Should We Stop Going For It On 4th Down?

[bartman83642]

I found something but only after searching through Popper’s Tweets.

Thanks. I was curious at what the percentage would look like when down 10, 2 scores with limited time left.

[Rugger05]

Week 15 began with the Los Angeles Chargers losing to the Kansas City Chiefs on Thursday night in a wildly entertaining game between two of the game’s best quarterbacks in Patrick Mahomes and Justin Herbert.

Afterward, something strange happened. The players’ performances took a back seat to the coaching, and specifically the decisions that Brandon Staley made on multiple fourth downs. The nerds yelled at the football guys for not understanding the math. The football guys yelled at the nerds for, well, being nerds.

The arguments intensified during the late window Sunday afternoon as John Harbaugh decided to go for two late in the fourth quarter of the Baltimore Ravens’ loss to the Green Bay Packers.

So what is actually happening? Are coaches relying too much on numbers? Are analytics ruining the game? Are we incorrectly judging process over results?

Let’s take a look.
How are analytics actually used with in-game decision-making?


Analytics in this context refers to the use of data to make better, more informed decisions.

NFL teams build their own analytical models. They are not perfect, but they can be quite sophisticated, taking into account time, score, down, distance, opponent, weather, stadium, timeouts, and other factors.

For example, on a fourth-and-3 from the opponents’ 26, a model can assess how likely the offense is to convert, how likely the kicker is to make the field goal, and what success or failure in either scenario would mean in terms of the team’s chances of winning the game.

When it comes time to make a decision, the coach has to act quickly. Many rely on a simple color-coded system. Whoever is in charge of game management relays a color to the coach through the headset. Green might mean go. Red might mean kick or punt. And yellow might mean it’s a toss-up. The coach can listen to what the model suggests and then make his own decision about what to do.
Why should the coaches even bother paying attention to the analytical models?


Because decision-making under uncertainty is really hard, and if coaches don’t have a sound process, they can fall victim to a number of biases. For example, they might fail on a fourth down in the first quarter. The second quarter rolls around, and they’re faced with another decision. They remember the previous fail and decide to kick even though all of their planning (and the data) suggested going for it again. That would be a simple case of recency bias, but there are many other biases that could cloud the coach’s judgment and lead to a bad process.

The analytical model can serve as a safeguard of sorts that reminds the coach what his original plan was. Often these types of things are decided in the days leading up to the game when the coach’s mind is clear. The planning aspect is crucial. If a team knows it’s going to be aggressive, and the offense is facing a third-and-9, the coach can tell the play-caller and quarterback that it’s a four-down situation. That way, everyone knows that a 6-yard completion isn’t necessarily a bad outcome. The opposite of this is when a team has to waste a timeout on fourth down because the coach can’t decide what he wants to do.

These are high-stress, competitive environments. Decisions need to be made in a matter of seconds. And leaning on gut feelings can lead to mistakes. That doesn’t mean the coach always has to do what the model suggests. He may have updated information. Maybe it’s a short-yardage situation, and during the week the plan was for a power run. But the guard who’s supposed to pull on the play just injured his ankle. That could affect what the coach wants to do.

But using the analytical model as a guide can be incredibly valuable.
Ok Mr. AP Calculus, then how do you explain what happened to the Chargers vs. the Chiefs?


A lot of coaching comes down to making smart bets that ideally will give your team a slight edge. For example, during the game-planning process, a team might notice that its opponent plays Cover-3 on a high percentage of early downs. So when they are putting together their call sheet for early downs, they’re going to want to include route concepts that beat Cover-3. Does that mean they’re definitely going to get a high percentage of Cover-3 on early downs in that specific game? Of course not.

No one strategy is going to work 100 percent of the time (or anywhere close to that). Staley went into the Chiefs game with a plan. He knew the strength of his team was Justin Herbert and a top-five offense. He knew that his defense ranked 26th, according to TruMedia’s Expected Points Added (EPA) per drive metric. It was unlikely that they were going to consistently stop Patrick Mahomes, Travis Kelce, Tyreek Hill and the Chiefs’ offense for the entire game. He made the assessment that his team needed to score touchdowns, not field goals. He was not reckless about it. The publicly available models backed his decisions. It was a perfectly reasonable approach. And he explained it well.



