Up by 2 points with 3:58 left against the Packers, Steelers head coach Mike Tomlin called for an onside kick. Was this a good decision?
When onside kicks are expected, they are successful only about 20% of the time. But unexpected onside kicks are successful a surprising 60% of the time. I think we can say this particular kick was certainly unexpected. And surprise onside kicks can be the most beneficial when a team is ahead late in the game. Possession is critical.
In this case, had the Steelers recovered the kick, they'd have a 1st down at about their own 40, which is good for about a 0.76 win probability (WP). An unsuccessful onside kick gives the ball to the Packers at the Steelers' 40, worth 0.58 WP for the Packers (which is 0.42 WP for the Steelers.) With a 60% success rate, the overall WP for the onside kick would be:
0.60 * 0.76 + (1-0.60) * 0.42 = 0.64 WP
A conventional kickoff gives the Packers a 1st and 10 at their own 30 or so (28 is the average, 33 is the median). This gives the Packers a 0.46 WP, which is 0.54 WP for the Steelers.
The onside kick is the better decision by 0.64 to 0.54 WP. These estimates are only league baselines, but they suggest it was probably a good call. What's most interesting to me is that a failed onside kick is hardly certain death--a 0.42 WP. There was plenty of time for anything to happen--a stop, a turnover, or a score. And sure enough the Steelers gave up a touchdown but came back with one of their own.
Also, a successful onside recovery doesn't seal the game. The Steelers would still need at least two first downs to clinch the win. Essentially, the Steelers traded 30 yards of field position for the chance to keep the ball out of the Packers' hands.
- Home
- win probability
- Was the Steelers' Onside Kick Smart?
Was the Steelers' Onside Kick Smart?
By
Brian Burke
published on 12/20/2009
in
research,
team analysis,
win probability
Subscribe to:
Post Comments (Atom)
Huh. I would have thought that the likelihood of GB getting a first down, sitting on the ball and kicking a FG would have been too high to justify the onside kick, but I guess not.
I'm not that excited about the 60% number in general. It's not just that they're unexpected onside kicks. They're also generally taking advantage of some alignment the receiving team is doing that is particularly susceptible to onside kick recovery (typically, too many guys standing more than 10-15 yards from the kickoff point).
Tomlin may have seen this advantage here as well--I haven't seen the play from a useful angle yet. But if he just did it due to the game circumstance and game conditions ("we hadn't topped them and they hadn't stopped us."), then we shouldn't assume a 60% chance of recovery.
Considering how awful both defenses were playing in that game, and how well Rodgers has been playing all year, I think the decision is much better than this analysis indicates.
I wonder how many people will get it backwards on Monday morning and say "You don't risk giving a short field to a good offense!"
By the way, the break-even onside recovery rate is (.54-.42)/(.76-.42) = 12/34 = 35.3%
Another thing to consider is that if you take that extra 40 yards of field position (by kicking away) you may not see the ball again,. Whereas if you don't you have a much better shot at one more possession.
At least that crossed my mind in a game with such poor defense.
I thought the total reversal of the commentators' reaction to the decision in principle once it became apparent that the onside kick was in fact just short of being the requisite 10 yards away was particularly interesting. Of course, that is sadly to be expected nowadays, especially with none other than Joe Buck and Troy Aikman behind the mic.
I think Joe and Troy are some of the best (and smartest) out there...not suprising they didn't get it right, though. My first reaction was that it was a bad play - now I think otherwise. I have the benefit of reflection and study, they did not.
I thought it was a great decision! First, to say the onside kick was unexpected would be an understatement. The only GB player within five yards was annihilated. The Pitt player just erred in his execution.
Second, with a short field, it would be much more difficult for GB to score with no time left on the clock.
Bottom line was this: If PITT recovers, they almost certainly win. If GB gets the ball, PITT will have another chance to win.
Alternatively, if PITT kicks deep, one potential outcome is GB scores with no time left on the clock. Tomlin basically removed that outcome from the range of possibilities.
one other thing the new rule doesn't allow for a "do over" when penaly occurs on the onside kicking team. That said I liked the play by Tomlin, by the way the Packer shold hve taken knee on the 3.
