Belichick – pattern recognition

Supported by Greenhaven Road Capital, finding value off the beaten path.

It’s Super Bowl week and what better way to celebrate that dig into the thinking behind one of the coaches (and general manager) at the game? The posts this week come from this short book.

In my readings there were five big ideas that stood out:

1. Inversion
2. Pattern recognition
3. Top-down support
4. Argue well
5. Seek counterfactuals

This post addresses the second point; pattern recognition.

…

Pattern recognition is a superpower because it saves you resources (time, energy, money). In the same way an outdoorsman knows where to fish, an accountant knows when things smell fishy, or when a chef knows how to fillet a fish, we can have our own superpower, our own pattern recognition. Belichick has Ernie Adams.

Mr. Ernie Adams, Belichick’s Belichick

Ernie Adams is the (nearly) silent partner.  David Halberstam writes:

“Ernie Adams was Belichick’s Belichick, the film master’s master of film. He was supremely knowledgeable about the history of the game – no play was ever forgotten, and his brain was like a little football computer, always clicking away, remembering which defense had stopped which offense, and who the coaches and the players had been.”9

Adams has superb pattern recognition. He’s so valuable to Belichick that in the bank of phones along the sideline there’s a direct line to Adams. Adams is the one who will report whether or not to challenge a play and Belichick will ask during the game “What have we got, Ernie?”

Before Super Bowl 49 against the Seattle Seahawks, the Patriots practiced plays they thought their opponent would run. The person who decided which plays to practice was Ernie Adams. “Ernie would diagram plays that he thought they would run against us or that he had seen from a previous game,” Belichick said. A good thing too.

In the final thirty seconds, with the Seahawks on the Patriots’ one-yard line, quarterback Russell Wilson threw an interception that turned near-defeat into victory for the Patriots. What’s amazing about that play was the practice that preceded it. In the NFL Network film, Do Your Job, the Patriots coaching staff and film narrator talk about how the Patriots practiced the exact play the Seahawks ran.  Here’s the dialogue from the movie.

Coach 1: “What’s funny, if you look at the play is that (Brandon) Browner is telling Malcolm (Butler) what he’s going to do. I’m going to jam this guy and then told Malcolm to just go.”

Coach 2: “If you remember, Malcolm’s matchup was really Kearse, but Browner had seen enough of those pick routes. He anticipated what was coming as well as Malcolm. So he was like, let me get up here and jam the point and you go. Malcolm had seen it in practice.”

Narrator: “Butler had seen it in practice, he just didn’t stop it.”

During this narration there are nearly identical videos of the game and practice footage where you see the before and after. In practice Malcolm gets picked and the receiver catches the ball. In the game Malcolm gets the pick. Here’s what Ernie Adams says in the video:

“You’re going to win or lose games at practice. There’s no such thing as being a game day player. You see situations come up on the practice field, you’ve worked on it, you know what it takes. When it comes up in the game, because you’re trained, you’re seasoned, you’ve seen it – you react and make the play.”

That’s good pattern recognition. At the final moments of the Super Bowl, the win is decided because one team has built pattern recognition into their strategy.

Stocks and submarines

Good in-the-moment decisions come from experience in consistent situations and learning from them. Football is great for this because it’s consistent. It’s the same number of players, the same rules, and the same dimensions. The more nebulous things in life require more work.

Warren Buffett uses pattern recognition to figure out what stocks to buy and when. At the 2016 Berkshire meeting, Buffett said, “pattern recognition gets very important in evaluating humans and businesses. Pattern recognition isn’t 100%, but there are certain things in businesses we’ve seen over and over.”

In an interview with Charlie Rose, Buffett said, “I’ve been reading IBM’s annual report every year for 50 years. This year I saw something that sort of clicked.” Fifty years! That’s how long it took for Buffett to recognize something valuable. It’s what Ernie Adams does too. Adams works 100 hours a week to build up pattern recognition.