Did it work out? No. The Chargers went 2-for-5 on fourth downs and twice failed to convert inside the Kansas City 6-yard line. Some might argue that Staley should have learned his lesson and stopped going for it after he failed once or twice. But that would have been a mistake. The third failed fourth-down attempt had nothing to do with the first or second ones. The Chargers went into the game with multiple fourth-down calls that they liked and a well-conceived plan. It didn’t work out. That’s football.

The most common mistake we can make when analyzing these decisions is hindsight bias. As Philip E. Tetlock and Dan Gardner wrote in “Superforecasting”, “Once we know the outcome of something, that knowledge skews our perception of what we thought before we knew the outcome.”

Going into Thursday night’s game, the Chargers ranked second in EPA on fourth downs. In other words, Staley’s aggressiveness on fourth downs had helped the Chargers all season. The only team that had benefited more from going for it on fourth down was the Ravens.

If there was a way to know beforehand whether the fourth-down attempt was going to be successful, that would make life a lot easier, and the decisions would be simple. Unfortunately, that’s not the case. Coaches have to make decisions based on the information they have at the time. And we should judge their decisions through that lens.
Ok, what about the Ravens’ decision to go for two at the end of the Packers game?


For those who missed it, Tyler Huntley ran in for an 8-yard touchdown with 42 seconds left to cut the Packers’ lead to 31-30. Harbaugh decided to go for two to try to take the lead rather than kick the extra point for the tie.

After the Ravens failed to convert, many were quick to rip analytics. The only problem? This wasn’t an analytics play. Harbaugh said as much after the game and again on Monday.



The Ravens were down four of their top six cornerbacks, and Aaron Rodgers had gone up and down the field against them for most of the game. Harbaugh had two choices:

1. Go for two, hopefully get it, and then stop the Packers from driving for the winning field goal in regulation.
2. Kick the extra point, stop the Packers from driving for the winning field goal, and try to win in overtime.

What made the decision especially tricky was the there were still 42 seconds left, and the Packers had a timeout. In other words, even if the Ravens had converted, they weren’t guaranteed a victory. There was still plenty of time left for Green Bay to drive into field-goal range. Had there been something like 15 seconds left, going for two would have been more appealing. Had there been more than a minute left, it probably would have been an easier decision to kick. But this was in a tricky grey area where it could go either way.

Critics of analytics often point to the “human element” that gets ignored. But that’s precisely what Harbaugh was taking into account here. The numbers didn’t tell him what to do. It was a toss-up. He talked to his coaches. He talked to his players. And he made a decision. It didn’t work out. But this is what good leadership looks like.



One area where it’s fair to criticize Harbaugh: not going for two after the Ravens’ first touchdown in the fourth quarter. They cut the lead to 31-23 with 4:47 left. That’s a situation where we generally see the analytically inclined teams go for two. The thinking is you have about a 50% chance to convert. If you make it the first time, then you only need an extra point the second time to take the lead. If you miss it the first time, you still have a chance to convert the second time and tie the score. The nightmare scenario would be not converting either time.

But given that Harbaugh was willing to take a risk on the final two-point conversion, he (and the Ravens) could have benefited and given themselves more margin for error by going for two after the first score.
Isn’t all this just overthinking it? What’s wrong with just taking the points?


Well, that often fails too.

On Saturday night, the New England Patriots trailed the Indianapolis Colts 20-7 and faced a fourth-and-goal from the Indianapolis 7. Bill Belichick opted to kick a 25-yard field goal with 8:57 left. That made it a 20-10 score, meaning the Patriots had to get a stop, score, get another stop and score again in a limited amount of time.

Here’s how Belichick explained his decision Monday.



The numbers suggested that the Patriots’ chances of winning at that point were low, regardless of which decision Belichick chose. You could argue that the only thing that mattered at that point was making the decision that offered the biggest bump in win probability. In other words, even though the chances of converting fourth-and-goal from the 7 were low, that was the Patriots’ only real avenue to increasing their odds of winning the game.



