Each week coaches around the country are faced with real-time critical calls. By evaluating how teams manage their win probability, you can learn how to optimize future decisions and give your team the best chance to win.
Let’s a take a look at one of the best and worst decisions from Conference Championship Weekend in the NFL.
This week’s BEST and WORST decisions come from the AFC Conference Championship game, Tennessee vs. Kansas City. As the stakes get higher, coaching decisions continue to play a major role in determining which teams advance.
Best Decision – CHIEFS
With 2:58 remaining in the 1st quarter, 2 timeouts, and trailing by 10, the Chiefs faced a 4th and 2 from the TEN 28-yard line. Coach Reid’s decision to go-for-it was the correct call and increased the Chiefs (pre-snap) win probability by +3.4%. Our Edj Power Indexes (EPIs) had the KC rush offense ranked among the best in the league (#4) versus a TEN rush defense that ranked slightly above league average (#13). The KC pass offense also ranked among the best in the league (#3) versus a TEN pass defense that ranked above league average (#12). Based on our EPIs and the game state, teams that go-for-it, go on to win the game 38.3% of the time and teams that attempt a FG, go on to win the game 34.9% of the time. It’s also important to keep in mind that +3.4% (pre snap) win probability increase, also represents about 10% of the Chiefs current win probability. This was a big call and Coach Reid made the correct decision to go-for-it. The Chiefs went on to successfully convert this 4th down GO attempt, score on a TD on this drive, and win the game! For a little more context, let’s take a look at the three most likely outcomes and their associated win probabilities, and use that information to determine a break-even point for this play (based on clock at 1st quarter with 2:50 remaining).
– (Pre Snap) FG Win Probability: If KC had decided to attempt the 45-yard FG, they would have been expected to go on to win the game 34.9% – – Unsuccessful on 4th and 2: With TEN taking possession at the TEN 28-yard line, KC is expected to go on to win the game 29.1% of the time.
– Successful on 4th and 2 (Actual Result): Facing a 1st and 10 from the TEN 26-yard line, KC is expected to go on to win the game 42.8% of the time.
Now, let’s evaluate these outcomes in terms of risk and reward. In this case, Risk is defined as the difference in win probability between the (pre snap) win probability associated with attempting a FG and an unsuccessful conversion on 4th and 2. The risk is 5.8% (34.9% – 29.1%). Conversely, the Reward is the difference in win probability between a successful conversion on 4th and 2 and attempting a FG attempt. The reward is 5.9% (42.8% – 34.9%). We can calculate a break-even point using: Risk / (Risk + Reward) or 5.8 / 11.7 = 49.5%. In other words, if Coach Reid thinks his team will convert this 4th and 2 attempt 50% of the time, the correct decision is to go-for-it. Otherwise the correct decision is to attempt a FG. Empirical data on 3rd and 2 attempts suggests that an average NFL team would be expected to convert this about 60% of the time. Coach Reid made another great call and showed why he finished the season ranked #2 in the 2019 EdjSports Coach Rankings.
Note: For this analysis we use fairly conservative assumptions (i.e. the pre-snap win probability associated with the FG option and KC only picks up 2-yards on the successful conversion. If any of these factors change (i.e. KC picks up more than 2-yards on the 4th down conversion), the break-even point is even lower.
Worst Decision – TITANS
With 7:07 remaining in the 3rd quarter, 3 timeouts, and trailing by 4, the Titans faced a 4th and 4 from the TEN 38-yard line. By deciding to punt, Coach Vrabel made an error that cost his team -5.1% in (pre snap) win probability. Our EPIs had the TEN pass offense ranked among the best in the league (#4) versus a KC pass defense that ranked fairly high as well (#8). Based on our EPIs and the game state, teams that go for it, go on to win the game 32.7% of the time and teams that punt, go on to win the game 27.6% of the time. The 5.1% in win probability that’s at stake on this play also represents about 16% of the Titans current win probability. This was a big call and Coach Vrabel made an error in punting. Let’s look at the three most likely outcomes and their associated win probabilities and calculate a break-even point for this play (based on clock at 3rd quarter with 7:00 remaining).
– Punt (Actual Play): With KC taking possession at the KC 27-yard line, TEN is expected to go on to win the game only 25.5% of the time.
– Unsuccessful on 4th and 4: With KC taking possession at the TEN 38-yard line, TEN is expected to go on to win the game only 21.0% of the time.
– Successful on 4th and 4: Facing a 1st and 10 at the TEN 42-yard line, TEN is expected to go on to win the game 38.5% of the time. The Risk on this play is 4.5% (25.5% – 21.0%). The Reward is 13.0% (38.5% – 25.5%). Using the formula from the previous example, Risk / (Risk + Reward), the breakeven point is 4.5% / 17.5% = 25.7%. If Coach Vrabel thinks his team can successfully convert this 4th and 4 attempt at least 25.7% of the time, the correct decision is to go-for-it. Otherwise the correct decision is to punt. Empirical data on 3rd and 2 attempts suggests that an average NFL team would be expected to convert this about 55% of the time. Coach Vrabel’s error turned out to be the worst decision of the Conference Championship games.
Note: For this analysis we use conservative assumptions (i.e. on a successful conversion TEN only gains 4-yards). If TEN had gained more than the bare minimum of yards necessary for a 1st down or if the punt had been returned past the 27-yard line, the break-even point is even lower.
Contact us for more information on how EdjVarsity can help you with your in-game decision making. Sign up for a FREE Trial.
Images via USA Today Sports.