Phil Simms played under Belichick and Adams for the Giants and said Adams “is a great source of information for Bill. There’s nothing like somebody that can stand back and get a different view of what’s going on.’’ To put it another way, Adams sees the patterns than Belichick missed.

Belichick learned to invert when he coached offense and defense. Adams is another point of view, another source for triangulation, another pattern recognizer. Pattern recognition can come from any situation with repetition. Even from your youth.

As a US Navy Intelligence Officer James Bradley was trying to figure out a mission for the spy submarine the USS Halibut. Bradley wanted the crew to listen to Russian Navy communications, but do so without being detected. He guessed there was a cable that “ran from the Soviet Union’s missile submarine base at Petropavlovsk, under the Sea of Okhotsk, and then on to join land cables going to Pacific Fleet headquarters near Vladivostok and then to Moscow.”12

He just didn’t know where.

The Sea of Okhotsk is big, 611,000 square miles big. Twice the size of Texas big. The Halibut had an underwater camera, but the crew couldn’t wander back and forth. That would take too long and be too dangerous. Bradley needed to figure out something else.

This is from Blind Man’s Bluff:

“Bradley cleared his mind of charts and maps, freed himself from official assessments, from the meetings, memos, and briefings that swamped the business of intelligence in Washington. He let his eyes close and his thoughts wandered into simpler journeys taken and simpler times, before the Cold War, before World War II, back to the waters of his childhood.

“There he found an answer that was beguiling simple and just strange enough to be true. It was buried in his memories of St. Louis in the 1930’s when he was a boy and his mother packed him up to escape the summer’s heat on river boat rides along the Mississippi River.

“…Young Bradley had taken to passing time with a steamer captains in the pilothouse, and from there he could see a series of black-and-white signs placed discreetly along the shore. Most of the signs marked mileage and location.

“But they were a few, he remembered now, they declared: ‘Cable Crossing. Do not anchor.’ These signs were there to keep some idiot in a boat from snaring and snapping a phone or utility cable in the shallows. Bradley’s eyes snapped open as he realized that what was true of the Mississippi just might be true of Okhotsk. That’s how they would find the cable, he thought. That’s how they would engineer one of the most daring acts of tele-piracy of the Cold War. Halibut would be led directly to her quarry by signs placed somewhere on a lonesome beach in the Soviet Union declaring: ‘Watch Out! Cable Here.’”  

That’s exactly what happened. Commander John McNish would guide the Halibut into Soviet waters and find the cable thanks to a sign on the beach.  Notice the similar language; see something you’ve seen before, practice something, look back in time. Good pattern recognition is about memories and experiences, it’s a tool anyone can carry. You just have to read or do.

Charlie Munger is described as a book with legs. The winningest college soccer coach, Anson Dorrance reads all the time too.  Pete Carroll read Grit by Angela Duckworth and The Obstacle is the Way by Ryan Holiday and invited both to talk to his team. Shane Parrish says that his popular Farnam Street blog is his attempt to “master the best of what other people have already figured out.” Other people’s experiences are your pattern recognition data.

You can also do to build pattern recognition skills. Elizabeth Gilbert thought about going to graduate school, but tuition was too high. Instead, thought Gilbert, why don’t I go out and live and write different things. “You can learn about the thing by learning about the thing,” Gilbert said, “or you can learn about the thing by doing the thing.”  NASA learns by doing; running simulation after simulation. The last simulation before the Apollo 11 landing included an error that the crew muffed during training, but got it right during the lunar landing. That’s good pattern recognition.

Note that pattern recognition has to happen in consistent environments. Randomness conceals accurate results. Sports are remarkably consistent, as the famous scene in Hoosiers shows, the basketball rims anywhere are the same height as those at Hickory.

In areas with more randomness it’s harder to build pattern recognition. Investors see this. As great as Buffett and Munger are, they aren’t perfect. That’s because things are beyond their control. Ditto for NASA. The simulations were good, but not perfect because of randomness and unknown variables.

…

Tomorrow we’ll look at #3, Top down support.