A look at ND performance 2008-2010

For those that have been following, I’ve been playing with yards per play differential (yppd) – or if ND gets 6.0 yds/play and our opponents get 5.0 yds/play, we have a +1.0 yppd. I’ve been using it as a way to measure a team’s strength as well as a way of picking winners.

I’ve tried a new cut of the data – looking at ND opponents based on their yppd “score” (I call it a score b/c it’s not only their yppd, but I adjust for strength of schedule). The results look like this:




Basically, above the x-axis is a win, below the x-axis is a loss. I’ve colored Weis and Kelly game differently.

What is interesting to me – looking at this, we look a lot better under Kelly than we did under Weis. Kelly has played a tougher schedule and, almost across the board, has done better across every quality of opponent. Consider:
- 3 of Weis’s 13 wins came across teams worse than any we’ve played under Kelly
- Kelly’s won ~80% of games against average teams (in the 0 yppd bucket), and is 4-4 against good teams (teams in the 1 or 1.5 yppd bracket… these teams typically win 70% of their games)
- Kelly has played 90% of his games vs average or better teams, Weis only 70%

Looking at these numbers, it’s hard not to see progress from Weis, at least for me. We are winning more games against better opponents.

Peeling back the numbers a bit more Kelly’s two losses against average teams (0 yppd) were Navy & Tulsa last year. The two losses against 0.5 teams were UM last year and USF this year. USC is in the 1 yppd bucket this year. UM and MSU are both in the 1.5 yppd bucket this year. Our remaining opponents are from average (Wake, MD are both in 0) to bad (BC is in -0.5) to very good (Standard is in 2).

If we beat Wake, MD, and BC, ND will have won all games against 0 yppd or worst opponents for the first time in 4 years. Which will bring us to Stanford, which will be our 2nd toughest game in this four year span.

One last interesting stat. Against 0 yppd teams, Weis teams scored on average 23-23. Kelly's teams have scored 32-19. Against teams in the 0.5-1.5 categories, Weis's average score was 31-29, Kelly's has been 25-24.

Picking today's winners - yppd fun

I've had a request to put together my "yards per play differential" (yppd) analysis for this week - so I just pulled together some numbers. I continue to evolve how I do these calculations, so it changes from week to week - but once I get to a place where i think it all comes together, i'll stop tweaking the approach.

With the model as it currently is, I develop % chance of winning based on historical performance against a set of statistics. Right now, the statistics I’m using are all yppd based and I’m using three different measures.

Across these three measures, ND is at +1.1 yppd, +2.4 yppd, and +1.4 yppd. These correspond to a 76%, 87%, and 83% chance of winning (the way each of these are calculated, I can apply 2008-2010 data to find historical win %). One estimate I’ve done has ND at 7.2 ypp and USC at 5.7 ypp – suggesting that we essentially have our way on offense and USC can move the ball, but not as well as we can. Some more details are below, but this looks like it translates into a ~35-17 win. Other ways I've crunched these numbers in the past suggest that the game may be closer than that - but in this iteration of how I'm crunching the numbers, it looks like ND should do quite well.

For some context, here are how other teams with a +1.1 yppd have fared (the first team is the one with +1.1).
In 2011, wins:
Marshall 24 Rice 20
Utah St. 63 Wyoming 19
South Carolina 14 Mississippi St. 12
Boise St. 41 Tulsa 21
Texas 17 BYU 16
Troy 24 UAB 23
Oklahoma St. 38 Texas 26
Tulane 49 UAB 10
Hawaii 44 Louisiana Tech 26
La.-Lafayette 20 Kent St. 12
New Mexico St. 28 Minnesota 21

In 2011, losses:
Nevada 34 Texas Tech 35
Middle Tenn. 33 Western Ky. 36
Iowa 41 Iowa St. 44
Central Mich. 13 Kentucky 27
Florida 6 Auburn 17

Ones that were at +2.4 yppd by the 2nd measure in 2011 (only wins listed, no losses occured):
Stanford 44 Duke 14
Oklahoma 47 Kansas 17
Southern Miss. 48 Rice 24
Georgia 27 Ole Miss 13
Clemson 43 Troy 19

