The former New York Jets head coach took heat for a surprising two-point attempt
During his time as the head coach of the New York Jets, Todd Bowles was known as anything but aggressive (outside of his blitz packages). The former safety and defensive coordinator brought an old-school mentality to the position, electing to take the safe route with nearly every game-management decision he faced. Conservative punts and field goal attempts were frustratingly common.
Now the head coach in Tampa Bay, Bowles came under fire for what was perceived as an overly aggressive decision in the Buccaneers’ recent playoff loss to the Lions.
With 4:37 remaining in the fourth quarter, Mike Evans caught a 16-yard touchdown pass from Baker Mayfield, cutting Detroit’s lead to 31-23. Tampa Bay had two timeouts remaining.
Conventional wisdom says the obvious move here is to kick the extra point and cut the lead to seven. Get a stop, score another touchdown, and you can send the game to overtime.
But Bowles shocked the world when he did something that would have been inconceivable during his Jets days – he took the aggressive route.
And not only was Bowles’ decision aggressive, but it was correct, contrary to the belief of those screaming from their couches across America.
Bowles elected to keep his offense on the field and go for the two-point conversion while trailing by eight points. The attempt failed, and it didn’t come into play anyway since Mayfield threw an interception on the following possession. Still, despite the failure of the conversion, Bowles made the correct decision, and I’m here to explain why.
The outrage over this decision highlights the lack of education that still exists in the general NFL community about the basics of analytics. Actually, scratch that – I don’t think “analytics” is the right word here. Analytics is a fancy term that seems to scare people away. Rather than calling it analytics, how about we just call it what it is: simple logic.
Point #1: Two-point conversions are simply better
Let’s start with this basic fact: going for two is more efficient than going for one in today’s NFL. I don’t think most fans understand this. With the extra point being moved back in 2015 and offenses converting two-pointers more efficiently than ever, going for two is the better play. At worst, it’s essentially equal to kicking an extra point.
Here are the league average conversion rates in 2023:
- Two-point conversions: 55% (70 for 127)
- Extra points: 96% (1,121 for 1,169)
At those conversion rates, a two-point conversion is expected to produce more points. Simply multiply the two-point conversion rate by two to get the average points per attempt, and you can see it outperforms one-point attempts:
- Two-point conversions: 70 for 127 = 140 points on 127 attempts (1.1 points per attempt)
- Extra points: 1,121 for 1,169 = 1,121 points on 1,169 attempts (0.96 points per attempt)
In 2023, two-point conversions produced 1.1 points per attempt while extra points produced 0.96 points per attempt. That means two-pointers produced approximately 15% more points per attempt.
To be fair, the success we saw in 2023 was groundbreaking. The 55% conversion rate was the highest in a season since teams converted 60% of tries in 2006, although there were only 35 attempts in the league that year. In 2022, the conversion rate was much lower at 47%.
Still, if you include all nine seasons that have passed since the extra point was moved back (2015-23), teams have converted 49% of two-point conversions compared to 94% of extra points. That still gives a slight lead to two-pointers: 0.98 points per attempt compared to 0.94 points per attempt (4% advantage for two-pointers).
So, in general, teams should be going for two more often. In the long run, you’ll probably score more points by going for two every time than by going for one every time. Whether or not this would be smart obviously depends on the quality of a given team’s offense and kicker, but you get the gist: two-pointers are more efficient than one-pointers. At the very least, they’re basically the same.
With that out of the way, let’s dive deeper into Bowles’ decision to go for two in that particular situation.
Point #2: The reward outweighed the risk
Here’s the biggest oversight made by critics of aggressive game-management decisions (whether it’s two-point attempts or fourth-down attempts): they’re always too afraid of the potential downside to see the potential upside.
These critics think the safer decision is always better just because it will fail less often. In actuality, playing it safe too often could lead to a gradual accumulation of lost value in the long run. Maybe you’ll never screw up, but you’ll never get those bonus points either, which could cost your team advantages that may have swung a game. Teams that are willing to take risks will fail sometimes, but they will probably net positive in the end.
Why? Because the reward is usually worth the risk.
That brings us to Bowles’ decision.
Let’s analyze it from a risk-reward perspective. Remember, the Bucs were down by 8 with 4:37 remaining.
- Potential reward: Make the two-point conversion, creating an opportunity to win in regulation on the next TD (assuming the XP kick is made)
- Potential risk: Fail the two-point conversion, forcing yourself to make a two-point conversion on the next TD to force overtime
- Bypassed option: Score two TDs and kick two extra points to force overtime
Bowles’ mindset was to make the two-point conversion to give his team an opportunity to win in regulation later on, which he viewed as a better path to victory than scoring two touchdowns and trying to beat the high-octane Lions offense in an overtime period on the road. He essentially wanted to give the Bucs a potential two-step path to victory instead of a three-step path.
