I can't see where you’re comin' from / but I know just what you’re runnin' from / And what matters ain't the who's baddest / but the ones who stop you fallin' from your ladder.
For a little over four years now I’ve had a summer time hobby of trying to predict plausible performance levels from various QBs for the upcoming football season. I have tried to root these projections as deeply into the bedrock of reality as is possible for a figment of one’s imagination and at this point there is a codex of sorts in the diary archives describing my methods. It’s fun to go back and see what worked and learn from what didn’t. There’s something there, man.
For Devin Gardner 2013 I laid out two stat lines hinging on two sets of assumptions—a reasonable/prudent set, and a ‘sexy’ set. The reasonable prediction: Gardner would complete 225 of 360 passes for 2900 yards, 23 TDs, and 10 INTs. In reality he went 208 of 345 for 2960 yards, 21 TDs, and 11 INTs. There’s a HEAVY dose a good fortune involved there but, hot damn, that’s pretty good. The assumptions here were basically looking at only QB stats and nothing else Devin had shown enough in his 5 QB starts during the 2012 season to perform at the “seasoned veteran QB” level which I think of as an incumbent with 2 years of experience in tow. That's a brutal benchmark, IMO but that's what I measure guys up against. That's what we want them to be.
Anyway, the sexy set of assumptions were:
- Devin has elite talent. I believe this one held. More on that later.
- The O-line would be fine despite the possibility of being “a touch weaker than last year (2012).” Eh boy…
- The offensive scheme would be well tailored to Gardner’s skill set and that of the support around him. This was sometimes true but not consistently often enough for Borges to keep his job.
Ok, so the necessary assumptions for DG to be the second coming of Vince Young vanished into the ether. But those last two assumptions about the support and scheme are really kind of baked into the reasonable prediction too. For my money, the fact that DG put up the numbers he was able to in spite of the glaring flaws of the team is a testament to just how good he can be if the conditions are reasonable.
The fact that there are so many straight-faced questions being asked about Devin Gardner’s incumbency status is ludicrous. Sure, numbers don’t tell the whole story but they tell a good part of it. DG went from being one of the darlings of the 2013 Manning Passing Academy to needing to prove his talent simply because he couldn't compensate for all of the flaws around him last season. He did as well as could reasonably be expected without adjusting for other very real headwinds.
[After THE JUMP: Gardner under the microscope.]
Yesterday I did some quick research on an improved red zone efficiency metric, which generated some discussion on other potential ways to look at a team's offensive productivity. One of the suggestions that immediately intrigued me was points per possession (thanks Gene). This metric is becoming more and more popular in basketball; I'm sure several of us have read a decent amount on this from Pomeroy, Gasaway (formerly Big Ten Wonk), etc., as Brian references their tempo-free stats on occasion. Dylan at UMHoops uses them as well, for those of you who follow (if you don't, you should).
Points per possession would seem like a pretty easy number to come up with. Well, total points scored is easy to find, but number of possessions... not so much (if anybody has a source for this data, I'd love it if you would share). To get there, I looked at all possible ways that a possession can come to an end (once again drawing on things learned from Pomeroy and Gasaway): Punt, Turnover (downs, int, fumble), Score (and thus a kickoff). So to determine total possessions, I used:
Punts + Turnovers + Kickoffs - Number of Games (each team has one kickoff per game that was not the result of a possession).
There are still some flaws with the above:
- I currently have no way to account for a possession that ends at the end of a half.
- A muffed and lost punt (or a Gordon/Vinopal special: pick -> lost fumble) will show up as a possession (turnover), but these probably shouldn't be taken into account when considering offensive efficiency.
- Points scored without the offense's involvement (pick-six, punt/kick return etc) should not count towards offensive efficiency. I am not sure whether or not the NCAA's "Team Scoring Offense -- Total Points" adjusts for this or not.
I welcome any suggestions or further critique.
(And since this is MGoBlog) Well, that's a lot of words, how about a...
|Rank||Team||Kickoffs||Games||Punts||Turnovers||Possessions||Points||Points Per Possession|
|28||North Carolina St.||28||4||20||6||50||151||3.02|
|31||San Diego St.||30||4||18||7||51||153||3.00|
|111||New Mexico St.||11||3||22||4||34||47||1.38|
|119||San Jose St.||12||4||29||8||45||36||0.80|
Michigan is near the top of the list (6th); this is no surprise. Also near(er) the top and of interest to some folks around these parts is Stanford (2nd), along with this week's opponent, Indiana (3rd), and the Buckeyes (4th).
I would love it if this sort of statistic would eventually make its way into the "mainstream." Again, it seems like basketball is leading the charge for tempo-free stats, but there's no reason that we can't look at it for football as well. Perhaps we could lay this up against the dreaded time of possession stat and look for correlation -- or lack thereof. I also think it would be an interesting metric to use along with the work that The Mathlete has done -- we could start to replace some of the assumptions (Top 20 offense, average defense, etc) with data.
As I said above, please feel free to rip this apart and tell me that I'm a flaming idiot, or offer suggestions, critiques, ways to improve.
Note: All original data for the above was collected here:
Pop quiz hotshot, who has the best offense in the Big Ten? If you don't know the answer or want to follow along with some simple stat manipulation, read and find out.
As usual, 12 data points is not enough to draw solid conclusions but if you didn't enjoy making statistical interpretations about college football you probably wouldn't be reading mgoblog.
