November 23, 2009
Why Bill Belichick Was Wrong
by Luckbox
I can't believe there's even a discussion about this. Everyone knows that Belichick was wrong. He made an arrogant decision that cost his team a victory. Yet rocket scientists and amateur poker players everywhere are breaking out their protractors to justify this game-losing decision.
Well, I spent about 10 minutes doing my own research. I present: Other numbers.
Over the last 6 weeks of the 2009 NFL football season, teams have attempted to go for it on 4th and 2 on 14 different occasions (including Belichick's dumb decision). Any idea how many times those 14 attempts have been successful? I'll give you a minute to guess.
Yep, it's zero. That's right. A big 0%. Eight incomplete passes, two failed runs, two interceptions, one sack and Belichick's pass that came up inches short. I wonder if Belichick knew how difficult it is to convert 4th and 2 when he made his bone-headed decision?
Anyone who understands football understands the difficulty in converting 4th and 2. It's a little too far to run meaning teams can play for the pass. In fact, Belichick made it even easier on the Colts by emptying the backfield eliminating any threat of a run at all.
Of course, many argue that had New England punted that Manning (because he's so super-awesome) would have lead a TD drive anyway. I wonder what the numbers say?
Over the last 6 weeks of the season, Indianapolis has had 37 drives starting from the 30 or behind and they've converted TDs in 10 of them. That's 27%.
Over the last 6 weeks of the season, New England's opponents have had 44 drives starting from the 30 or behind and they've convered TDs in 5 of them. That's 11%.
What about when New England and Indianapolis play each other? Here are some more numbers.
In this year's game, Indianapolis had 12 drives starting from the 30 or behind and they converted TDs in 3 of them. That's 25%.
Since 2006 against New England, Indianapolis had 29 drives starting from the 30 or behind and they converted TDs in 6 of them. That's 21%.
Of course, Indianapolis got the ball from the 40 yard line or closer twice against New England since 2006 and, predictably, both resulted in TDs, including the game-winner a week ago.
So what do the numbers tell me?
4th and 2 conversion rate last 6 weeks: 0%
Indianapolis' TD rate from 70+ yards vs. NFL last 6 weeks: 27%
New England's opponents' TD rate from 70+ yards last 6 weeks: 11%
Indianapolis' TD rate from 70+ yards vs. NE last week: 25%
Indianapolis' TD rate from 70+ yards vs. NE since 2006: 21%
I don't know about you... but those numbers certainly don't support going for it on 4th and 2. Looks like NE had a chance somewhere between 75% and 85% of stopping Indianapolis on a long TD drive but very little chance of either converting the 4th down or stopping the Colts if they failed on the 4th down conversion. And that doesn't even include the fact that the Colts would have been attempting this drive with just 1 timeout and just 2 minutes left in the game which likely lowers the Colts' percentages.
Of course, lots of people will say I'm cherry-picking. Guess what? That's what statisticians do. They cherry-pick numbers to support their case.
This was a bad decision by Belichick. It's clear that the decision was last-second which only compounded the mistake. Had New England really intended to go for it on 4th down no matter what, there's little chance they throw the ball on 3rd down. Not only would a run have been a surprise on 3rd (giving it a better chance of succeeding) it also would have forced the Colts to burn their final timeout and it would likely have gotten them closer for their 4th down attempt (which, given the spot, would have made all the difference).
When New England failed on 3rd down, most of the offense was running off the field and New England was then forced to burn their last timeout which prevented them from challenging the spot on 4th down (although I'm not sure a challenge would have worked). If the plan was to go for it on 4th all along, wouldn't the offense have known that? After all, they had already burned a timeout in that same drive!
Not to mention, going for it in this situation brings all kinds of factors into play, including crowd noise and the potential of the home crowd to influence any close spot in favor of the Colts.
Of course, context doesn't matter to the people with their fancy equations. Context is irrelevant, right?
The Patriots conversion rate on fourth down against the Colts was 75% over the last four meetings.
Using a stat of "70 + yards" is misleading at best and I think plain wrong. "70 + yards includes all the drives that start inside the 30 as well as inside the 5 yard line which clearly wasn't going to be the case. To make your case as you stated it you need to narrow down the length of the drive from 70 to 100 yards to 65-75 yards, or even 60 to 80 yards.
Honestly with the way the Colts were playing I liked the call.. either way THAT play is not what cost the game IMHO.. It was a lot of turnovers in the red zone that caused this game to even get there. Props for Indy for stepping up there. If Bill wants to go for it there then I support him.
I'm obviously a huge Pats fan. I'm still not decided on whether the call was great, marginal, or horrible.
However, your assertion that the call was made due to arrogance is specious at best.