Having said that, the Patriots’ defense was playing well, and Carson Wentz was struggling. The model above showed the decision as a toss-up.

Belichick made the decision that he thought would give his team the best chance to win — just like Staley and Harbaugh. But Belichick didn’t receive much criticism because he opted for the safer, more conventional option. This is usually how it goes. When a coach makes a decision that’s different from what we’ve most often seen in the past 50 years, he gets crushed. When he makes the more conventional choice, even if it doesn’t work out, the criticism isn’t nearly as harsh.

It’s worth pointing out that in the same game, Frank Reich went for it on fourth-and-1 from the Colts’ 43 with 5:38 left. Wentz converted a quarterback sneak, and Indianapolis was able to take nearly two more minutes off the clock while forcing the Patriots to use a pair of timeouts. Those who criticized Staley on Thursday night seemed to stay silent after Reich’s decision. Why? Because it worked out.
Are there other factors that coaches have to consider?


Yes. One that we probably don’t talk about enough is job security. Ultimately, NFL owners get to decide whether to retain or fire head coaches. A coach can preach “process over results”, but if he’s losing games, getting ripped by the media and working for an owner who doesn’t care for analytical models, well, he might not have a job for long. That’s why, in my opinion, it’s so valuable for coaches to explain the thinking behind the decisions they make.

Another factor is the coaches’ relationship with the players. There’s value to fostering a culture of aggressiveness — specifically when there’s a consistency to it. This was very much the case with Doug Pederson and the 2017 Philadelphia Eagles team that won the Super Bowl. Offensive players appreciated that their coach had confidence in them. Defensive players didn’t sulk when the decisions failed. They knew the aggressiveness was part of the team’s identity and would be more helpful than harmful over time.

In the clip above, Mark Andrews looked like he was all-in on Harbaugh’s approach. And after Thursday night’s game, Keenan Allen supported Staley’s decisions.
So where does that leave us?


We should always remember that the coaches could be privy to more information than the public, and that information could factor into their decision-making. But that doesn’t mean they always deserve the benefit of the doubt. Some NFL coaches would have you believe they are among the most important members of society. But as we see on a weekly basis, they are not always the best and the brightest. And specifically, they don’t always have coherent plans when it comes to in-game decision-making. It’s perfectly fine to question and criticize their decisions.

If you’re someone who doesn’t believe in the value of using data to help make better-informed decisions, feel free to continue to “Lol analytics!” every time you see something bad happen in a football game.

For the rest of us, let’s listen to the coaches when they explain their decisions and judge process over results.

[Rugger05]

On Sept. 10, before he had coached a down of regular-season football for the Chargers, Brandon Staley was asked about how aggressive he planned to be on fourth downs. The Chargers were two days away from opening their 2021 season on the road against the Washington Football Team.

“I’m a big fan of trusting our quarterback, of putting the ball in his hands and putting pressure on the defense. I believe in that,” the rookie head coach said. “From a strategic standpoint, there are a lot of reasons why. Fear is definitely not going to be a motivator for us, one that we make decisions by. We’re going to be able to live with our results because we trust our process and we trust our players. … Sometimes you’re going to have to live with the criticism. That’s part of the job.”

We are now 15 weeks into the NFL season, and every word Staley said in that response has held true.

If you are clamoring about his decision-making — if you are surprised or put off or confused — then you have not been paying attention.

Since Staley accepted the Chargers’ head coaching position in January, this has been his vision. This has been his philosophy. This has been the identity he is determined to forge.

He is committed to the win probability advantages gained by going for it on fourth downs in specific situations. The process matters. The math matters. The results do not. This is who he is. It is not going to change. So get used to it.

Brandon Staley is coaching to win. He has one of the most gifted and talented quarterbacks in the league in Justin Herbert. And, damn it, that is the player he is going to let decide the game.

The Chargers lost to the Chiefs, 34-28, on Thursday night at SoFi Stadium. Tight end Travis Kelce turned a 4-yard catch into a 34-yard touchdown, slicing through the Chargers defense with yards after the catch for the walk-off score in overtime. With a win, the Chargers would have taken over first place in the AFC West. They let a massive opportunity slip away.

