A more interesting tidbit from the new rules, from my perspective, is the introduction of the one-point safety. It is now theoretically possible to see a game with a score of 6-1.
Team A scores a touchdown (6 points) and attempts to kick an extra point. It is blocked. Under the old rules, the play is over; under the new rules, either team can pick it up and run. Team B does so, picking it up and running downfield. Before reaching the goal, the Team B player fumbles the ball and a chasing Team A player picks it up. The Team A player is surrounded by Team B players and attempts to evade them, but only succeeds in accidentally backtracking into his own end zone, where he is tackled.
The result of the play is a safety, which is usually worth two points - but since this is a two-point conversion attempt (after the kick is blocked), the safety is only worth one. So the score is 6-1 in favor of Team A.
The odds of such a crazy play are infinitesimal. One-point safeties have happened a few times in college, where it's been legal for decades, but have always been scored by the offense. I am eagerly awaiting the first one-point defensive safety.
I've always thought coaches were far too conservative by never going for 2 and punting on almost every third down. There's actually a high school coach that starting using a more aggressive strategy. They never punt and always onside kick
https://www.youtube.com/watch?v=AGDaOJAYHfo
Punting on 3rd down would be overly conservative since you get 4 downs. I've always thought the NFL was a little to conservative. But then at the college level I am a fan of Georgia Southern. The Triple Option coaches(Paul Johnson at Ga Tech, Jeff Monken at Army...both formerly of Georgia Southern) their philisophy is if with that type of offense, if you can't move the ball 2-3 yards a play either with a Fullback dive, a QB keeper/midline option, or pitching it to an A-Back. Then we have bigger problems than just turning the ball over.
This definitely works at the highschool level, but becomes much more risky and nuanced in major collegiate BCS-level games, or especially in the NFL. On-sides kicks are incredibly difficult to recover when your opponent has been coached to be ready for them. We have a "hands team" for this specific reason.
Making things even more difficult, there is game film available to pretty much all college/NFL coaching staffs within hours of a game's completion. If you do stuff like this, your opponents are going to be ready. I know we (Clemson) are always ready for Georgia Tech and Paul Johnson to pull stuff like this, because that's what he does. They're able to make bank on this against lesser opponents, but it doesn't always fly against an equally matched (or better) foe.
Are you talking about trying to convert on 4th and short or the offense in general? Because it works well enough against better opponents. (Ask Florida who lost to a "lesser" team with the option, or Georgia Tech who you would think would be ready to for the option got taken to the wire by Georgia Southern).
If you are fan of the nuts and bolts of offenses then the 1990s Paul Johnson Option playbook is really interesting. Especially the section about the quarterback and the attributes of a successful option quarterback.
http://fastandfuriousfootball.com/wp-content/uploads/college...
GT and Paul Johnson still punts and goes for the PAT pretty often. It's very situational, though he does go for it on 4th far more than most BCS-level teams. He's not doing what the aforementioned highschool coach does, though.
Also, I'll take the bait and point out that Florida has been pretty bad for a number of years now. They do get ranked highly every year because of their success the last decade+, but they haven't been a real contender for some time now.
Yeah 4th down is where Paul Johnson likes to roll the dice. Although I'm not sure he would consider it rolling the dice. being able to pick up 2-3 yards a run on the Fullback dive is the cornerstone of the offense. Establishing the dive is the first stage of the flexbone spread option offense.
As to Florida they were still a Power 5 sec school with SEC level talent. They also had one of the better Defenses in FBS that year, and were still made to look foolish against a FCS school that quite honestly was not that good that year(24 Scholarship players out for that game due to injuries). To me that is the prime example of why that offense is so good.
It's great when it works, but without the dive you are screwed. In the present era, it's a gimmicky offense that fails spectacularly if your opponent is talented and prepared. You need only look at Paul Johnson's bowl record at GT to see what happens when a more evenly matched team is prepared for him. He victimizes schools that only have a week (or less) to gameplan for him. He only got his second bowl win in a good while last year, after many failures.
