While we are the leader of staffing analytics in football, one of "side jobs" we enjoy tackling here at MAS is being a devil's advocate for our coaches. They are constantly being peppered with on-field ideas about managing a football game through analytics and do not have the time nor resource to check every suggestion that comes to them. We do, and we enjoy breaking down the hypothesis of these suggestions.
One of the very broad hypothesis game management tools frequently noted is trying to score on a two-point conversion attempt. In taking on this concept we felt that we needed to try and answer 3 simple questions.
1. What is the average 2 point conversion value versus and extra point
2. What is the frequency of a 2 point conversion effecting the outcome of a game
3. What is the winning percentage of head coaches in games with an outcome effected by their successful or failed 2 point conversion attempts)
Let's start with the absolute basics. Every year since 1974, the expected average value of an extra point attempt has exceeded the average value of a two-point attempt. Every year. Rather than go back year by year for half a century, lets take a look at the last couple of years when we have seen the narrative of two-point conversions analytics being "more valuable" than extra point attempts and the conversion attempts rising.
The Data - 2021 through week 3 of 2023
EXPECTED POINTS PER ATTEMPT
2021 - 115 total 2 point conversion attempts and 51 were successful. That equals a 44.3% success rate or .886 expected points per attempt. Extra Points averaged a 96.9% make rate for a .969 points per attempt.
2022 - 312 attempts. Nearly double that of 2021 with a success rate of 134 out of 312 or a drop to 42.9%. That equals a .886 expected points per attempt. Extra Points averaged a 97.1% make rate for a .971 points per attempt.
2023 - Through 3 weeks of FBS v FBS games there was 87 attempts (on pace for over 350) and the conversion rate was 34 of 87 or 39.1%. This is the lowest success rate number in the last 10 years through week 3 with a .781 points expected per attempt. Extra point success rate was down to 96.1% through week 3 with an expected point per attempt value of .961.
The gap in expected points grew in 2022 relative to the 2021 numbers and now in 2023, in the easiest part of the schedule, we are see the further regression of two-point conversion value relative to extra point attempts.
FREQUENCY OF GAMES
1 point. That is the difference in a successful two point conversion attempt versus a made extra point. Obvious, I know but I want to emphasize this because in order to have a two-point conversion affect game outcome, you need a close game. Since 2021, there have been a total of 1672 FBS vs FBS games. Out of those 1672 game, 68 were one point game or 1 in 25. Can you, can I, can the head coach guess which game is going to be close enough to even utilize a two-point conversion strategy? No, of course not. Therefore, most successful and unsuccessful conversion attempts have zero effect on game outcome. You are chasing a ghost, trying to determine which game outcome is going to need a two point conversion to win it prior to the 4th quarter.
WINNING PERCENTAGE OF GAMES INFLUENCED BY TWO-POINT CONVERSIONS
Of those 68 aforementioned games since 2021, only 9 have had a two-conversion attempt clearly have a direct effect the outcome of the final score.
2021 - In the games with 115 attempts there were a total of 5 game outcomes in which the success or failure of the attempt impacted the game. Of those 5 games the record of the team that attempted the conversions was 1-4.
2022 - 4 games out of 756 were effected by a total of 312 attempts and the record of the attempting team was 1-3.
2023 - Zero. 87 attempts and none of them, successful or not affected game outcomes.
Total Impact Record - Since 2021 through today that gives Head Coaches a 2-7 record in utilizing two point conversions to effect game outcome.
Going forward we fully expect conversions rate to drop as defenses become increasingly prepared for the attempt. Even though we believe there is no analytics edge in attempting a 2 point conversion, it is normal with analytics edges that they regress in effectiveness by the change in behavior and response to using them.
CONCLUSION
We have one more step in our analysis and that is comparing success and failure rates to the situational quality of the offense versus the defense. Our research in teh NFL is VERY CLEAR that conversion rates correlate with the quality of the run and pass offense vs the quality of the run/pass defenses.
We currently feel that given the data and trends since 2021, attempting a two-point conversion in any situation not dictated by score and time pressure has a negative expected value to scoring and winning.
Dave Bartoo
Managing Partner
503.504.8710
コメント