clock menu more-arrow no yes mobile

Filed under:

What Advanced Stats Tell Us About UGA Football After Week 3

The good, the above average, and the not-ugly-but-still-room-for-improvement after 3 dominant performances. Also some Jane Austen and T.S. Eliot because this is my column, and I’m gonna chase my bliss.

NCAA Football: Arkansas State at Georgia
Robert Beal demonstrates how to lower success rate allowed and increase havoc rate in one image.
Dale Zanine-USA TODAY Sports

It is a truth universally acknowledged, that football team in possession of a fortune of talent and an easy schedule, must be in want of an opponent-adjusted data set. And while yours and my Georgia Bulldogs are no Mr. Darcy — who can be Colin Firth other than Colin Firth? *pining stare into the middle distance * — they are certainly in want of an accurate representation of their value. Let us go then, you and I, to the column where stats lay like a patient etherized upon a table: What Advanced Stats Tell Us About UGA Football After Week 3.

It’s also universally acknowledged that UGA is a talented, efficient team. There are, however, many good teams each year. So the most relevant question is probably: is UGA not good, but great? In the interest of not literally burying the lede I’ll start with a “Maybe?” and go from there.

The biggest issue in evaluating every team at this point in the season is that we have no context for any result. Did UGA dominate Vandy because they’re good, or because Vandy is awful? (yes) Is Alabama generic Alabama good, or by-gawd-that-man-had-a-family great? (probably the second one) Is Trevor Lawrence regressing, or is his hair slowly attaining sentience and drawing him into increasingly complex existential debates, thereby effecting his ability to throw the deep post? (¯\_(ツ)_/¯, but let’s go with probably) Despite the granularity with which systems like SP+ include historical data to influence their ratings, it is impossible to fully quantify, even with the stats we have, the quality of a team after 3 games.

So at times like this, there are several things we can do to try to unravel a team’s place in the college football hierarchy. Generally, I would refer to these as heuristics, as they allow anyone to learn independently, but at the risk of sounding like a condescending know it all — way past that point, I know — I’ll refer to them as concepts or processes. I take the time here to establish this terminology because I think each term emphasizes the nature of what I’m talking about: not what to think but how to think. The purpose of this column and my podcast will never be predictions. Rather, I (and my co-host, Justin) want to help readers, listeners, and ourselves in becoming more informed and intelligent fans. Last week, I talked about two of the aforementioned concepts in my column on fundamentals and priors. Today I want to look at a few more of these, starting with what I would call filtering.

When I think about how good UGA is at this point in the year, the first thing I do is try to filter them out using a series of tests that we know historically that great teams pass. If UGA doesn’t meet one of these characteristics, they may be good, but they probably won’t be great. This process is a subjective one, and I’ll admit based more on my intuition than any statistical insight. But the point here isn’t to make any final decision on if a team is great, but rather to weed out teams who are not. So just add “in Nathan’s unverified anecdotal experience” to the next sentence. Great teams usually:

  • Dominate not great teams
  • Have one element of either offense or defense they can rely on in a tight game
  • Have, at the least, a reliable quarterback
  • Have a strength on one of the lines of scrimmage

So, if we compare what we know about a team to these standards, we can’t necessarily say if a team will be great, but we can determine teams who don’t make the grade. And happily for Dawg fans, UGA passes each of these tests easily. They’ve atomized each opponent they’ve played this year, they have the best running game in the nation — more on this later, they have a “game manager” QB who is currently completing 75% of his passes, and arguably the best offensive line in the country. UGA may not be great, but we definitely can’t say that they aren’t. And that’s the first step.

After trying to “eliminate” a team from greatness, my next process is always to examine their dominant results. Sometimes, the final score of a game doesn’t tell us much about how close the game was in reality. Fluke plays, poor coaching decisions, and many other factors can lead to outcomes that aren’t representative of the flow of a game. It’s helpful to look deeper into the numbers to see if a team was truly dominant, or just lucked their way into a win. For an example of a dominant performance on Georgia’s part, look no further than the statistics from the Bulldogs latest game:

UGA vs. Arkansas State Box Score
Nathan Lawrence

In our week 2 edition I talked about how predictive success rate, and success rate margin, are to the outcome of a game. UGA had a +32% SR margin on the day. Teams when nearly 100% of their games when they have a SR margin of +30%. We don’t need to look much farther than that to know that UGA should have won this game. But the fact that Arkansas state only ran successful plays 29% of the time — well below the historical average of 45% — gives us an even clearer picture: this was a bloodbath. So, again, we don’t know if UGA is great, but our process still hasn’t eliminated them from contention.

The final, and most useful in my estimation, process we can use in determining early team quality is to rank predictive statistics across the entire FBS. This won’t help us adjust for schedule, but it can give us an idea who is at the very least “doing their job,” even against bad opponents. Unfortunately, we (#sbnanalytics) don’t yet have national rank- OH WAIT WE TOTALLY DO I MADE NATIONAL RANKING SEASON SUMMARIES AND SOON I’LL HAVE RADAR CHARTS. I DON’T THINK MY WIFE WILL READ THIS SO LET ME SAY I’M AS PROUD OF THIS IMAGE AS I AM OF MY WEDDING CALL MY BLUFF SAMANTHA:

UGA Season Statistics Week 3: Offense
UGA Season Statistics Week 3: Offense
Nathan Lawrence

That’s a lot to take in, I know, but one thing you’ll notice very quickly is how many single-digit numbers there are next to rows with the word “rank” in them. That tells us from the jump that this is a formidable Offense. Look a little closer, and you can see that UGA’s run game is more than formidable. #1 in rushing Success Rank, #5 in YPP rush, #8 in standard down Success Rate. No matter how close a game gets, this team has one thing it knows it can do: run the damn ball. So not only do we now know, objectively, that Georgia passes our previous “have one element to rely on” test, we also know that they’ve produced at a comparatively high level through three games. Could that change? Sure, and theoretically, my hatred for Tim Tebow may one day cool from the blue-hot inferno that currently makes up the core of my soul. But until we get new data, we have to rely on what we already know: neither of those things are likely to happen.

So, just as with any question worth asking, we’re left with yet more questions than answers at the end of our exercise. Is UGA great? Maybe, probably, it depends on injuries and luck and a million other quirks of fate. Does Kirby make good in-game decisions? Do freshman continue to develop? Does Nick Saban retire and start a politically active alt-Country band? We can’t answer any of those questions with certainty (Nick Saban is quietly a Blue Dog West Virginia Democrat, I’m just saying). Not very confidence-inducing, I know. But here’s the thing Dawg fans: there are about 130 teams in the country who can’t pass the tests we just put UGA through. So I’m not telling you UGA is going to win the CFP. I’m just telling you there’s a chance.

Anyway, here’s some more stats you degenerates:

UGA Season Stats Week 3: Defense
UGA Season Stats Week 3: Defense
Nathan Lawrence
I’m just saying, it’s big, like a placemat.
UGA vs. Arkansas State Placemat
Nathan Lawrence