Again with UF, I am assuming you are talking about 2013 when they lost to GA Southern. The problem is, they lost to a lot of teams that year (4-8). They were a fundamentally bad team in the midst of a five game losing streak when they lost to GA Southern (not to take away from what the Eagles accomplished).
That GA Southern loss wasn't because of their offensive scheme. It was more about a team feeling that they failed fantastically, that some of their coaches were going to be removed, and that it's going to be seen as a lost season. The only two offensively strong teams they played wore them out (Missou 36-17 and FSU 37-7). Throw in a loss to an awful Miami (21-16) and an even worse Vandy (34-17!?!?).
I've never liked Expected-Value type calculations because they don't take variance into account. You could have a theoretical messed-up lottery where the payoff is greater than the odds, where EV would tell you to blow all your money on the lottery even if the odds are totally against you winning. Similarly, this wouldn't really imply that coaches should go for 2 all the time (like... you've just scored a TD to tie the game at 6 at the end of regulation, and a tie would mean you'd miss the playoffs).
There's a cool equation called the Kelly Criterion to solve this but it only works for single bets. I don't know how to use it to compare between two bets, or for a case like the extra point where you are essentially betting nothing for a 92% chance to get one point.
You may have missed it in this article, but he does address variance (in the section explaining external incentives coaches include in their decision-making).
I believe you are rephrasing the oft-repeated adage, "The [stock] market can stay wrong longer than you can stay liquid."
- I see that the defense can now score two points by intercepting a two-point play and returning it. Doesn't this lower the utility of the two-point play as used in the article? Or have I misunderstood?
- The article assumes the probabilities won't change. It isn't hard to imagine that teams will get better at executing two-point plays if they attempt more. Also, am I right in thinking that they are typically attempted by a team in a desperate position at the moment? If the existing stats show attempts by inferior teams or teams under pressure, they may be skewed.
- Also, since it is now more likely to miss the one-point field goal, you are more likely to fall a point behind the opposition. They score TD and convert, 7-0, you score TD and miss, 7-6, they score TD and convert, 14-6, you score TD, now what are you going to do? You may as well go for the 2-pointer to make up the lost ground, depending on how late in the match it is.
> - I see that the defense can now score two points by intercepting a two-point play and returning it. Doesn't this lower the utility of the two-point play as used in the article? Or have I misunderstood?
It's such an unlikely event that it shouldn't enter into anyone's calculation except in extreme cirumstances (e.g., you only need 1 point for certain victory, so why risk it). I doubt it will happen once per season.
> It isn't hard to imagine that teams will get better at executing two-point plays if they attempt more.
Teams will get better at defending it too. Teams might acquire personnel especially for the play, both on offense and defense.
> am I right in thinking that they are typically attempted by a team in a desperate position at the moment?
Often but not always desparate. Generally teams defaulted to extra points (1 point) and only used 2-point conversions when there was some certainty it would help. For example, if you are ahead by 1 point and there is time only for one more drive (offensive possession) by the other team, an extra point, which would increase your 1 point lead to 2 points, won't help you because the other team can score either 0 or >=3 points (i.e., scoring 1 or 2 points is impossible, except in a very unlikely event). You might as well go for 2 and earn a tie in case the opponent scores 3 (a field goal).
Also, a little vocabularly help so you look smart around the water cooler:
As for your second point, a 2 point conversion is basically just like any goal line play which they practice many times and encounter in games often. So I would say they have maximized their ability for this play and it wouldn't change if they attempted it more.
I agree here, from my observations teams struggle far more from super short yardage (<3 yards) worse than they struggle from ~10 yards.
The lack of space before the goal line limits the routes you can run. The defense clogging the middle becomes easier because there is no deep threat to allocate safeties and linebackers to.
I think the more 2 point plays that are run, defenses are going to get better at stopping them. I also wonder what percentage of successful 2 point plays were fake kicks, where special teams defense was out there. Rather than actual defensive starters. I wager that if offenses have to face real defenses in this situation they will perform worse.