As everyone knew, going into the OSU game Michigan had the best scoring offense in the Big Ten. Unfortunately that 10 spot we put up drops us all the way to 4th. How do we drop so quickly from 1st to 4th? What it really means is that we are in the 1st tier of offenses and a virtual tie for 2nd. If we had made that field goal (or gotten a safety) we would have been 2nd place in the Big Ten.
So how does the Big Ten stack up? Well, Wisconsin is the best scoring offense in the conference. Penn State, Michigan State, Michigan and Ohio State make up the rest of tier 1. Purdue*, Northwestern, Indiana and Iowa are the tier 2 offenses. Finally, Minnesota and Illinois bring up the rear.
|Points per game||Standard Deviation|
The main point to take away is that our offense was comparable to the Big Ten's offenses this year. Would you have said that last year? The other important thing to note is the standard deviation. Michigan was the most inconsistent of all Big Ten teams. Shocking statistical analysis there. Isn't it a good thing we can look at the numbers to see things we could never have known by watching the games?
Cupcakes aren't a high fiber diet
Again as everyone knew, part of that number 1 ranking was built out of baby seal carcasses. Michigan wasn't the only team that played a cupcake though. How can we adjust for these blowout games?
Well, one possibility is to look at performance against average points allowed. However, this takes some work and is already covered in great detail by The Mathlete. I prefer a quick and dirty approach. We take out the high and low score for each team to get more of a sense of what the consistent performance of the offense is.
|Adjusted PPG||Std Dev|
The only change in the adjusted points per game is that Michigan drops from 4th to 6th. It still remains in tier 1 though, along with Wisconsin, Penn State, Michigan State, Ohio State and this time Purdue. Northwestern falls more in line with the tier 2 offenses along with Indiana and Iowa. Minnesota and Illinois are still tier 3, although Illinois should probably be it's own tier 4. By the way, who wants to guess how many of the high scores that got eliminated were scored against Michigan?**
What does this all mean?
Everybody will have their own interpretation of these stats. When combined with the eyeball test, I think that it means our offense has made a lot of improvement over last year. It's not quite the offensive juggernaut we hope to see soon, but a lot of that could be explained by a true freshman QB and Molk's absence. We'll see how much more they improve next year, but I think there is a real reason for a lot of hope on the offensive side of the ball.
* Purdue is hard to judge because it is basically in between tier 1 and tier 2. There is a gap between the tier 1 teams and Purdue so I made them tier 2. I probably should have included the Boilermakers in tier 1 though, as we'll see in the next section.
** Trick question. Surprisingly, only Wisconsin scored their season-high against us. Although, Illinois and Indiana came within a touchdown of their season highs when they played us.
|School||Drives||Red zone %||Rank||PPT||PPT rank|
|Ohio State||18||83 %||57||4.67||72|
Well, hell. What do you make of that? I can't see much of a pattern there at all. I guess maybe the teams that play more of a smashmouth style are higher up on the list? I'm willing to chalk this up to limited sample size and uneven competition, and just come back to this in a few weeks. It does make me question how much sample size and opponent quality affected the defensive numbers as well. What do you think?
weighted percentage = (red zone TDs + 0.5 * red zone FGs) / total red zone trips
Yes, I know that a TD usually winds up being worth 7 points, but a 2:1 value for TDs vs. FGs seemed like a good starting point. Why did I bother doing this? Well, mostly just to see if numbers justified my perception that regardless of how the defense as a whole plays, it's really tightened up inside the 20s. How do we measure up? Well, I put the whole Big 10 on a...
|School||Drives||Red zone %||Rank||Weighted red zone %||Weighted rank||PPT|
|Wisconsin||13||92 %||102||81 %||110||5.46|
|Michigan State||12||100 %||111(t)||88 %||113||5.92|
|Ohio State||5||100 %||111(t)||90 %||118||6.40|
Note - "Red zone %" and "Rank" are the defensive numbers straight from the NCAA website, and the weighted numbers are mine.
So what does this mean?
It's still early in the season, but we can start to see a few things.
- First, I was right - Michigan is near the top of the conference, and has done a pretty good job of keeping folks out of the endzone when they get inside the 20.
- Holy hell, MSU. If we get in the red zone, we should get points - probably 6 of them.
- Iowa's interesting - every red zone trip, they've given up a score. However, they've done a damn god job of limiting people to field goals. (Admittedly, that's on only 7 drives.)
- Penn State's been pretty darn good, allowing a TD on only 25% of their red zone trips.
- Before you start gloating about OSU being at the bottom of the list, look at the number of drives. That's right, they're allowing an average of 1.25 red zone drives per game. Of course, they gave up touchdowns on almost every trip, but small sample size blah blah.
- Virginia Tech checks in at #4 nationally, allowing TDs on only 4 of their 17 defensive red zone trips. They must put something in the water in Blacksburg, cause that's ridiculous.
- Oklahoma and Florida are #2 and 5, respectively. It's just not fair to put defenses that tough opposite offenses with the kind of firepower they have (assuming their QBs are healthy, anyway.)
EDIT - I added in the "PPT" column. This is the "points per red zone trip" metric discussed in the comments, and it's what I used for my red zone offense post here. This didn't change the rankings too much - if I reranked based on the new metric, Wisconsin would leapfrog Northwestern and Illinois by a slim margin, but that's it. Also of interest is that using the PPT metric, OSU gets jumped by Arizona and Louisiana-Monroe, leaving the Bucks dead last in NCAA D1A. That makes me smile inside, even if it is just an artifact of a small sample size.