Super Bowl 32 was a memorable one. A tight game featuring John Elway and Brett Favre would have made a memorable regular season game, but as a Super Bowl it was spectacular. To me the most interesting thing about the game was how the winning score happened. It was allowed intentionally.
The game was tied at 24. The Broncos began a drive with 3:27 left to play. After a big Elway pass and several Terrell Davis runs, Denver put Green Bay in the Field Goal Choke Hold. Eventually, Denver fought its way to a 1st and goal from the Green Bay 8. A holding call on Shannon Sharpe moved Denver back to 1st and goal from the 18. Another Davis run set up 2nd and goal from the 1 with just 1:47 to play. Rather than allow Denver to run down the clock any further, head coach Mike Holmgren elected to allow the TD on the next play to give his offense a better chance to respond with a TD of their own.
In the wake of my previous five-part analysis of intentionally allowing a TD, I learned what the Internet jargon tl;dr stands for. I promise to make this one shorter. Previously, I looked at situations in which an offense that's trailing by 1 or 2 points could run out the clock before kicking a field goal to win. In many cases, depending on the time, score, field position, and number of timeouts remaining, it makes sense for the defense to allow a TD rather than try to force a stop and a FG attempt.
This time I'll examine similar situations where the score is tied. The considerations are a little different than when the defense has a 1 or 2 point lead. A tie score means that the defense can't be relatively assured of a win in the event of a miss. And given a successful FG to break a tie, a FG in response only re-ties the game.
*Remember that this analysis cuts both ways. Whenever it's a good idea for the defense to intentionally allow a TD, it's also a good idea for the offense to take a knee prior to scoring.*
I used the same basic methodology and assumptions as the previous analysis. Starting with a model of when the defense could get the ball back after forcing a stop and FG based on the number of timeouts on hand, I estimated the chances it could respond to a FG with a score of its own given the time remaining. The plots below show the raw and smoothed plot for the FG and TD rates for teams down by 3.
Next, I estimated the chances of the team on defense winning given a FG failure, which includes either scoring within the regulation time remaining or winning in overtime.
Based on standard FG probabilities, I computed the win probabilities for the force-the-FG strategy.
I also estimated the win probabilities for teams that need a TD to survive in the endgame based on time remaining. This estimate assumes the team in the choke hold will start with the ball near their 20-yard line. (Specifically, I used an average of starting field position between the 15 and 25 to incorporate a larger and more reliable range of data.) The chance of a TD in this situation is different than for when the team was down by only 3 following a successful FG, because the offense will obviously play differently needing nothing less than a TD to survive.
In a nutshell, the win probability of the team in the FG choke hold shakes out like this:
wp|offense makes FG = p(response TD) + 0.5 * p(response FG)
wp|offense misses FG = p(TD or FG) + 0.5 * p(no score)
wp(force FG) = p(FG success) * wp(offense makes FG) + (1 - p(FG success)) * wp(offense misses FG)
wp(allow TD) = 0.5 * p(response TD)
When all the numbers are crunched, the results can be compiled into a graph like the one graph below, which plots the win probability for a defense with two timeouts remaining. The colored lines are the wp for forcing the FG based on the field position and time remaining. The black line is the wp for allowing a TD. Whichever line is higher for the given combination of time and field position should typically be the recommended strategy.
The red diamonds mark the final few plays of Denver's winning drive. Point #1 was a 1st and 10 at the GB 32. This was well outside the choke hold zone. But 2 plays later, an Elway pass to fullback Howard Griffith gave DEN a 1st and 10 at the GB 8 (#2). At this point, GB was inside the choke hold zone, and would typically have been better off allowing a TD. But on the next play Shannon Sharpe was called for holding, putting DEN back at the GB 18 (#3), barely outside the envelope.
Denver gave the ball to Terrell Davis, who took the ball to the GB 1 but ran out of bounds to stop the clock. The next play was the allowed TD (#4). [Although this was a 2nd down and not a 1st down, the clock stoppage due to Davis running out of bounds means that the same analysis can be approximated by adding the time of one play (about 6 seconds) to the time remaining.] It looks like Mike Holmgren made the right decision, especially with Brett Favre, in mid-90s form, at quarterback.