> I also wonder what percentage of successful 2 point plays were fake kicks, where special teams defense was out there. Rather than actual defensive starters. I wager that if offenses have to face real defenses in this situation they will perform worse.
Fair point, but I don't think I've ever seen any team at any level of play attempt a true fake FG on an extra point. There is, however, a related concept called a "Swinging Gate"[0] that's relatively prevalent at the high school level, rare at the college level, and almost completely unused in the NFL. Interestingly, this rule change completely eliminates the possibility since teams won't be able to reset into a traditional field goal formation at the 2 yard-line.
> I see that the defense can now score two points by intercepting a two-point play and returning it. Doesn't this lower the utility of the two-point play as used in the article? Or have I misunderstood?
Probability of a 100-yard interception return is very low.
> Teams will get better at defending it too. Teams might acquire personnel especially for the play, both on offense and defense.
There really isn't a functional difference between a stopping a 2pt conversion and stopping a red zone drive from the 2 yard line. Defenses already know how to do this. It's a matter of play calling, reading, execution, and the raw will of each team at this point.
> Am I right in thinking that they are typically attempted by a team in a desperate position at the moment?
No, not at all. You can reach certain score combinations that favor a 2pt conversion. Say the score is 13-14, and the team with 14 just scored a touchdown. Now it is 13-20. That team might elect to attempt a 2pt conversion to raise their score to 13-22, which would require the other team to score TWICE in to get ahead or tie the game. If they don't make the 2pt, they're still up by a touchdown.
90+ yard plays are extraordinary, so the defensive score will not be a large factor.
The main thing is that they are tweaking the 1 point kicks to be slightly less certain, if there is an imbalance I think they will quickly adjust things again.
Even if the expected point value for a 2-point conversion is higher, you don't need to involve resk-averseness to explain why a coach may still prefer 1-point kicks. The quantity to be maximized is not expected point value of the play, but probability of winning, which can be expressed as
P(success) * P(win after success) + P(failure) * P(win after failure)
where P(failure) = 1 - P(success). (More generally, it's the sum over all outcomes of P(outcome) * P(win | outcome).)
You also should account for the fact that coaches are often considering more than just points, and beyond that are often just irrational. Coaches should go for it on 4th-and-short more often than they do, but getting stopped on 4th down is demoralizing and makes the coach look foolish, which shifts their decision-making. Getting stopped on a 2-pt. conversion might also reflect poorly on the coach, so they are averse to calling that play even if it is statistically the right move.
I find this psychological side of coaching very interesting, and sports are a great venue for examining the psychology of decision-making. A great example of outcome-bias is the sports world's reaction to the Seahawks decision to throw the ball at the goal line in last year's Super Bowl. I would argue it was the proper decision as it maximized their chances of winning the game (it allowed them time for one passing play and two running plays, as opposed to just two running plays). Unfortunately for them, a very low percentage outcome occurred (interception), and that outcome has strongly biased how the play call has been judged.
"coaches are [...] often just irrational. Coaches should go for it on 4th-and-short more often than they do, but getting stopped on 4th down is demoralizing"
If it is demoralizing enough, and the game (or the season, if being demoralized has a long enough half-life) still has plenty of time to go, that may mean the coaches are acting rational by not playing the short-term better option.
Extreme example: suppose a missed two-point conversion makes your team become absolutely worthless for the rest of the game. If so, you probably shouldn't go for a 2-point conversion at 6-0 in front with 30 minutes to play, even if its success rate is 95%.
The criticism of the play is that Marshawn Lynch nearly always is able to gain yardage with heavy contact. If you have a running back that is able to do this, and do this consistently, why would not go for the run?
Lynch is a great power runner, but even he is not automatic to score on two attempts. If I remember right, he was actually 0/4 on the year inside the 2 yard line.