Ones at +1.4 yppd by the fourth measure (again, no losses):
LSU 19 Mississippi St. 6
Boise St. 40 Toledo 15
Utah St. 63 Wyoming 19
Utah 54 BYU 10
Rice 28 Memphis 6
Oklahoma 38 Missouri 28
Nevada 17 San Jose St. 14
Northern Ill. 51 Western Mich. 22
Michigan 28 San Diego St. 7
Cincinnati 27 Miami (OH) 0

In short, for the first measure (+1.1):
11-5 this year,
Avg win: 33-19
Avg loss: 25-32
For the second measure, all wins (5-0), 42-17 score.
For the third measure, all wins (10-0), 37-13 score.

I see the % as more accurate than the specific game outcomes because they pull from a larger data set. But right now, it looks like a 35-17 win should not be out of the question. And by most measures, we look to be the better team. Before we get too excited, other measures I’ve done shows ND at more of a 60% chance. But we should be feeling good overall.

Now – for the rest of the league. Last week, this methodology was 40-12 (I may have double counted a game.. I’m too lazy to check). When the chance of winning was 60% or higher, the model was 35-7.

This week, here are the predictions, listed from blowouts to close games (the first team is the one the % is relevant to). Some interesting games in the list, such as Clemson /UNC being close, MSU/UW being neck and neck. Washington giving Stanford a run for its month. Or Alabama blowing out TN.

Institution Opponent Name % chance of winning
TCU New Mexico 100%
Oregon Colorado 99%
Oklahoma Texas Tech 99%
Tulsa Rice 99%
Nebraska Minnesota 97%
Temple Bowling Green 96%
Penn St. Northwestern 94%
Virginia Tech Boston College 95%
Alabama Tennessee 97%
Tulane Memphis 95%
Texas A&M Iowa St. 93%
Arkansas Ole Miss 94%
Vanderbilt Army 89%
Toledo Miami (OH) 89%
Notre Dame Southern California 82%
La.-Monroe North Texas 88%
Northern Ill. Buffalo 88%
Houston Marshall 88%
LSU Auburn 84%
Iowa Indiana 86%
Virginia North Carolina St. 84%
Boise St. Air Force 86%
UTEP Colorado St. 66%
Utah St. Louisiana Tech 82%
Wake Forest Duke 82%
La.-Lafayette Western Ky. 75%
Nevada Fresno St. 78%
Central Mich. Ball St. 74%
Illinois Purdue 74%
SMU Southern Miss. 69%
Ohio Akron 80%
Florida St. Maryland 72%
Kansas St. Kansas 71%
Oklahoma St. Missouri 66%
Utah California 66%
Georgia Tech Miami (FL) 64%
Middle Tenn. Fla. Atlantic 64%
Stanford Washington 64%
Washington St. Oregon St. 63%
South Fla. Cincinnati 54%
Michigan St. Wisconsin 52%
Clemson North Carolina 50%
Western Mich. Eastern Mich. 54%
East Carolina Navy 49%
New Mexico St. Hawaii 50%

...
...

PS Not that I have to say it - but by no means is this meant for betting - I'm just having fun with numbers and there's a good chance my excel model is buggy and this is all wrong.

Predicting remaining W's on ND's schedule

For those that have noticed, I've been playing with a variety of stats this year doing some analysis for fun. I've been a bit facinated with yppd - yards per play differential (e.g. if we average 6.0 yds/play and our opponents average 5.0 yds/play, we are at +1 yppd).

Well, I've decided to expand my model and I've included a number of things, including total yards/game, home field, and winning % - to come up with a predictor model. Essentially, by combining these variables, I've create a formula that correctly 'predicts' the winner to a game approximately 80% of the time (based on data from 2008-2010). For 2011 so far, this formula would predict the winner 85% of the time (likely due to the tougher part of the schedule not yet being played).

The predictor works by having a % chance of winning - 0 to 100%. If it's above 50%, I say it predicts a win, below 50%, a loss.

Some more detail on its accuracy -



[% chance of winning] [% of times right] [# of Games in sample]
0-19 94% 950
20-39 77% 709
40-49 54% 538
50-59 56% 405
60-79 76% 756
80-100 94% 940



So, in the middle of its predictions, it's right about 55% of the time. But is right 3/4 to more than 9/10 once it gets outside the 60% predictions.

