Who are the SteezBros?

Welcome to SteezBros! SteezBros is a Steeler Blog run by three brothers who are all huge Steeler fans. If the true definition of fan is fanatic, we fit the bill. We were born into Steeler fandom. Love of the Steelers goes back many generations in our family. It's in our blood. You can read our "Welcome to SteezBros" post in the archive if you are interested in more information about us and the blog. Thanks for reading and check back often!


Monday, September 28, 2009

Right or Wrong: Analyzing the Decision to Go for It on 4th Down

Mike Tomlin is taking some heat for the decision he made at the end of the second quarter in the 23-20 loss to the Cincinnati Bengals. Leading 13-0 with 1:15 to go in the first half, the Steelers faced a 4th down and 4 from the Bengals 35 yard line. Tomlin decided to go for it. The conversion attempt failed and Cincinnatti capitalized by driving down the field and scoring a field goal at the end of the half.

Writers from the Post-Gazette are calling it a terrible decision. But we shouldn't be surprised. Members of the media often defer to the standard NFL thought that going for it on 4th down is always risky and taking the field goal or punt is the safe - read correct - decision. If you watch football at all, the announcer almost always prefers the safe play and questions the coaches decision making when they go for it on 4th down.

The coaches usually follow suit, choosing punts and field goals over fourth down conversion attempts. ESPN writer Gregg Easterbrook often explains that coaches choose the "safe" option on 4th down because it saves them from criticism. If a 4th down attempt fails, the media blames the coach. If the team punts and the defense can't hold, it's the players fault. As we're seeing today, this sentiment probably has some truth to it.

But, the announcers and media members are wrong. In certain situations, the decision to go for it on 4th down is not risky. It's playing the percentages. Which is why I want to spend a little time analyzing the decision made at the end of the first half.

If you read extensively about the NFL, you've probably seen quotes that coaches go for it on 4th down less than what researchers say is optimal. According to research by the football statistics web site, Advanced NFL Stats, the optimal decision chart for 4th down looks like this.

If you are interested in the full analysis of how this chart is derived, read the four-part post here, here, here, and here. This is the cliffs notes version. Using real data, the researchers assigned every yard line an expected value of points. If you have the ball on your opponent's one yard line, you have a very high expected point value (because you are one yard from scoring a touchdownn). If you have the ball on your own one yard line, the expected value is actually negative, because it is more likely that your opponent will score before you will. Using this data, the researchers created the decision making chart above. On the vertical axis is yards to go on 4th down and on the horizontal axis is position on the football field. For example, on 4th down and 4 from the opponent's 15 yard line, the team should go for it if it is 4th and less than 4 and should kick a field goal on 4th and longer than 4.

According to this chart on 4th down from the opponent 35 yard line, the Steelers should go for it if less than 10 yards is needed for conversion. It was 4th and 4. The decision, according to the math, was absolutely the right decision.

I work in finance so naturally, I'm biased towards following the math. But, even I recognize that not every decision should be made purely on empirical data that averages every team in the league. Adjustments need to be made for the situation in the football game and specific team strengths and weaknesses.

At that point in the game, the Steelers had dominated both offensively and defensively. The offense was averaging 8 yards per offensive play, double what they needed to get on the 4th down play. The defense was allowing less than 3 yards per play, indicating that even if the conversion failed the defense could prevent the Bengals from gaining the 35 or more yards necessary to attempt a field goal. Additionally, if the Steelers gained just one first down, they would be in position to get a field goal (or touchdown), making it a three possession game. With only one half to go, the Bengals would have had an extremely difficult time making a comeback. Thus, all these signs, along with the empirical data, point toward going for it.

On the flip side, if the conversion attempt failed, the downside was two-fold. One, the Bengals would have a chance to score before the half. Two, the momentum of the game could change on that score. The first scenario is taken into account with the research above. The second scenario is purely qualitative. Had the Bengals went into the half having generated 0 points, 60 total yards, and losing by 13, their mentality would be much different than putting together a last minute drive and cutting the deficit by 3 or 7 points. It may not be captured quanititatively, but no one can convince me it doesn't matter. Thus, there was a definite downside in going for it.

The other two options were to kick a field goal or to punt. A field goal from 52 yards is a 50-50 proposition at best. If missed, the Bengals would have only needed 25 yards to get into field goal position themselves. Taking into account Jeff Reed's struggles last week, the field goal attempt was probably the lowest percentage play.

Punting was the other serious option the Steelers considered. At the time, I thought it was the best option. A punt would have pinned Cincy deep with little time left. Unless something strange happened, the halftime score would have been 13-0. Not an insurmountable lead, but certainly tough given the way the teams were performing on offense and defense.

Knowing what we know now, the Steelers should have punted. Their lead would have been safe at 13-0 and perhaps the Bengals wouldn't have played as hard in the second half. But these kinds of decisions should be evaluated based on the information known at the time. It's similar to blackjack. Of course hitting on 16 is a bad decision when you go bust, but it's the right decision if the dealer is showing a 10. Same situation here. It was labeled a bad decision because the dealer flipped a 10 on our 16. But, given what we knew at the time, I can't fault the coaches for making the decision that research said was the right one.

If only the coaches would have made the "right decision" earlier in the game on 4th down from the Bengals 1 yard line.

1 comment:

Anonymous said...