How rare are these circumstances? Not that rare. Just in the tie scenario alone, since 2000 there were 256 1st downs that would qualify, or about 20 per season. Some of those may have been in the same game, so I'd guess it's about game per week where it's worth considering intentionally allowing an endgame TD when tied. There were another 166 1st downs in which the defense clung to a 1- or 2-point lead inside the choke hold zone. That makes about 10% of all games where it's at least worth considering, if not actually allowing, the intentional TD. There were 37 total situations in the 2012 season alone.
Remember the tearful PIT victory in BAL earlier this year? Charlie Batch led the Steelers to a dramatic win over their rivals with a late game FG drive. With the score tied at 20, PIT entered BAL territory at the 2-minute warning, with a 1-10 at the BAL 44 (#1 on the chart below). BAL was called for roughing the passer, which gave PIT a 1-10 at the BAL 19 (#2). The Ravens also had an injury on the play, which by rule forced them to forfeit their final timeout. With none remaining and 1:46 left to play, BAL was well inside the choke hold zone.
Allowing a TD at this point would have more than doubled their chances of winning. Instead, PIT was allowed to burn down the clock to 3 seconds remaining before nailing the game-winning FG as time expired. Even after a 5-yard false start penalty (#3) on the first play of the final series, PIT was still well within choke hold territory.
You might say that many situations like this don't really offer a big advantage. After all there's not much difference between a 6% and a 12% chance of winning. That's one way to look at it. The other way to look at it is that whatever chance you might have, you might be able to more than double it, and in some situations do much better than that. If the football gods offered you a 6% chance or a 12% chance of winning, you wouldn't just say, "Screw it. Let's just go with 6%."
And one more reminder here at the end of the article: If a defense does intentionally allow the TD, a choke hold-aware offense could always take a knee prior to getting to the end zone and set up for the chip shot winner, taking the defense's wp to near zero. That's going to be a risk, and there are a few wrinkles to note.
First, a defense should be aware of when it is about to enter the choke hold zone, so that they can say if there is a 1st down inside this yard line, just let the guy score. Don't even wait for the next play on 1st down --especially if there is no clock stoppage on the conversion because that just costs you another 45 seconds. The offense will not yet be prepared to talk about taking a knee, and can easily be suckered into the end zone. This would have been the case in point #1 above..."Ok men, if they get inside the 20, just take a bad angle and let the guy get to the end zone..."
Second, this approach might help defenders to make a big play. Not having to worry about making a tackle after the catch, they can play tighter coverage and shade underneath the routes, increasing their shot at defending or even intercepting a pass.
Lastly, this analysis is just as valuable to the offense as it is to a defense. These are the exact same situations in which it would be better not to score a touchdown. Teams should have a plan for when to take a knee. They can use these charts or do their own analysis. Either way, they should be prepared, because these situations happen more often than we realize.
Here are the generalized charts for when to intentionally allow a TD when tied. Each chart is for each possible number of timeouts remaining. Wherever possible, I've made assumptions that would favor the conservative approach of always trying to force a stop and FG attempt. Therefore, these charts should be interpreted as the minimum criteria for intentionally allowing the TD. In other words, in ambiguous circumstances, you should lean toward allowing the TD.
The envelope of the choke hold zone shrinks with one timeout remaining, and is very small outside the 2:00 warning.
With two timeouts in its pocket, a defense would only want to allow the TD inside the 20 with between 2:00 and about 1:00 remaining.
With all three timeouts left, it's never a good idea to intentionally allow the TD. And if the 1st down occurs prior to the 2:00 warning, there's the bonus of keeping one timeout for the offense.
One final note: By no means do I claim that this is the final word on the question. It's simply the first draft. It's a framework for approaching the question. Later, we might want to change our assumptions or refine the approach, but this can serve as a foundation.
- Home
- win probability
- When to Intentionally Allow a TD When Tied
When to Intentionally Allow a TD When Tied
By
Brian Burke
published on 1/15/2013
in
coaching,
research,
strategy,
win probability
Subscribe to:
Post Comments (Atom)
"This was well outside the choke hold zone. But 2 plays later, an Elway pass to fullback Howard Griffith gave DEN a 1st and 10 at the GB 18 (#2)."
I think that should be GB 8.
