Tuesday, July 01, 2008

On 3rd Down does the game turn.....

I have often rambled about how I think 3rd down is the most important part of football, and, at one times, I have rambled even less coherently about the interest I have in statistical analysis.

Of course, the first down is the currency of football. It keeps drives alive, defenses on the bench, helps field position, and creates points. You make up long, game-winning drives with long strings of first down.

Well, one ramble ran into each other when I was randomly perusing some game notes from last season, and saw that the Falcons actually had some unpublished 3rd down data buried in these little reports.

I'm looking now at the data from the Tulsa game notes, which includes the whole season except that game. (Its page 55 of this document, if you want to see).

It details how the Falcons did on 3rd down, on offense and defense, depending on how many yards there were to go.

Control yourself, please.

No, actually, something really amazing is buried in here. Let's look.

First, let's remember that we were 8-4 at the point and had very good, though not great offensive numbers. The first thing I looked at helped bear that out.

We made 30% of our conversions on 3-10 or more. I don't have data on other teams, but our opponents only made 15.6% from that distance. And, 30% is about what Temple did on ALL its 3rd downs. And it seems remarkable by intuition.

Twenty times we converted from more than 10 yards--that's right on the ass of being twice a game. And, we were 43.5% from 7-9 yards to go. That would be third in the MAC on its own. 30 times we faced 3-7+ and converted.

You might also know that we didn't exactly light the world on fire on 3rd down, (4th in the MAC in all games), and if you know anything about the law of averages, you are saying "hey, were's the bad part that balances this out."

The answer is that we converted only about 61% of our chances fom 1-3 yards, a testament to an anemic running game. (I perceive that we have struggled with these short yardage situations since we put the spread in, but I don't have data to support it). By comparison, our opponents converted 74% from 1-3 yards on 3rd down.

The chart below gives us a look at the data:

You don't need too much analytical ability to draw the following conclusions:

1. We could be a real juggernaut if we were better in short yardage.

2. We were way better than most from longer distances.

3. We didn't experience as much fall off from 4-6 to 7-9 as you would expect.

So, that's one thing. Simply put, our team converted well at long distances but less well from what should be gimee territory.

Then, something leaped out at me. Our distribution is completely out of whack. Our our third down attempts, 66, or NEARLY 40% were from more than 10 yards. Are you kidding me? I would guess that if I told you that there was a team that ran 40% of its third down plays from beyond 10 yards and asked you to guess that team's record, you'd go a while before you hit 8-4.

Our opponents, on almost the same total number of third down plays, had only 32 attempts from beyond 10 yards. The converse is also true. We had only 36 attempts from 1-3 yards. Yes, that's right, we had darn close to double the 3rd down attempts from 10+ yards to go than we did 1-3 yards.

The only conclusion I can draw is that we performed poorly on first and second down, something that is pretty obvious. I lay that at the feet of our less than stellar running game, but it has to include some incomplete passes too, especially because it appeared to continue even when the running game stabilized late in the season.

I don't have data, but it seems to me that in the early days of the spread, we were 3 and short bunches of times. We were a ball control machine, and very effective setting up those makeable third downs.

Now, where is this all heading. There are, in fact, two possible conclusions from the data, one optimistic, and one a little chilling.

  1. If we can get our RBs healthy, we will improve down and distance on third down, and then also do better on those down and distances and be able to hold onto the ball and generate first downs and the increased offensive firepower we need to compete with CMU. That's with a new O-Line, of course.
  2. Last year's offensive success (and the 8-4 record) is built on something unsustainable, to wit, an excess of luck on 3rd and long that kept drives alive. Can we do that again? Doesn't seem like it? If this number comes back to the mean, will we see our numbers fall?
I don't have an answer for that, but its really interesting data. One last little conclusion. We couldn't stop the run, either. And we gave up 74% on 3rd down and 1-3. Staying out of those situations will help us get off the field, too.

No comments :