The NFL’s Wild Card Weekend featured many exciting games filled with big plays, huge momentum shifts, and lots of second-guessing of the coaches. Let’s take a look back at three key fourth down decisions from Colts-Bills, Ravens-Titans, and Browns-Steelers using Sports Info Solutions’ Win Probability Calculator. SIS’s Win Probability Calculator factors in the current game state along with the strengths of the two teams playing.
Scenario: The Colts are leading the Bills 10-7 with 1:53 left in the 2nd quarter. The ball is on the Bills’ 4-yard line, and it is fourth and goal. The Colts had all three of their timeouts while the Bills only had two. The Colts decide to try for a touchdown and bypass the option of kicking a field goal. The resulting play is an incomplete pass and the Bills take over on their own 4.
Win Probability Breakdown: Before the decision to send the offense back out onto the field or to attempt a field goal, we consider the Colts’ win probability to be equivalent to the win probability if they make the optimal decision given the situation. At this point, the Colts’ win probability if they decided to send the offense back onto the field was 67%:
|Decision||Outcome||Colts’ Win Probability||Change from Pre-Snap Win Probability|
|Go for it||Touchdown||80%||+13%|
|Turnover on Downs||57%||-10%|
|Kick Field Goal||Made Field Goal||66%||-1%|
|Missed Field Goal||55%||-12%|
Only the outcome “Colts Touchdown” represents a positive change in win probability. Because of the large potential gains in win probability, the Colts clearly made the right choice to forgo the field goal and try for a touchdown. A field goal comes with little risk, but even a make results in a 1% decrease in win probability. This makes the field goal attempt an inferior decision regardless of make or miss. Unfortunately for them, the Colts did not score a touchdown and turned the ball over on downs on a ball barely off the fingertips of rookie wideout Michael Pittman Jr. However, their overall win probability was still at 57% when the Bills took over on their own 4, so even the negative outcome still left them in a good position to win the game.
Verdict: The Colts made the correct call to try for the touchdown, but ended up with an unfavorable outcome.
Scenario: The Titans are trailing the Ravens 17-13 with 10:06 left in the 4th quarter. The ball is on the Ravens’ 40-yard line, and it is fourth-and-2. Both the Titans and Ravens had all three of their timeouts remaining. Instead of trying to get the first down, the Titans punt the ball, and the Ravens take over at their own 15-yard line.
Win Probability Breakdown: At the time of the decision to go for it or punt, the Titans’ win probability if they decided to send the offense back onto the field was 43%, but instead they sent the punting unit on the field which dropped their win probability much lower to 36%. For reference, since 2016, teams have had a success rate of approximately 58% on fourth-and-2. Even with an off day, given that the Titans had Derrick Henry in their backfield, they would have felt even better about their chances of converting on fourth-and-2. The potential scenarios are laid out below:
|Decision||Outcome||Titans’ Win Probability||Change from Pre-Snap Win Probability|
|Go for it||Convert||58%||+15%|
|Turnover on Downs||20%||-23%|
It becomes clear that a Titans conversion is the only outcome that increases their win probability. While a failed fourth down attempt would be damaging, a Titans punt sacrifices win probability at a time when they are already behind and need to embrace variance. The difference in win probability between a conversion and a punt (22 percentage points) is much larger than the difference between a failed attempt and punt (16 percentage points).
Verdict: The Titans should have tried for the first down but instead played it too safe. The reward for a successful conversion far outweighed the risk.
Scenario: The Steelers are trailing the Browns 35-23 with 15:00 left in the fourth quarter. The ball is on the Steelers’ 46-yard line, and it is fourth-and-1. The Steelers had three timeouts remaining while the Browns had two. Instead of trying to get the first down, the Steelers punt the ball, and the Browns take over at their own 20-yard line.
Win Probability Breakdown: The Steelers were facing an uphill battle with a win probability of just 11% before the play, so they needed to be as aggressive as possible. Since 2016, teams have had a success rate of 65% on fourth-and1, so the Steelers had a high chance of getting the first down. Instead, they decided to accept a delay of game and punt which dropped their win probability to 7%. The three potential outcomes that we will investigate are a Steelers conversion, a turnover on downs, or a punt:
|Decision||Outcome||Steelers’ Win Probability||Change in Pre-Snap Win Probability|
|Go for it||Convert||13%||+2%|
|Turnover on Downs||4%||-7%|
Based on the score and time left in the game, the Steelers had very little chance to win regardless of their choice as evidenced by the single-digit changes in win probability. The only chance to increase their win probability would have been a successful conversion, but the Steelers did not even give themselves a chance to increase their win probability by choosing to punt.
Verdict: The Steelers should have tried for the first down. Their chances of winning would be low regardless, but by punting, the Steelers did not even take the opportunity to improve their chances of winning.
Unlike the Titans and Steelers, the Colts at least chose the best option to increase their win probability. Even though the outcome was an unfavorable one, the Colts should be applauded for doing their best to win the game while the Titans and Steelers refused to take the risks needed for them to win or get back into the game.