Love this stuff
I was thinking the other day, being English, about the fact that in Soccer Managers/Coaches have started to come from non-playing backgrounds and wondered why that was. I'm pretty sure it is a direct result of players having the ability to not work once they are done playing. They have earned enough money to just retire or do something else. This realistically only became an option for all reasonable players (even in the second division in England) in the late 1990s. The game has made strides in tactics/nutrition/fitness since these non players got into and respected in the game more and more. As a result I wondered when the same might happen in the NFL so that the above analysis is finally used on the sidelines by teams. Every team should have a "Brian Burke" clone on the bench to tell them what the numbers say they should be doing in certain situations. Denver last week being a reasonably classic example of poor management in several different areas. I think that time is coming as more and more players have enough money to not be in the game anymore.
But what happens when the other team's Brian Burke clone knows that your Brian Burke clone says you're going to intentionally allow the TD? And yours knows that theirs knows? The advantage of knowing all this stuff will have diminishing returns as more and more teams take advantage of the information.
I'm sure I said it last time on this, but this is stuff any half-decent Madden player knows. I'm always amazed at how NFL coaches don't seem to have any instinct when it comes to game and clock management.
In addition to reading this article, coaches should be made to play end game scenarios in Madden until they understand the effect of timeouts, running out of bounds, incomplete passes and when to allow TDs/kneel on the ball.
On a similar note, I wish QBs and WRs would understand that in an end-game scenario when you are chasing the game, getting tackled in bounds for a five yard gain is much worse than an incomplete pass and going to second-and-ten. Wide receivers, if you can't get out of bounds and you aren't downfield, drop the ball.
tldr
tldr is another way for people to say they are lazy, and blame you for their own laziness.
They expect information to be downloaded to them like the matrix. screw the people who aren't interested enough to read this
"But what happens when the other team's Brian Burke clone knows that your Brian Burke clone says you're going to intentionally allow the TD? And yours knows that theirs knows? The advantage of knowing all this stuff will have diminishing returns as more and more teams take advantage of the information. "
Then keep going off sides repetitively and pushing the QB into the endzone until the other team realizes they can't finish the game at all until they decline the penalty and take the touchdown.
"Then keep going off sides repetitively and pushing the QB into the endzone until the other team realizes they can't finish the game at all until they decline the penalty and take the touchdown."
??? This makes no sense whatsoever.
What about other unique scenarios? The one from the SF@NE game from earlier this year sticks in my head. The Pats were down 38-31 and went for 4th and 1 from their own 12 late in the game. They failed, giving the ball to the 49ers a mere 12 yards away with 2:20 left. The Pats used both of their TOs and the 2MW, Akers made a FG, and the Pats got the ball back at 1:56 down 41-31. They drove down, kicked a FG with 0:38 left, then didn't recover the onside kick.
In this scenario, I would think the Pats would benefit from allowing a TD. They would still have been down two scores, but being down 14 would have given them an extra 24 seconds, two timeouts, and the 2MW. The timeouts are especially crucial considering the Pats could score a TD to bring it to 45-38, not recover the onside kick, and still get the ball back with only ~50 seconds coming off the clock. With Belichick's history of intentionally allowing TDs, I was quite surprised to see him not allow one in this example.
> ??? This makes no sense whatsoever.
I think Anonymous is referring to deliberately drawing penalties to advance the other team to the goal line. (Though I'd suggest dead ball fouls offsides so that the clock stays stopped.) Because it's zero-sum, there's potential for a scenario where one team will keep dead-ball fouling, and the other keeps declining the penalty.
That said, I'm not sure there's a whole lot to be hand from blatantly intentionally drawing penalties after first and goal.
> With Belichick's history of intentionally allowing TDs, I was quite
> surprised to see him not allow one in this example.
Going further out of the box, another potentially interesting option would be to long snap into the end zone to play for a long pass or take an intentional safety. (Come to think of it, this sounds like a fake punt. Does Zoltan Mesko have the same sort of QB chops that Johnny Hekker showed?)
Sloppy post hoc thinking is that down 9 after the safety free kick and a stop is much better than down 10 after the stop and Akers field goal, but the salient question is really how much you lose trying to convert the downs.
As for the 4th down play itself, the break even point was 5%, and I seriously doubt a fake punt would be a better bet than Brady and co.
Belichick has taken an intentional safety before, against Denver in 2003. The Pats were down by 1 with ~3:00 left. Facing 4th down from their own 1 the Pats took a safety and free-kicked instead of punting from their own 1 and ended up winning the game.
http://sports.espn.go.com/espn/wire?id=1653659
ropke, ian, etc.