The Chargers went 2-for-5 on fourth down in the game. Two of those failed fourth-down attempts happened inside the Kansas City 6-yard line in the first half. The third came on a fourth-and-2 inside the Kansas City 30-yard line in the third quarter.

The hindsight vultures will say the Chargers left nine points on the field, assuming kicker Dustin Hopkins connected on the three field goals he could have attempted on those three fourth downs. But that logic is flawed because it is based on results. The win probability math shows that all three of these decisions, in and of themselves, increased the Chargers’ chances at victory.

The decision to go for it on fourth-and-goal from the Chiefs’ 5-yard line on the opening offensive possession increased the the Chargers’ odds of winning by 3.1 percent, according to Ben Baldwin’s fourth-down model. Herbert delivered a perfect ball to Donald Parham in the back of the end zone. In a scary scene, Parham slammed his head into the turf on his fall. He dropped the ball. And he was stretchered off the field and transported to a local hospital for further evaluation.

If Parham makes this catch, no one questions this call. It would have been regarded as a great decision, like the Chargers’ fourth-and-goal attempt from the 4-yard line two weeks ago at the Bengals that resulted in a Keenan Allen touchdown. But the rationale for whether a decision is good or bad cannot rely on the result of the play. It must rely on the inherent value of the decision itself.

The decision to go for it on fourth-and-goal from the Chiefs’ 1-yard line on the final play of the half increased the Chargers’ chances at winning by 2 percent, according to Baldwin’s model. Daniel Sorensen batted Herbert’s attempt to Allen at the line of scrimmage. But, again, the result of the play does not make it a bad decision.

The decision to go for it on fourth-and-2 from the Chiefs’ 27-yard line early in the third quarter increased the Chargers’ chances at winning by 4.4 percent, according to Baldwin’s model. Herbert threw to tight end Jared Cook on a short route to the left. Linebacker Nick Bolton jumped the throw and broke it up.

All these decisions improved the Chargers’ chances at winning. They also converted on two fourth downs. One, a Herbert sneak, led to a touchdown in the second quarter. Another, a fourth-and-1 handoff to Joshua Kelley, set the Chargers up with a first-and-goal from the Chiefs 4-yard line.

“We love to be put in those situations,” Herbert said.

“Leave the game in our hands,” Allen said. “I don’t want to kick the ball and watch it and hope.”

What lost them the game was a lack of execution.

Mike Williams dropped what should have been a touchdown pass two plays before the failed fourth-and-goal attempt on the opening drive.

Cook dropped what should have been a touchdown pass — which, to be fair, was slightly high from Herbert — two plays before the failed fourth-and-goal at the end of the half.

Kelley fumbled at the goal line on an ill-advised attempt to jump over the pile with the ball extended.

The defense failed to contain Kelce without Derwin James on the field. James only played 26 defensive snaps in the game because his hamstring injury flared up. He was off the field for the Chiefs’ final six possessions. Kelce only had one catch for 14 yards when James was in the game, according to play-by-play data from TruMedia. Kelce had nine catches for 177 yards and two touchdowns with James off the field.

The Chargers gave up too many loose plays in the final stages of this game. They let Patrick Mahomes escape the pocket and create with his legs and arms. The final six plays of the Chiefs’ game-tying drive in the fourth quarter came in off-script situations, including Mahomes’ 32-yard scramble. Mahomes also hit Tyreek Hill for 20 yards on a third-and-10 after evading pressure on one of those plays.

Hill finished with 148 yards receiving. Mahomes threw for 410 yards.

“It’s the recipe that he uses to play well,” Staley said of Mahomes after the game. “And when you allow loose plays by the quarterback, then you see numbers like he had today.”

Staley’s fourth-down decision-making is one of the primary reasons the Chargers are in this position, fighting for a playoff spot down the stretch. They do not beat the Chiefs in Week 3 if Staley does not go for it on two fourth-and-9 attempts in the fourth quarter. They might not have beaten the Raiders in Week 4 if Staley does not go for it on a fourth-and-2 in the fourth quarter that set up a touchdown to increase their lead to two scores. They do not beat the Browns in Week 5 if Staley does not go for it and successfully convert three fourth downs in the second half that led to two touchdowns.

