their game plans on Tuesdays, and get them to players by Wednesday morning, when the first full meetings, practices and walkthroughs take place. Eayrs and Macdonald have meetings as early as Tuesday evening to map out what Eayrs refers to as "critical variables that are going to impact the coaching decisions in the game." These include tendency and situational research on that week's opponent, plus examples from around the league and in college football of how teams made decisions in key scenarios.
"He's so passionate about it, he's so good at it," Macdonald said. "Some analytics guys, they kind of just live in their brain. Brian … he can explain (and) simplify things in a way that makes it functional for us in terms of how we operate and how we make decisions."
Instead of waiting to hold separate situational meetings later in the week (some teams wait as late as Friday), Seahawks assistant coaches build the information they receive Tuesday into their installations, walkthroughs and practices on Wednesday and Thursday. Eayrs said that they go over everything again with coordinators and players on Friday and Saturday, including tape on situational strategy used in practice.
Between this film study and his opposition research for the schemes and tendencies that help Macdonald and the coaching staff build a game plan, Eayrs has to process and organize an enormous amount of information -- and make it teachable to coaches, who then do the same for players.
"The living hell that I'm in is that every single football game played (that) has a clock, I have to be watching it. And I'm not gonna miss it," he said, smiling. "Every single game is an opportunity to think this through."
In Macdonald, Eayrs has not just found a thought partner in prioritizing strategy and analytics, but also someone who shares his belief that technology can help in the pursuit of small advantages.
Last offseason, the NFL expanded its partnership with Microsoft, the technology and software company co-founded by the late Paul Allen (who owned the Seahawks until his estate sold the team to tech billionaire Vinod Khosla and his family and investor group in August). As a result of this expansion, some teams -- including the Seahawks -- are more deeply integrating AI tools like Copilot into their day-to-day operations. (Some teams use other AI software as well.)
Eayrs uses this, alongside other technologies, to help organize his buckets of film and other information, sort new and old data, and run models for any scenario Seattle may face. Meanwhile, Macdonald has specific ideas for how he eventually wants Copilot to be a regular part of the team's workflow. The Seahawks, who are a partner of Microsoft AI, are already testing some of those ideas. For example, Macdonald records all of his meetings and uses the tool to package specific notes and highlights from the recordings so he can be clearer in his delivery and ensure players and coaches are receiving consistent messaging week over week.
More time, for a football coach, can become a massive advantage during a season in which one day of strategy and planning blends into another until suddenly, it's January. Macdonald came up through the coaching ranks as a college graduate assistant and Baltimore Ravens coaching intern, when he was often relegated to grunt work such as logging and filing statistics, play charts and other time-consuming tasks. Such staffers around the NFL who use AI and other digital programs now run hundreds if not thousands of queries to get to answers that he remembers used to take him days to find.
"You might get to the same answer, but if you get to it three days earlier, that's super valuable," Macdonald said. "Now you can rep those things and you can give those (players) extra experience and extra looks going into the week."
Could AI learn his defensive system? Could it understand how the scheme needs to be taught? More importantly, could it learn others' schemes well enough that it could help Macdonald and his coaches in their game planning? These are questions Macdonald wants answered.
"I think one of the things we've learned in the AI space is you just have to go. You have to dive in headfirst," he said. "There's going to be (bad) queries. Sometimes you're gonna have bad data in, bad data out. … This is a process that we're gonna adopt and buy into, and really be able to build this thing for the long haul, and that includes revising how we're going to drive the AI in terms of the data we're looking for."
Macdonald, said Eayrs, has been very clear with the organization that he wants to be at the forefront of the league in applying newer technologies to football and constantly evaluating the team's processes. The head coach admitted to a bit of an obsession with improvement.
"That was something that always bothered me about my experiences coming up," Macdonald said. "'Why is there so much resistance to trying -- why are we not trying to find the absolute best way to do things?' It still drives me crazy to this day. How can you live with yourself if you're just squeaking by?
"A cuss word around here is, 'well, this is the way we've always done it.'"