Here’s the key factor: If the two-point attempt failed, the Bucs could try another two-point attempt later on, and if they made it, they would end up in the same position as if they just kicked two extra points. And the odds of them converting at least one out of the two two-point attempts were good enough to make the reward worth the risk.
Let’s go step-by-step. To make this simple, we’ll use the 2023 league average conversion rates of 55% on two-point attempts and 96% on extra points. These rates surely differ game-to-game based on the offenses and defenses on the field, the kicker, the game location, the weather, and other variables, but using these baseline rates will make this exercise easier to digest.
Tampa Bay had a strong chance of getting 1-of-2 two-pointers to force overtime
At a 55% conversion rate, the odds of Tampa Bay making at least one out of two two-point attempts would be 80%. This means Bowles’ decision only had a 1-in-5 chance of resulting in the Bucs not at least getting to overtime in the event they scored two touchdowns. The risk really wasn’t as great as many people probably thought it was.
Picture it like this: Out of 100 tries, the Bucs would convert the first attempt 55 times. That leaves 45 tries in which the Bucs fail the first attempt and try again on the second to tie the game. They would be expected to convert 55% of those 45 tries, which is about 25 – giving us 80 tries out of 100 in which the Bucs get the one two-point conversion they need to at least tie the game.
Compare that to two one-point attempts. At a league-average conversion rate of 96%, the odds of the Bucs going 2-for-2 are actually just 92%. They’d make the first extra point 96 times out of 100. Out of the 96 tries in which the Bucs make the first kick, they’d make the second kick 96% of the time, which is about 92. That leaves us with 92 instances out of 100 in which the Bucs go two-for-two on extra points to tie the game.
So, Bowles’ preferred path actually sacrificed only a 12% chance of two Tampa Bay touchdowns being enough to at least get to overtime instead of losing in regulation (92% to 80%). Here’s another way to look at it: out of 100 scenarios, there were only 12 where Bowles’ decision would cost the Bucs in comparison to just going for the kick, whereas there are 88 scenarios out of 100 where it wouldn’t cost them. Those are the same odds of Kevin Durant making a free throw.
The potential reward was extremely strong
Now let’s talk about the potential reward that Bowles acquired in exchange for that small risk.
Tampa Bay converts the first two-point attempt 55 times out of 100. They’d make the extra point on the second touchdown in 96% of those 55 instances, which is 53. This means Bowles’ decision gave the Bucs a 53% chance of essentially winning the game in regulation if they scored two touchdowns (there may have been some time left for Detroit, but it likely wouldn’t be much).
On top of those 53 wins out of 100 scenarios, the Bucs still had more winning scenarios in their back pocket. As we discussed earlier, there are 25 scenarios out of 100 in which the Bucs fail the first two-point attempt and still make the second one to tie the game and force overtime (as they fail 45 times out of 100 and then convert 55% of the 45, which is about 25).
Out of those 25 overtime scenarios, if we assume the Bucs win half of the time, that gives you 12.5 more scenarios in which the Bucs win. Add that to the 53 scenarios where they win in regulation and that’s a 65.5% chance of winning in the event of two touchdowns with Bowles’ decision. (You can also throw in a rare 1% scenario where the Bucs make the first two-point attempt, miss the next extra point, and still win in overtime, which makes it 66.5%.)
If the Bucs took the safe route, their overall chances of winning would take a significant hit since they would completely eliminate the possibility of winning before overtime. They would guarantee the need for a three-step path to victory. It gets overlooked how costly this is. This path feels safe, but it really isn’t – encapsulating the flaws in conservative decision-making.
After scoring two touchdowns, the Buccaneers would still have to win the overtime period, which we are assuming is a 50-50 proposition.
So, with two extra-point attempts on the two touchdowns, the Bucs would have a 92% chance of making both kicks to tie the game. They would then have an implied 50% chance of winning in overtime. This means they would win half of the 92 scenarios in which they got to overtime, giving them an overall 46% chance of victory with two touchdowns.
Compare that to Bowles’ decision, which yielded a 66.5% chance of winning if Tampa Bay scored two touchdowns.
And there you have it. Bowles’ decision gave his team a significantly better chance of winning the football game.
Let’s round everything up.