The Chargers were converting fourth downs at a 87.5 percent clip through the first five weeks of the season. That was not a sustainable rate, and mathematical regression has hit since then. The Chargers are 8-of-18 on fourth downs — 44.4 percent — since Week 6.

The results have changed. The decision-making process has not — and will not — change.

“That’s going to be the mindset no matter who we play,” Staley said Thursday night.

“I know that the quickest way to win a game like that is to score touchdowns, not field goals, especially considering who’s on the other side,” he added. “And to me, when you feel like you’re in an advantage situation, when you don’t feel like it’s a gamble, when you feel like this is an advantage for you, then that’s going to be our mindset. And I don’t think that any decision that we made tonight, I made tonight, was a gamble. We felt like it was an advantage situation for us, and that’s why we did it. If we didn’t feel like that was the case, then we would have kicked a field goal or we would have punted. That’s the way we’re going to do things around here, and I know that our team embraces that mindset, and we’re going to continue to do it every game moving forward.”

Staley will go for it on fourth down again, and he will fail again.

The one thing he will never do is coach not to lose.

He told you that before the season started. It’s not his fault if you weren’t listening.

[chargeroo]

When it's fourth and goal and the los is inside the five, the defense has a lot less ground to cover. Not any worry that a speedy receiver will go deep on them. I think going for it on fourth down in that situation must have a lower percentage of completion than going for it from other parts of the field.

[BayAreaBoltz]

Brett Kollmann does a good job of breaking down why almost all of Staley's 4th down calls from the KC game were correct. Again, not all, but almost every one (and why).

[Velo]

Brett Kollmann does a good job of breaking down why almost all of Staley's 4th down calls from the KC game were correct. Again, not all, but almost every one (and why).

Who is Brett Kollman?

[Xenos]

Velo here’s another good article breakdown:

https://theathletic.com/3030201/2021/12/23/hate-nfl-analytics-based-decisions-the-math-adds-up-more-often-than-you-might-think/

In the moments immediately following the Chargers’ overtime loss to the Chiefs last Thursday night, the Fox Sports studio crew assembled to offer their closing thoughts on the affair.

It was no mystery where the conversation would immediately turn. The Chargers went for it on five fourth downs in the game but converted just two. Two of the three failed attempts came inside the Chiefs’ 6-yard line.

The analysis did not seize on the failed execution in those instances. The panel instead immediately eviscerated head coach Brandon Staley’s fourth-down decisions in and of themselves.

“All those points you took off the board, big mistakes, I think,” Terry Bradshaw said. “Big mistakes.”

“Neither one of us can spell analytics,” Howie Long added, “but it took a beating tonight.”

That brief exchange offered a perfect illustration of what’s wrong with the way fans and many in media view the use of data and analytics in pro football. The topic is almost always framed in one particular way: As a driving force behind failed decisions that were critical to the outcome of games.

Instead of characterizing failed fourth-down decisions as “leaving points on the field,” why not describe them as what they really are: attempts at winning the game.

“I know that the quickest way to win a game like that is to score touchdowns, not field goals,” Staley said.

If you think this is merely a matter of semantics, you’d be wrong. In fact, the manner in which we talk about data in football matters more than you think.

Because integrating analytics often becomes a flashpoint in postgame analysis and can still be seen as controversial, the coaches who make these decisions must have a certain amount of chutzpah to go against the proverbial grain. After all, who wants to be the guy getting trashed on national television?

Frank Frigo, chief innovation officer for Champion Gaming and a founder of EdjSports, offered an example of the hesitancy of coaches to go all-in. He recounted this week the story of a prominent NFL assistant who had fully embraced the benefits of analytics and, seemingly, was poised to become a leader in this area. Emphasis on “was.”

“He was super into all of this,” Frigo recalled. “He read everything that Football Outsiders wrote, wanted to know everything about our models. When he interviewed for head-coaching jobs, he was on the phone with me on the eve of the interview because (teams) were starting to care about this.

“So, lo and behold he gets a head-coaching job and he just turned into a different person. He got the job, got his contract, and all of a sudden, he was not going to go in there and be the guy who was doing stuff that was controversial and getting his head ripped off in the media every week if it didn’t work. He completely changed his tone.”

