Category Archives: Stats Geek!

When stats aren’t convincing

Interesting assertion from the AJ-C’s Brandon Adams yesterday:

Georgia football podcast: One stat could determine SEC East winner

Beginning of the show: In three of the last four seasons the winner of the SEC East has collected more than 40 sacks as a team.

Now that’s true, as far as it goes.  Missouri cracked the 40-sack mark in 2013 and 2014, while Florida did the same in 2015.  However, the Gators, while leading the East in sacks last season, only managed 31 on the way to winning the division.  So, maybe the key is simply topping the division rather than a specific number.

Except when you go back to 2012, you find Georgia’s 32 sacks were only second-best in the East, behind South Carolina’s 43 (I think some dude named Clowney was a Gamecock back then).

There really isn’t much of a rhyme or reason to this, then.  Which shouldn’t come as a surprise when you think about how chaotic the division has been, in the sense that the favorite never seems to win it these days.  In fact, scrounging around the site, I couldn’t find a single statistical category over the past four seasons that the East winner topped in all years.

None of which is to say I wouldn’t love for the Dawgs to manage forty-plus sacks.  I would venture to say were that to happen, it would be an indication that the defense enjoyed a dominant 2017.  But as far as that guaranteeing a trip to Atlanta in December, who knows?



Filed under Georgia Football, SEC Football, Stats Geek!

Today, in greasing the skids

What do you get if you use ESPN’s preseason 2017 Football Index (I know; just roll with me here) to rank SEC schedules in order of difficulty?  This:

Here are the SEC’s toughest schedule rankings, based on ESPN’s Power Rating Index:

1. LSU and Mississippi State 83.6; 3. Vanderbilt 82.5; 4. South Carolina 80.3; 5. Florida 80.1; 6. Texas A&M 79.8; 7. Alabama 79.3; 8. Arkansas 78.3; 9. Auburn 78.1; 10. Tennessee 77.4; 11. Kentucky 74.6; 12. Ole Miss 73.7 13. Georgia 73.5; 14. Missouri 68.7  [Emphasis added.]

Favorable schedule, loads of starting experience on defense, a starting quarterback and coaching staff with a year under their belts… you can’t deny the stars are aligning for Georgia’s chances in 2017.  I won’t pronounce it a given that the sea has parted, but it’s gonna be a lot harder to excuse another year without winning the division.


Filed under Georgia Football, SEC Football, Stats Geek!

Pythagoras is impressed with Kirby Smart.

If you had “5-7 was more likely than 11-2” as the pick in the Oh, How Georgia’s 2016 Season Could Have Gone Pool, Matt Melton is here to tell you that you may be on to something.

And here are the APR standings sorted by division with conference rank in offensive touchdowns, touchdowns allowed, and APR in parentheses. This includes conference games only with the championship game excluded.

Finally, SEC teams are sorted by the difference between their actual number of wins and their expected number of wins according to APR.

Look who’s number one!  The question is whether that could be chalked up to coaching or luck.  We’ll see what regression to the mean has to say about that in 2017.


Filed under Georgia Football, Stats Geek!

Notre Dame: best 4-8 team of 2016…

… or best 4-8 team ever?

By the way, Mr. Connelly projects Notre Dame as an almost four-point favorite when the Dawgs come calling in September.


Filed under Notre Dame's Faint Echoes, Stats Geek!

When they were bad, they weren’t so bad.

Bill Connelly dives into what advanced stats say the best and worst games of 2016 were.

Ask, and you shall receive. Below is a list of what the stats say were the top 50 games of 2016. I added one slight tweak, though. Along with percentile performances, the stat profiles also include a postgame win expectancy figure that basically says “based on this game’s stats, you could have expected to win this game X percent of the time.” If your postgame win expectancy was 50 percent, that means it was a perfect tossup, per the key stats.

If we’re truly judging the most high-quality games of the year, then in my mind they should be games in which both teams not only played well (per percentiles) but also played almost perfectly even. So the closer each team’s win expectancy was to 50 percent, the better the game.

So here are the top 50 games based on what I’m so cleverly calling the Great Game Score — the teams’ combined percentile ratings minus a win expectancy factor…

Best games make for a fun list, although from a pure entertainment standpoint, my favorite game of last season, the USC-Penn State meeting in the Rose Bowl, only finished fourth, but it’s when you get to the worst games that you get the crack that stings, Dawg fans.

And in case you’re curious, here are the 20 worst games of 2016…

The main surprise for me: no Georgia-Nicholls State! Too close, I guess.

Last season was so mediocre, they couldn’t even do awful well.


Filed under Georgia Football, Stats Geek!

The SEC and yards per play, 2016 edition

Let’s welcome back our old friend, regression to the mean!  Matt Melton returns with his annual conference yards per play analysis.

… Since 2005, I have collected YPP data for every conference. I use conference games only because teams play such divergent non-conference schedules and the teams within a conference tend to be of similar quality. By running a regression analysis between a team’s Net YPP (the difference between their Yards Per Play and Yards Per Play Allowed) and their conference winning percentage, we can see if Net YPP is a decent predictor of a team’s record. Spoiler alert. It is. For the statistically inclined, the correlation coefficient between a team’s Net YPP in conference play and their conference record is around .66. Since Net YPP is a solid predictor of a team’s conference record, we can use it to identify which teams had a significant disparity between their conference record as predicted by Net YPP and their actual conference record. I used a difference of .200 between predicted and actual winning percentage as the threshold for ‘significant’. Why .200? It is a little arbitrary, but .200 corresponds to a difference of 1.6 games over an eight game conference schedule and 1.8 games over a nine game one. Over or under-performing by more than a game and a half in a small sample seems significant to me. In the 2016 season, which teams in the SEC met this threshold? Here are SEC teams sorted by performance over what would be expected from their Net YPP numbers.


I’d like to get excited about some of this, namely the part about Florida overachieving, but looking back on last year’s analysis,


… the depressing truth may simply be that Jim McElwain is a good coach.  Of course, maybe it just means the Gators are overdue to come back to earth.

On the other hand, those stats certainly give every indication that Bret Bielema may be the SEC’s most underrated coach.  Take that for what it’s worth.

In any event, keeping Matt’s caution about the significance of over or under-performing by more than a game and a half in a small sample in mind, what’s particularly striking about the 2016 numbers is that a majority of conference schools posted variations under .100, and half were under .050.  If I didn’t know any better, that looks to me like conference YPP was a pretty good measuring stick.


Filed under SEC Football, Stats Geek!

Another advanced stats dump that explains much about 2016

Bill Connelly breaks down last season’s top running backs’ statistical performances by eleven different categories.  Care to guess how many times a Georgia player’s name crops up?  Let’s just say it’s not a good number for a team that was supposedly built around the running game and leave it at that.


Filed under Georgia Football, Stats Geek!