I had the Colts to win, so it was clearly an excellent decision by the Belly Check.
If you really believe the Patriots had little to no chance of making it on fourth down, then I guess there's no arguing with you.
I don't see how bad offenses going for it in desperate situations needing a 4th and 2 conversion to have a chance to win really compares to this one.
Alan...
That's the best you can do in response to my research? That's pretty weak.
You seem to miss my point entirely. Weren't the numbers you presented also inclusive of bad offenses in desperate situations needing a 4th and 2 conversion?
Aren't all of the numbers chosen by the rocket scientist including contexts that are vastly different from the situation presented?
The point of this post was to clearly demonstrate that the "math" is irrelevant in this circumstance. It was just as easy for me to choose my sample set as it was for the rocket scientist.
It was a bad decision.
That's the best I can do because you won't change your opinion no matter what evidence I present.
I do have one more stat: How many times this season have the Colts needed a touchdown to win in the final minutes of the game? Once.
Manning got it done in 32 seconds.
(Yes, I realize I'm the same way. My opinion is very unlikely to change, either.)
I feel like my point is being missed.
Everyone is free to their opinion, right or wrong. I think it was the wrong decision and I think the result of the game helps support that.
What I have an issue with is someone pointing to some mathematical equation and saying, "See, he made the right decision."
There are lies, damned lies and statistics. You can choose whatever numbers you want to "prove" you're right. I choose my numbers and "proved" I was right.
The math does not settle this no matter how many fancy equations are cited.
Okay, I get your point. In this case, though, that's not what's happening. The equation, without knowing the result it would lead to, came first. Then (what some think are) appropriate values were filled in, and that lead to the conclusion that he made the right decision.
It was *not* a question of looking for numbers that would support going for it. I really doubt the site this all came from (Advanced NFL Statistics) has any vested interest either way in the result. They just presented the math as they see it, not by cherry picking numbers to get the conclusion they want.
And my point is, it's not as cut and dry as "Everyone knows that Belichick was wrong. He made an arrogant decision that cost his team a victory." Clearly, a lot of people disagree with that.
Let's remember where this discussion started. I made a joke about Andy Reid not making a dumb decision like Belichick did.
Shortly thereafter, you tweeted:
"Conventional wisdom is often wrong. http://bit.ly/3um1aB"
Almost immediatley thereafter, BloodyP tweeted:
"@metsfan is right. http://is.gd/51tnz"
Amusingly, it was the exactly same column. Math geeks everywhere latched on to this one guys fancy math to "prove" that Belichick made the right decision.
But let's look at his choice of numbers:
1) His conversion rate is based on "historical" numbers. He never explains it, but I'm going to assume he's using every instance in the history of the NFL of 4th and 2? Is that right? Sounds pretty unreasonable to use the conversion rates of 4th and 2 in 1947 to in any way determine what would have happened today.
2) He says that historically, teams score a TD from that spot on the field and less than 2 minutes just 53% of the time. Really? The Colts score a TD from that spot on the field with 2 minutes left just 53% of the time?
As I point out, there are significantly better stats to look at that apply more directly to the Pats and the Colts. And, when used, these stats make it much more clear that going for it was the wrong choice.
-------------------
And let's put this in poker terms. A football game is a lot more like a poker tournament than a cash game because it's an elimination situation.
When you're in a tournament, is it always correct to make a decision based on the math? Especially when the math is so close?
Of course not. If one decision will more quickly lead to your elimination, it's often correct to make the other decision even if the math slightly favors the former.
Do you disagree?
1) I agree, historical data shouldn't be used. I guess there isn't enough data to get a good idea on their chances on 4th and 2. Many people think their chances were pretty damn good.
2) No, the Colts score from there more than 53% of the time. But they're also higher than the 30% chance of scoring from down the field after the punt. The better you think the Colts offense is, the MORE you should be going for it... if you think you have a good chance to convert the 4th and 2.
As for the poker related argument, you're right. There are times in tournament situations you shouldn't take an edge because of the risk of getting knocked out (some would actually argue with that, but I agree). However (and again, this depends on you believing the math) the numbers presented here are the chances of *winning the game*, not the chances of making it on one particular play (or individual poker hand).
I think it really all comes down to what you think their chances to make it from 4th and 2 are, and the chances of the Colts scoring from down the field.
Regarding the current 0-15 streak on 4th and 2 (ignoring 4th and greater than 2: Lions made it from 4th and 5 today)... Even Albert Pujols has had 0-13 streaks. Doesn't mean you should sit him. Just takes a trip to NY to break him out of it: http://www.albertpujolsclub.com/albert-pujols-news/pujols-breaks-slump/
Bet the Broncos convert from 4th and 2 against the Giants tonight. :)
