Mark Simon also contributed to this article.
Two days later, Matt LaFleur’s decision is still inspiring discussion for which there is debate over what exactly the right choice should have been.
Trailing 31-23 with 2:09 remaining in the fourth quarter of Sunday’s NFC Championship Game, the Packers faced a fourth-and-goal from the Buccaneers' 8-yard line after an incomplete pass (in a potential scramble situation) on third down.
The Win Probability tool that Sports Info Solutions provides to its clients gave the Packers a 12.7% chance of winning if they went for it and a 9.4% chance of winning if they attempted a field goal. Note that that is specific to a Packers-Buccaneers matchup. In a matchup of an average offense versus an average defense, going for it was an even more clear choice (17.9% chance of winning by going for it versus 12% by kicking the field goal). The Buccaneers' offense and defense being as good as they were (fourth in Points Earned and first in Points Saved) narrowed the odds a bit.
Head coach Matt LaFleur opted to have Mason Crosby kick the field goal. Crosby converted. But the Packers' defense didn’t come through. Green Bay failed to get a stop on a pair of Buccaneers third-down attempts. Alas, LaFleur now finds himself the subject of considerable second-guessing from those who think that Aaron Rodgers could have put his team in the end zone on fourth down.
Not every win probability model showed LaFleur’s choice to be as clear of a decision as the SIS model.
ESPN’s model slightly favored going for it and analytics writer Seth Walder noted that taking into account the quality of the Packers offense might have tilted it even a little more in that direction.
We preferred go, but more narrowly than I would have thought.
WP go: 10.0%
WP kick: 9.5%Needed a 21% chance to justify going for it. League average is 23%. But the Packers are also better than league average.
— Seth Walder (@SethWalder) January 24, 2021
Why was it a close call to begin with?
Consider that the Packers were going up against a Buccaneers offense and defense that rate great by just about any measure.
“The trap we fall into is equating tying and winning,” said Walder. “If they go for it, they need to convert it, then convert the two-point conversion, then stop the Buccaneers and win in overtime. The other way is to kick the field goal, stop the Buccaneers, drive and score a touchdown to win.”
For reference, ESPN and SIS include similar inputs in their win probability calculator equations. Commonalities between the two models include:
Time Left
Score Margin
Which team has the ball?
Where is the ball?
What’s the down and distance?
How many timeouts does each team have?
The models differ in that ESPN factors in a pre-game win probability for each team whereas SIS uses a relative strengths measurement of the team’s offense and defense.
Another source of in-game win probability information, DeckPrism Sports, which supplies in-game odds to sports betting operations, had win probabilities with two differences from the other sources. It had a much higher win probability for the Packers whether Green Bay attempted the field goal or went for it and it also calculated LaFleur’s optimal choice being to kick.
Green Bay gains 2.5% win chance, 27.5% vs 25%, by kicking a field goal 4th and goal from 8. Been awhile since FG was correct, but Lafleur nails it.
— Matthew Davidow (@DavidowMatthew) January 24, 2021
Lafleur's job as a game manager is to give his team the best chance he can. He did so here, very clearly. 2.5% well beyond Deck Prism margin of error. This was obv a very hard one to get right (chk the replies). That's ‘nailing' it in my book.
— Matthew Davidow (@DavidowMatthew) January 25, 2021
If GB went for it, we had their chance to score a td at 36.5%. Its all close, and depending on play call options I wouldn't call it a ‘mistake' to have gone for it. Unlike the usual 3% errors in the 1st half, or McDermott's love of kicking tonight.
— Matthew Davidow (@DavidowMatthew) January 25, 2021
Our live odds would have had it breakeven at the 4.5 yard line. ie kick from the 5, go for it from the 4. A good coach could surely improve that from knowing play call options.
— Matthew Davidow (@DavidowMatthew) January 25, 2021
Davidow would not elaborate on how DeckPrism’s model works (understandably so, since it’s a proprietary model). Giving the Packers a 25% or 27.5% chance of winning requires a level of optimism that neither the SIS model nor the ESPN model were willing to give. In fact, the SIS model wouldn’t even have given the Packers that high a win probability if Green Bay had the ball on the Buccaneers' 1.
Packers Win Probability–Going For It On Fourth Down–By Yard Line (SIS Model)
| Yard Line | Chance of Winning |
|---|---|
| 1 | 20.9% |
| 2 | 17.8% |
| 3 | 17.0% |
| 4 | 15.9% |
| 5 | 14.9% |
| 6 | 14.2% |
| 7 | 13.4% |
| 8 | 12.7% |
Conclusion
So, while it is true that the Packers would have had a stronger incentive to go for the touchdown if Rodgers had scrambled on third down, the SIS model indicates that the Packers still had an uphill battle, which relates back to what Walder said:
Scoring a touchdown would’ve still required a two-point conversion to tie, and even if the game was tied, Tom Brady’s Buccaneers would’ve been much better positioned to kick a field goal to win the game in regulation. Moreover, even if the Packers were to have everything go right, including getting a stop and forcing overtime, the game would still just be a 50-50 proposition at that point.
However, kicking the field goal was going to lock in the Packers' win probability at around/under 10%, and with two minutes left in a football game where your team is behind, there aren’t going to be many chances to increase that number. The clock was the Packers' enemy just as much as the Buccaneers were, and this is why they had to take the risk—even if failing would’ve plummeted their win probability to near zero—because it was their last, best chance to embrace variance.
Teams can afford to sacrifice win probability when they are ahead to minimize variance. In other words, it’s okay to punt on fourth-and-2 when you are up by 17 in the second half because the clock is your friend and you want to keep the win probability swings to a minimum, even if your win probability goes from 80% to 79% by punting. While this sort of behavior is debatable, it is certainly understandable.
When you are behind, the opposite is true. The later in the game it gets, the more that risk aversion becomes a weakness. Instead of giving themselves a chance with the MVP pulling the strings, they lay down.
In a game where Rodgers outplayed Brady—who threw three interceptions—but the Packers had fumbling issues, coverage breakdowns, and pass protection problems, the juxtaposition of the aggressive decision at the end of the first half by the Buccaneers (which added 1.8%) with this conservative one by the Packers might have been enough to tip the scales in this matchup.