If Tomlin were Belichick you would have instead stated that he was a genius for realizing that the Stealer offense was more capable of scoring twice than the Stealer defense making one stop. Tomlin knew that his own battered and decimated Stealer defense was INCAPABLE of stopping a LONG and TIME consuming drive by the Packer offense, and that it would be better to have time of possession in the hands of Roethlisberger. That is why he tried an onside kick so early; to make sure his porous defense was only on the field as LITTLE as possible! At best you get the ball back and get a TD. Tomlin knew at worst if the gambit to regain procession of the ball during the onside kick failed (and it did) and Green Bay scored (which they did) that the Stealer offense would get the ball back with MORE time on the clock, than if they had kicked it deep in to the opposite goal line. Thus allowing more of a chance for his Pro Bowl QB to mount a successful comeback drive. (Which he did.)...... That is why Tomlin gets paid the big bucks and has a Super Bowl ring while WE are all Monday morning quarterbacks!!!!!
@Edward Lee:
Don't mention about field goals to the Packers. =(
The very fact Aikman hated the call made me believe Tomlin was probably right.
But the Anon two before me makes the great point that the no "do-over" rule for kicking team fouls probably skews the data.
But with 4 minutes left, that's far from certain death, I'd be curious to see this analysis if there were only 2 minutes left in this spot.
I also think Tomlin is somewhat vindicated just on the face that this was just one yard from being a successful kick.
A couple notes. To respond to the first commenter, GB would need more than just a couple 1st downs and FG to seal the game. PIT had all 3 TOs, plus would need less than a minute for a leisurely FG drive to take the lead back.
Regarding the rule change to remove the do-overs--this should mostly be accounted for in the data. Most stats services will ignore plays nullified by penalties from their records. In this case, I used all attempted onside kicks for the baseline data. So ones that went out of bounds in previous years but allowed a do-over are included as failures.
Jonathan-thanks for the break-even success rate. That's not very high.
To those who question this analysis because this is only a league-average baseline: I agree, but this is still very, very valuable. You can talk about the mitigating considerations all day long--Steeler defense this, Aaron Rogers that, Crosby can't kick, etc. But these are all adjustments.
What are you adjusting from? Without this baseline, you are making free-floating relative adjustments in mid-air. Intuition only gets you so far, and it's very often misleading--just like it was to Buck and Aikman (and Dungy).
It's like this word problem: How long will it take you to get home? Given:
a. Traffic is light, so it will be a little quicker by about 5 min
b. You're low on gas, and probably need a refill, taking 10 min
c. You're hungry, so a drive-thru might take about 5 min
Well, what's the answer? You'd say "about 10 min longer than usual." Ok then, what's "usual?" 10 minutes? An hour?
*You can't even begin to answer the question without the baseline.*
This is only partially related, but the logic behind Tomlin's decision made me think about a number of topics that are common to this blog. In general, it seems that the blog advocates a more aggressive offensive game plan (more emphasis on passing, going for it more frequently on 4th downs and more onside kicks), than most teams are currently deploying. While I don't disagree that is probably the most efficient way to score points, it seems that style of play might alter the overall dynamics of a game to the point that it may become detrimental. I concede that I do not have any statistics to back it up, but it just seems that games where this strategy is used, like the Packers-Steelers game, turn into shootouts where the winning team is the one that just happens to have the ball last. There isn’t much game management, except for in the last 4 minutes, and each team seems equally likely to win. This contrasts with the strategy that the Cowboys appeared to utilize against the Saints on Saturday. They had a more traditionally “balanced” offense that controlled the clock and kept the ball away from the Saints’ offense (not to mention a 3-0 turnover advantage)- and it worked. I know that underdog strategy has been covered before, and that it was determined that more aggressive play calling is the best chance for an upset, but is there something to be said for tempo? It is probably difficult to measure, but is it possible that it might be beneficial to not be so aggressive for the sake of ball control, keeping your defense fresh, and possibly frustrating the opposing offense?
John H ... I agree with your comments above and have tried on a few occasions to say the same thing. I just hope this "style of play" doesn't end up in the NFL changing its OT rules. I see nothing wrong with putting your defense out first and expecting them to get a stop. The NCAA OT rule sucks if you ask me, but the NCAA is not the NFL, so its ok. I do like the idea Brian proposed for OT where each team "bids" where they would let the other team start and the low bid plays defense first.
KenYonLV
Brian, just to add to your response to the comment that GB just needed a couple of first downs and a FG to win. First, like you said, the Steelers had enough timeouts to make that unlikely (I thought they had 2). Second, although this doesn't fit with a specific statistic, in order for GB to milk the clock like that, they would have had to run the ball effectively, which was unlikely. For all that the Steelers defense has looked like a sieve lately, it's been the pass defense. Teams have largely given up rushing against the Steelers because they can't, and it's so much easier to pass on them.