It’s worth noting that this coach, who shall remain nameless, is no longer a head coach. Frigo likened the arguments around analytics to our political discourse — lots of talking and not much critical thinking.

“There’s a famous quote on this,” he said. “It goes like this: ‘The problem with the world is foolish people are certain of themselves and really smart people are uncertain of themselves.’ Really smart people aren’t afraid to say, ‘I don’t know’ or ‘this is more likely than that,’ whereas a lot of people just want definitive answers. Fans will gravitate to somebody with conviction. But we (data analysts) don’t talk in that language.”

It feels like there are better odds that the Lionswill qualify for the playoffs than our altering today’s rancorous political discourse. But maybe it’s not too late to change the conversation around analytics.

Consider this story an effort to do that in some small way. Here’s an idea: What if we flip things on their collective head and discuss situations where analytically based decisions worked rather than failed? In the end, logic strongly suggests that if you consistently make decisions that empirically improve your team’s chances of winning games, you are going to win more games over the long term.

So, at least in this piece, that’s what we’re going to focus on. Here are a few prime examples from the 2021 season, compiled with a little help from EdjSports data analyst Ian O’Connor and The Athletic’s resident analytics expert, Ben Baldwin.


Week 1: Bengals 27, Vikings 24

Here’s an interesting dilemma. The Bengalswere in a tied game with Minnesota and faced a fourth-and-1 at their own 48-yard line with 39 seconds remaining in overtime. The Vikingshad two timeouts left, something for a coach to consider when deciding whether to punt when a field goal could win the game.

But the most important piece of information here is that going for it on fourth down would add roughly 11.3 percentage points to Cincinnati’s win probability based on EdjSports’ computer model. Head coach Zac Taylor gave the go-ahead, and quarterback Joe Burrow hit C.J. Uzomah for a 32-yard gain. The Bengals kicked a winning field goal three plays later.

Week 2: Ravens 36, Chiefs 35


This example highlights a point that data folks are often indignant about. The end of a game is a strange time to suddenly place tons of confidence in a defense that has let you down all day. Punting the ball to a team that has moved the ball up and down the field, in the hopes that your defense will now find a way to slow the opponent, is very flawed logic, at minimum.

The Ravens were protecting a one-point lead, then saw their drive stall at their own 43-yard line. They faced a fourth-and-1 with 1:05 remaining. The Chiefs, meanwhile, had one timeout, Patrick Mahomes and a whole lot of firepower. Kansas City had already scored 35 points. Ravens head coach John Harbaugh called for a “go” and saw Lamar Jackson run off the left edge for the necessary yard and a first down. Baltimore then ran out the clock and enjoyed a win.

Despite the reality that Kansas City had amassed 405 yards that day, this wasn’t just a “gut decision” by Harbaugh. Baltimore’s analytics department is among the most integrated in the league, so Harbaugh unquestionably knew the numbers that supported his decision. EdjSports’ computer model preferred the fourth-down attempt over a punt by a margin of 16.4 percentage points in relation to the impact on win probability.

A sound decision all around — even if it hadn’t worked.


Week 3: Chargers 30, Chiefs 24

Staley took a lot of heat for his decision in last week’s Chiefs game, but there was precious little conversation about his decision-making in the Chargers’ first meeting with Kansas City.

The score was tied at 24 with 48 seconds remaining after Chargers quarterback Justin Herbert threw an incomplete pass to Mike Williams on a third down. That left the Chargers with a fourth-and-4. They were then flagged for a false start, backing them up 5 yards. Staley faced a difficult choice: Attempt a fourth-and-9 from the Chiefs’ 35-yard line or try a roughly 52-yard field.

He opted to go for it, and the Chargers drew a pass-interference call that allowed them to gain an automatic first down and pick up 15 yards. Williams caught a 4-yard touchdown pass from Herbert two plays later to put the Chargers ahead for good.