Why do people think the nfl "doesn't know this". This article is about what happened in the superbowl. In fact it happened in at least 2 superbowls - and we did have exactly a 'let them score, but the rb tried not to' scenario.
This has been known at least 15 years ago in the nfl.
I'd say it is almost a moot point now, seeing as a team knows not to take the TD (i.e. ny giants in SB), but to instead just to the 1 and kneel. Then keep kneeling and kick the FG on the last play.
Anon - all well and good, but clearly the majority of the NFL don't know it. Ahmad Bradshaw didn't know it. Tom Coughlin didn't (why even bother running plays when you are on the 6 yard line and you don't want to score a TD? Just kneel on it).
The amount of bad clock management we see, the amount of coaches who appear completely unprepared for any slightly unusual situation, I don't think 'the NFL' knows it. Some coaches do, this is true. These are the ones who reap the rewards. As for the rest, they should thank Brian. He's given them a competitive advantage (or at least, removed their disadvantage) for free.
Ian.
I know it.
You know it.
Anyone who watched the Superbowl in 1998 knows it. It is actually occurred in two superbowls. I have seen better kept secrets.
Also, whether the NFL knows this or not (they do) is beside the point.
Regarding the alleged "competitive advantage". Has the team giving up the TD intentionally ever won the game? (i know it is stated that there is a 12% chance, but has it ever actually occurred?)
The broncos won their SB in question, and the Giants won their SB in question. So, they are 0-2, which is statistically not inconsistent with a 12% chance. Neither is it inconsistent with a 5% chance.
Can you model the case where you are down by 1 point. How much time needs to be left to play for the stop vs tricking the opponent into going up by 8 quickly?
"Regarding the alleged "competitive advantage". Has the team giving up the TD intentionally ever won the game? (i know it is stated that there is a 12% chance, but has it ever actually occurred?)"
Why does that matter? It's totally irrelevant to the discussion. There have probably been situations where one team scored a touchdown and SHOULDN'T HAVE, but did anyway, even though the defending team TRIED to stop them.
I know for a fact that there was one Vikings game I watched in 2010, a loss to the Jets, where Favre threw a pick-6 to a Jet defender with 1:30 remaining. Took the score from 20-22 to 20-29. Down 9 points, the chances of the Vikings coming back and winning (recovering an onside kick in the process) were minuscule... but they weren't zero. Had the Jets knelt on the ball without scoring the touchdown, they would have been GUARANTEED a victory. It didn't end up hurting them of course, but they did give the ball back to the Vikings.
The point is, it doesn't matter what your reasoning is prior to intentionally (or unintentionally) allowing a touchdown. The only question is, how many teams have won given situation X, and how many have won given situation Y? Regardless of how they got there, teams are more likely to win if they get the ball back down 7 with ~1:30 left to play than if they allow their opponent to attempt a ~30 yard field goal as time expires with the score tied.
"Has the team giving up the TD intentionally ever won the game?"
"The only question is, how many teams have won given situation X, and how many have won given situation Y?"
lol. Umm, that was my question. I don't know why you think it is totally irrelevant.
In a related vein, should Pete Carroll have called an intentional non-TD (not a kneeldown but a running play in which the. RB is instructed to stop at the half yard line if necessary) on first and goal with 40 or so seconds left?
this topic brings up another interesting point: team goals vs player goals.
And you do see this all the time, a player trying to make a huge play to be the hero (looking at you rahim moore) instead of doing the mundane task of fulfilling his teams plans.
For a team, there is a small benefit to executing the 'no td' rule in these situations, however, for the RB who can score the "clinching TD" at the end of a SB, it means a ton of glory (which means bigger contracts, more endorsements, etc) which is obviously of a huge benefit to the player. So the individual makes the decision - do i take the TD for myself, or shun it in what is most likely a meaningless increase in the chance to win.
I can see players taking the glory in spite of their coaches orders, especially in playoffs/superbowl games.
@Anonymous: Who makes personnel decisions? GMs with input from coaches. If a player kneels at the 1 they'll know the player either executed an unusual strategy correctly in the heat of the moment or made a smart play independently. If a player scores a touchdown and the coaches aren't happy about it that surely won't look good to them.
GB 8, yes thanks. Fixed. Also fixed the two "its" that should have been "it's."