If the Seahawks throw an incomplete pass on 2nd down, they have time left for two running plays. If they run on 2nd down and don't score, they only have time for one more running play. So by passing on 2nd down they get an extra play. Since it would leave them with the same 2 chances to run with Lynch, the 2nd down play can be judged in isolation: the upside is a chance at a touchdown (win the game), the downside is the risk of a turnover (lose the game). I am making these numbers up, but that is probably a 30-50% shot at a touchdown and a low single digit percentage of a turnover. Worth it.
To expand on the fateful play itself, the odds of an interception on a short inside slant with a screen are very low. Browner jammed his receiver effectively which allowed Butler to sell out on jumping the route and then somehow hold on to the ball. Amazing awareness/preparation and flawless execution by the Patriots. Russell Wilson also lead Ricardo Lockette (his receiver) just a bit too much with the throw and Lockette doesn't break up Butler's interception as hard as he could have. If any one of these things doesn't happen, it's probably a touchdown or at worst an incomplete pass.
It brings up another interesting question: many say Belichick was foolish for not calling a timeout to save a bit of time for a desperation drive. But you could also argue he was trying to put the pressure on Seattle and bait them into a bad play call. I think he got lucky. So really, he was the one who made the bad decision, not Carroll/Bevell/Wilson. Once again, outcome bias.
I would argue, from the coach's perspective, the quantity to be maximized is not the probability of winning but the probability of retaining their job. 2-pt conversion calls are are judged harshly when they fail, however unfairly, and it's likely in a coach's best interest to just stay conservative and avoid scrutiny, even if that means playing sub-optimally.
I think the more pressing thing to consider is P(injured player | play call). If you consider the fact that your quarterback is on the field for the 2 pointer, but not the 1 pointer (typically), the slight difference in point expectation is likely not worth the risk to your franchise player.
The risk to the QB is much smaller on a 2-point conversion. Often QBs are injured standing in the pocket, waiting for receivers to get open far downfield. On a 2-point conversion, that situation occurs much less: There are more running plays because of the short distance required (Seattle Seahawks fans might have somthing to say about that), and passing plays generally are over quickly because the receivers can't run more than 10 yards downfield.
Also, as others have said, 1 point is a high return on investment for one play.
On that note, if I were to make one change to the game, it would be to get rid of kick off and kick return. Yeah, it's exciting to watch, but as a player I hate it. It's the most danger you will ever put yourself in on the field. Concussions abound.
Point expectation is a good way to think of it, but consider the expected points of a 2pt (or 1pt) play vs the other 50 snaps a qb takes everywhere else on the field, where the expected points are less than 1.
An integer point difference on a single play is huge.
On the other hand, he is on the field for 55-60 snaps anyway. The additional 4 or so from the two-point conversions isn't going to change his risk profile much.
With 16 games, an extra 4 snaps per game means 64 extra snaps per season, which is a bit more than playing one extra game. That seems like a noticeable risk increase, season after season, as well as increased wear-and-tear which would possibly lead to earlier retirements.
Regardless, if you now try to convert more of your touchdowns, and you have Eli throwing the ball, that's going to result in more 2 point conversions for the other team.
They already practice and develop plays for short yardage. The short down field at the goal line is different than first down conversions, but it isn't all that different than a red zone offense scheme.
Team A scores a touchdown (6 points) and attempts to kick an extra point. It is blocked. Under the old rules, the play is over; under the new rules, either team can pick it up and run. Team B does so, picking it up and running downfield. Before reaching the goal, the Team B player fumbles the ball and a chasing Team A player picks it up. The Team A player is surrounded by Team B players and attempts to evade them, but only succeeds in accidentally backtracking into his own end zone, where he is tackled.
The result of the play is a safety, which is usually worth two points - but since this is a two-point conversion attempt (after the kick is blocked), the safety is only worth one. So the score is 6-1 in favor of Team A.
The odds of such a crazy play are infinitesimal. One-point safeties have happened a few times in college, where it's been legal for decades, but have always been scored by the offense. I am eagerly awaiting the first one-point defensive safety.