The payoff for attempting the fourth down has to be viewed in its totality. It’s not just that it allowed the Chargers to ultimately score the winning touchdown (that was hardly guaranteed). But possessing the ball longer also took an additional 16 seconds off the clock. That proved significant when the Chiefs got the ball back at their own 25-yard line with just 32 seconds remaining and one timeout left. And, of course, the Chargers’ touchdown also made them much less vulnerable to a potential tying final possession (because the Chiefs needed a touchdown, not a field goal).


Week 8: Packers 24, Cardinals 21

Sometimes it isn’t just one decision but a series of them that can be the difference.

This was the case in this Packers victory, with Green Bay converting two key fourth downs that directly led to crucial points in a one-score game.

The Packers faced a fourth-and-1 on the first play of the second quarter when they trailed 7-0. Rather than settle for a pretty routine field goal from the Cardinals’ 13-yard line, head coach Matt LaFleur called for an A.J. Dillon run. The Packers converted, then scored a touchdown two plays later. The Packers later converted a fourth-and-3 in the third quarter from the Arizona 7-yard line, setting up another touchdown a play later.

Green Bay did fail on a fourth-down attempt with 3:26 left in the fourth quarter, but as we explained earlier, coaching decisions are sometimes best viewed in a broader context rather than individually. Just as a season’s worth of sound, analytically based decisions can boost a team’s winning percentage, the same can happen over the course of four quarters in an individual game.

In this instance, the Packers scored 14 points as a direct result of those fourth-down decisions. Had they settled for field goals in each instance, they would have walked away with a maximum of nine points. That’s critical in a game Green Bay won by just three.


Week 15: Colts 27, Patriots 17

Most of the conversation around this game has centered on the decision by Patriots head coach Bill Belichick to settle for a field goal with nine minutes remaining rather than attempting a fourth-and-goal from the 7-yard line while trailing by 13 points.

But what’s been much less discussed is that Colts head coach Frank Reich, in contrast, was particularly aggressive on fourth downs throughout the game. That loomed large in the outcome. It was a great contrast in styles and approach, with Belichick — despite being perhaps the greatest strategist in the game’s history — still having a sizable blind spot when it comes to analytical decisions.

As one league source said this week, “It’s crazy how much more (the Patriots) would win if he made three to four better decisions each game … X’s-and-O’s-wise, obviously he’s dominant.”

As for Reich, he called for three fourth-down attempts in the game, with the Colts converting each one. Those drives amounted to just three points, but they affected the game in other ways. Take, for instance, the ensuing drive after New England’s final field goal cut the Colts’ lead to 20-10. The Colts faced a fourth-and-1 with 5:38 left from their own 43-yard line. Might look risky to try it, but the data suggested going for it was prudent (it’s also worth mentioning the Colts lead the NFL in fourth-down conversions, 68.2 percent).

The Colts converted and were rewarded. The Patriots went on to burn two of their three timeouts after the next two plays. By the time New England forced a punt, just 3:55 remained and the Patriots still needed two scores.

Here’s hoping the next time analytics takes center stage on a post-game show, the talking heads present more even-handed analysis. Maybe we aren’t going to find much common political ground, but can we at least endeavor to change our language on analytics in football?

[Bolt-O]

Staley is a client...

[Guest]

Ok so why weren't the Bolts going for it on 4th down against the Texans?

[richpjr]

Ok so why weren't the Bolts going for it on 4th down against the Texans?

I didn't see the winning percentage chances during the game, but the 4th downs were longer than the ones against the Chiefs.

[northerner]

Ok so why weren't the Bolts going for it on 4th down against the Texans?

this game showed the downside of FG's = 15 points with 4 drives inside their 35 yard line (3 FG's and one TD). so you are damned if you do, and damned if you don't.

i still like that our coach is on progressive side of the analytics. this is going to benefit us down the road.

[richpjr]

this game showed the downside of FG's = 15 points with 4 drives inside their 35 yard line (3 FG's and one TD). so you are damned if you do, and damned if you don't.

i still like that our coach is on progressive side of the analytics. this is going to benefit us down the road.

In the end, the players have to execute. We got 2 TDs on 5 tries against the Chiefs for 14 points. If we kick FGs and make all of them, we have 15 points. Pretty much a wash. However, 2 of the 3 misses were plays that should have been made and could have swung the game our way. I far prefer playing to win than playing not to lose.