Bowles goes for two
66.5% chance of winning if two touchdowns scored
- 53% chance of taking the lead in regulation (55% to convert first 2PT attempt → 96% to take lead with XP on second attempt)
- 12.5% chance of still winning in overtime after missing first 2PT (45% to miss first 2PT attempt → 55% to convert second 2PT attempt → 50% to win in OT)
- 1% chance of still winning in overtime after making first 2PT and missing second XP (55% to convert first 2PT attempt → 4% to keep tied after missing XP on second attempt → 50% to win in OT)
33.5% chance of losing if two touchdowns scored
- 20% chance of failing both 2PT attempts to lose in regulation (45% to miss first 2PT attempt → 45% to miss second 2PT attempt)
- 12.5% chance of losing in overtime after missing first 2PT (45% to miss first 2PT attempt → 55% to convert second 2PT attempt → 50% to lose in OT)
- 1% chance of losing in overtime after making first 2PT and missing second XP (55% to convert first 2PT attempt → 4% to keep tied after missing XP on second attempt → 50% to lose in OT)
Bowles goes for one
46% chance of winning if two touchdowns scored
- 46% chance of winning in overtime after making both XPs (96% to make first XP → 96% to make second XP → 50% to win in overtime)
54% chance of losing if two touchdowns scored
- 46% chance of losing in overtime after making both XPs (96% to make first XP → 96% to make second XP → 50% to lose in overtime)
- 8% chance of losing in regulation after missing at least one of the two XPs
Recap
Ultimately, this decision was correct primarily because Tampa Bay had the luxury of two shots at going for two to tie the game. Their chances of making at least one of those to tie the game were good enough to make it a worthwhile gamble for the major reward of being able to win in regulation and avoid overtime. This meant that, overall, Tampa Bay’s odds of winning were significantly higher by going for two on the first attempt even if it lowered their odds of getting to overtime.
Bowles sacrificed a 12% chance of losing in regulation (from 92% to 80%) in exchange for acquiring a 53% chance of winning in regulation (from 0% to 53%). This was a smart trade-off that massively increased Tampa Bay’s chances of winning.
It wasn’t long ago that a team successfully executed this strategy. Mike Vrabel and the Tennessee Titans pulled it off to upset the Miami Dolphins in Week 14. With 2:46 remaining in the fourth quarter, Tennessee scored a touchdown to cut Miami’s lead to 27-19, and the Titans went for two. They converted the attempt to cut the lead to six. Later in the quarter, the Titans scored again and were able to take the lead with an extra point, giving them a 28-27 win.
This strategy is especially appealing if you’re a road underdog, as both teams were.
I put the overtime victory odds at 50% in our exercise for the sake of making it easier to digest, but in reality, Tampa Bay’s odds of winning in overtime were probably much lower. The Lions are a better team and were playing at home. Outplaying them play after play in a 10-minute period would be difficult for the Bucs, especially if they lost the coin toss and let Detroit’s offense take the ball first (needing to rely on a coin is another reason to try and avoid overtime).
If you redid our exercise and lowered the overtime victory odds, then the advantage in favor of going for two on the first attempt would grow even larger. For instance, if we lowered Tampa Bay’s expected chances of winning in overtime to 40%, their chances of winning (with two TDs) would be 64% if they went for two on the first attempt compared to 37% if they went for one. That’s a 27% margin compared to the 20.5% margin we arrived upon when assuming the overtime odds were 50-50.
On one play from the two-yard line, anything can happen. If you’re the underdog, it’s smart to make the game come down to one play instead of a full overtime period.
Analytics don’t have to be scary
Bowles’ decision may not have led to victory (although the whole process didn’t get to play out since there wasn’t a second TD to give them another two-point shot, which was the crux of the whole thing). Still, the fact that a famously conservative coach like Bowles was willing to make a decision like this on the playoff stage was a sign that analytics are here to stay in the NFL.
Since that’s the case, I think it’s important that the general NFL fanbase gets a better grasp of the thought process that goes into these decisions. That was my motivation for writing this article.
Many fans are quick to dismiss analytics-influenced decisions just because they differ from the conventional wisdom that has dominated the NFL for decades. But if you take the time to think these decisions through, they usually make a lot of sense. And you don’t need overly complicated formulas that are beyond the comprehension of the average fan. It’s really basic logic when you break it all down.
Unfortunately, television broadcasts and the NFL media in general have not done a great job of educating people on the thought processes that go into these decisions. I hope this piece did a better job of that. Hopefully, this helped clear the air for a few football fans out there who were previously in the dark about why an NFL coach might make a decision like the one Bowles made.
