Third-and-Long: More Fun With Numbers

Mark Richt was asked for a simplistic answer to a complex question earlier this week, so perhaps it's not fair to really analyze his answer. But hey, what else do we have to do?

The question was this: What do you need to do to turn around the defense?

Here's what Richt's answer was:

“I think when we get people in the third-and-long situations, the possession down, we need to nail it more often. We need to get off the field when we get them there. ... We’ve gotten ourselves in that position on enough occasions where if we could just finish it I think the outcome of a couple of our games would have been different, too.”

Those pesky third downs. Richt talks about them often, and Willie Martinez essentially echoed his comments later in the week.

The lack of success on third-and-longs is demonstrated by the lack of sacks and turnovers, he said. It's been, as Richt said, the defining difference between wins and losses in some of Georgia's games.

But has it really?

Let's take a look at the real numbers. (Percent = conversion percentage.)

Opp.	Third Down	Percent	3&Short	Percent	3rd&Long	Percent
OSU	6-15	40%	5-6	83%	1-9	11%
SC	6-17	35%	4-8	50%	2-9	22%
Ark	3-14	21%	3-4	75%	0-10	0%
Ariz St	4-13	31%	1-2	50%	3-11	27%
LSU	5-14	36%	2-5	40%	3-9	33%
Tenn	5-11	45%	2-3	67%	3-8	38%

The number of third downs forced is a good indication of overall defensive success. If the opponent is facing a lot of third downs, Georgia has probably done well on first and second down.

So what do we find? Well, aside from last week's blowout loss, the overall number of third downs forced has remained essentially the same.

Georgia has played, to be general in our assessment, three good defensive games (OSU, ASU and LSU) and three bad ones (SC, Ark. and Tenn.). But there's essentially no correlation between Georgia's success on third down in the good games with the bad games.

But all third downs aren't created equal. It's those third-and-longs Richt is looking for -- those are the ones where Georgia's defenders can pin their ears back and make the big play. So we might assume in the good defensive games, there were a lot more third-and-longs, right?

Nope.

Opp	Total	3&Long	Percent
OSU	15	9	60%
SC	17	9	53%
Ark	14	10	71%
Ariz St	13	11	85%
LSU	14	9	64%
Tenn	11	8	73%

In Georgia's three best defensive games, 69 percent of all third downs forced were third-and-longs (i.e. third-and-5 or longer). In the bad defensive games? That number is 64 percent -- down a bit, but nothing substantial.

In fact, Georgia's only game with any measurable increase in sacks -- LSU -- didn't approach the success Georgia enjoyed in creating third-and-longs in games against Arkansas and Tennessee, the two worst performances by the defense.

And while Richt and Martinez seem to indicate that Georgia is having more trouble creating third-and-longs recently, the truth is that their two worst games at doing that were their first two. The game in which they created the second-most third-and-longs, meanwhile, was probably the worst defensive effort.

So maybe third downs don't really have much to do with how successful the defense really is, despite what Richt and Martinez have said.

What seems to be more important is how successful the Bulldogs have been on first and second down.

First off, let's look at the number of first downs that the opposition has picked up on a third-down play.

Opp	Total 1Downs	1D on 3rd	Percent
OSU	17	6	35%
SC	26	6	23%
Ark	17	3	18%
Ariz St	14	4	29%
LSU	19	5	26%
Tenn	24	5	21%

In Georgia's three strong defensive performances, 70 percent of all opposing first downs came on first or second down. In the three bad performances, that number jumps up to 80 percent.

Compare that to the SEC's top defense at Florida. For the season, 30 percent of the first downs the Gators have allowed came on third down. That's about the number Georgia is at in their "good" defensive performances. In essence, it's a sign of how successful the team has been at creating third downs or, more to the point, how unsuccessful they have been at holding teams on first and second down.

Of course, a better measure of that would be the number of third-downs created compared to the number of overall first downs -- i.e., how many sets of downs resulted in a third down. Here's Georgia's stats on that. (*Note, first downs include those that began a drive.)

Opp	First Downs	Third Downs	Percent
OSU	29	17	59
SC	38	17	45
Ark	33	14	42
Ariz St	26	13	50
LSU	31	14	45
Tenn	38	11	29

In both of Georgia's best defensive performances, it turned at least 50 percent of first downs into third downs two plays later. Its worst defensive performances were also the two games in which the opposition needed to get to third down the least.

Then let's look at how many big plays come on third down.

Opp.	Big Plays	On 1st	On 2nd	On 3rd
OSU	2	0	2	0
SC	2	1	0	1
Ark	9	7	1	1
Ariz St	3	1	1	1
LSU	4	0	4	0
Tenn	6	4	0	2

Georgia has allowed a total of 26 plays of 20 yards or more this season. Twenty-one of them have come on first or second down (81 percent).

So what does that mean?

It means Georgia's real problems aren't on third down. Martinez said the team's goal is a 67 percent success rate on third down. As it stands, they're successful 65.5 percent of the time -- not quite his goal, but not too far off. In fact, Georgia ranks 35th nationally in stops on third down -- not great, but not awful.

What we can say is that Georgia is simply not very good on first and second down. A lot of this starts with the same problems we witnessed against Tennessee. When teams have a chance to keep their game plan open -- running play action, for example -- Georgia's defense is lost. On third downs, when teams are more restricted in what they can do, Georgia is just as successful as we might expect. In fact, when the Bulldogs do get teams into third-and-long, they make the stop nearly 80 percent of the time.

Georgia doesn't have a problem getting off the field when they get third-and-long. The problem is that the Bulldogs aren't getting to third down enough, and that's less about execution than it is about recognition. When the Bulldogs know what to expect on D, they're good. When they don't, the recognition and adjustments simply aren't happening.

Read Richt's quote again: “I think when we get people in the third-and-long situations, the possession down, we need to nail it more often. We need to get off the field when we get them there. ... We’ve gotten ourselves in that position on enough occasions where if we could just finish it I think the outcome of a couple of our games would have been different, too.”

The numbers just don't support it. Georgia is doing just fine -- not great, but acceptable -- when they get to third down. Richt needs his defense to do better the rest of the time.