Anyway - the point of everything I wrote before this is... the numbers in this model are complex, but are a decent predictor of who wild win.

So - let's apply this to ND's schedule this year. We get:


[Opponent] [Prediction] [% chance]
South Fla. win 64%
Michigan loss 19%
Michigan St. loss 40%
Pittsburgh win 78%
Purdue win 82%
Air Force win 81%
Southern California win 55%
Navy win 88%
Wake Forest win 51%
Maryland win 81%
Boston College win 103%
Stanford loss 19%



BC is above 100% because there's a home field advantage factor that sometimes pushes % above 100% and below 0%.

Now, the model gets better as the year goes on, so the numbers will change as we and everyone else plays more games. But, based on what's been played so far, Michigan and Stanford will be our toughest games. USC and (gasp) Wake are our hardest games left outside of Stanford.

For fun, here is what the model would predict for past years:

2008 (predicted 7-6)

San Diego St. win 94%
Michigan win 92%
Michigan St. loss 32%
Purdue win 77%
Stanford win 75%
North Carolina loss 35%
Washington win 94%
Pittsburgh loss 46%
Boston College loss 33%
Navy loss 42%
Syracuse win 96%
Southern California loss 3%
Hawaii win 65%


2009 (predicted 3-9)

Nevada loss 46%
Michigan win 56%
Michigan St. loss 49%
Purdue loss 42%
Washington win 70%
Southern California loss 33%
Boston College loss 50%
Washington St. win 98%
Navy loss 33%
Pittsburgh loss 11%
Connecticut loss 50%
Stanford loss 15%


2010 (predicted 8-5)

Purdue win 87%
Michigan win 66%
Michigan St. loss 28%
Stanford loss 21%
Boston College win 52%
Pittsburgh win 56%
Western Mich. win 83%
Navy loss 27%
Tulsa win 50%
Utah loss 46%
Army win 64%
Southern California loss 44%
Miami (FL) win 58%

Defense vs AFA - drive compression

A term I've made up for when playing an option offense is "drive compression" - or that what you look for your team to do is to get better over the course of the game... and drives become "compressed" throughout the game, or become worse and worse.

If I take a 2-drive, rolling average of AFA drives against us in terms of total yards and yards per play, we get:
[total yards] [yards per play]
92 10.2
144 6.5
106 4.8
106 5.3
106 6.6
29 4.1
21 2.1
72 2.9
60 2.9
84 12.0
145 13.2

You can see the first two drives, they really ripped into us - over 10 yards/play. Then, they continued to gain yards (over 100 yds per two drives), but the yds per play dropped significantly.

Then, our defense really shut their offense down, that is - until our subs came in and they started gaining a lot of yards again.

This could be due to them being worn down by our size or us getting used to their execution speed. But it is good to see the team get better as the game went on. Hopefully, the team is able to carry over this performance into Navy so they don't get off to as strong of a start.

Are we getting better? A progress report

Some fun with statistics -

This year, 4 of our 5 games have been against familiar opponents (UM, MSU, Pitt, PU). Looking at stats from 2008-2011, this has been our best year by a significant margin. Looking at 4 measures:
1) Record
2) Points differential
3) Yards differential
4) Yards per play differential

1) Record:
This is our best year by one game. In the last three years, we’ve split games with these four teams, going 2-2 each year. Always beating PU, going 1-2 with the other three. This year, we are 3-1.

2) Point differential
This is our best year by a significant amount – we are +11/game. 2008 was +4, 2009 was -1, 2010 was +2.5.

3) Yards differential
This is our best year by a significant amount. +96 yards differential game. About 100 yds/game better than last year – with 10-15 yds due to offense, and 90 due to defense. 2010 and 2009 were pretty much the same (-5/6 yds/game). 2008 was the worst, -40/game. An interesting note - last year, only 15 teams averaged +96 yards/game differential or better.

4) Yards per play differential
Also our best year (and are running the ball more!). We are at +0.6. Up from +0.3 last year and -0.8 in ’09 and -0.3 in ’08.

So, all in all, by the numbers, we are looking much improved.