The strategy of a football game is often built out of a long string of ideas, attempts and errors. Macdonald believes that any worthy process should come with the risk of error; that a person should accept this risk because any error will probably lead to a new answer. Football coaches, he said, always want to be right. But the process of first being wrong, then becoming right, is important. It informs his willingness to try new things. It also informs his mantra for his team.
"There's more energy behind that. There's more excitement. There's more unknown," he said. "When we talk about 'chasing edges,' you're implying there is an edge and sometimes you'll go over (it). Part of our culture is that we try to make it a space where if you do go over, that's OK. If our intent is right, we are able to learn from it and move forward.
"That's where you make those marginal gains. That's what we're trying to create."
On game day, Eayrs is up in the coaches box and on the headset with Macdonald and assistant head coach Leslie Frazier. Typically, he'll have play sheets and computer monitors with him, including one that features Microsoft's game analytics dashboard.
Every team has equal access to this dashboard, which digitally charts the game's action for them in what looks like an Excel spreadsheet and helps assistant coaches and analysts compile analytics and build situational charts based on what their head coaches and coordinators need. Staff used to have to chart all of this by hand.
"You would have a sheet, and like the first seven to eight columns are just descriptives of the game that you can see on the scoreboard," Eayrs said. "… Somebody had to manually keep writing that down as fast as they could."
This season, analysts like Eayrs can send the models, charts and other data they collect in real time from their screens directly to those used by the coaches on the sideline or in the other booth. It's called the Sideline Viewing System. Previously, if a team had a situational expert in the booth, he had to verbally relay this information to coaches through the headset.
"If you think back to the old days," said Eayrs, "somebody in the box would have to say, 'hey, I've seen 13 of these, five of these and four of these.' They'd go, 'wait, what was that first number again?' … Now we don't have to do that. … We can just stream all of that down to the sideline and say, 'see your work sheet? That's how many times that thing has happened that you've asked about."
Eayrs' role during a game is to anticipate what decisions Macdonald and Frazier will have to make, a few plays before they have to make them. "I only deal in matters of fact on game day," he said. "I don't involve what I think we should do or what I feel we should do. That's really the intuition and the art of what our coaches bring to the table. … It (is) a matter of saying, 'hey, this thing might happen three to four plays from now. So be thinking through your decision if it goes this way, or that way.'"
Sometimes Macdonald goes with the analytics, and sometimes he doesn't.
But none were needed to make the decision to go for two points after Seattle's overtime touchdown against the Rams last season. Every person in the stadium knew that converting would mean winning the game. Kicking an extra point would tie it; not converting the 2-point play would mean losing.
What the crowd did not know was how much discussion Macdonald, Eayrs and the rest of the staff had throughout the week about the importance of Seattle winning that game -- one they believed would be close -- and the potential that they would need to do so via one or two pivotal, possibly unusual plays. Though Seattle had not run a 2-point conversion play in a game that season, they talked through the possibility. Offensive coordinator Klint Kubiak had a bucket of 2-point plays he was itching to use and had even rattled off "four or five" of them to Eayrs and Macdonald during their game planning meetings. (They ran three in that Week 16 game; all were successful, including the now-infamous fourth-quarter call reversal during which Zach Charbonnet picked up what turned out to be a live ball in the end zone. No man or machine could have predicted that one.)
"The value was so much higher in winning than tying," said Eayrs, "that we had brought it up before the game, (that) this is a game that the value of tying really is not good for us. We'd much rather do something and try to go for the win." As the game progressed, Eayrs tracked the numbers that suggested an ever-increasing probability that it would go into overtime. Everyone, he believed, was prepared -- because of all of the work done in the margins beforehand.
But as Eayrs watched Darnold and the Seahawks offense drive down the field in response to a Rams touchdown on the opening possession of overtime, he thought about how much better all of that planning sounded in their meetings than it felt in real life. Suddenly, the man responsible for watching every single snap of every football game, who meticulously plots out every possible game scenario every season, could hardly watch the one in front of him.
This article originally appeared in The Athletic.
