Meta: Diary Analysis

Submitted by Finance-PhD on

A great deal of discussion exists around the offense. I was thinking about a way of determining how good or bad production has been over the last three years. The areas I was thinking would be the best in the evaluation are rushing yards, passing yards, and points. Naturally things like conversion percentage are good and could also be included but so many things go into that (and we have had a number of diaries on those issues).

To normalize the data a bit I was thinking of looking at production as a percentage of the defense average. For instance, Michigan passes for 250 yards, rushes for 120 yards and scores 14 points against a defense that only gives up 200 passing yards, 150 rushing yards and 10 points on average. That week would then be 125%, 80%, 140%. We can then have three graphs that show week over week how we do and see if we are trending up or down over time.

My concern is that Michigan playing is a one off event and we are comparing that to a 12 game season average so that same defense could have high variance so that production is not really different statistically from 100% across the board.

There was a great deal of rework done on the tread about youth on the O Line so I am hoping to get some ideas on data analysis now so that the obvious issues could be addressed and no one (or perhaps very few) expects a rework later.

LSAClassOf2000

November 12th, 2013 at 4:01 PM ^

I suppose my question then is how far back do you want to go - I have all the normalized tables for rushing and passing for 2011 to now right now now, along with scores and scoring margin. I think that it would actually be interesting to do a scattergram of rushing and passing as axes and categorize the games that way. 

I was actually thinking about doing a diary on this very thing in the next week or so, but I can share anything that I have if you're looking to do something. 

Finance-PhD

November 12th, 2013 at 4:21 PM ^

I was thinking 2011 just to have a run of a few years with these coaches. I wouldn't want to pick up too much noise.

I really enjoy your diaries so I can wait until you post. I was just thinking that instead of everyone complaining about this or that, let's look at the actual numbers.

The scatterplot is an interesting idea. I would just also want to have a table or something then so we can see if there is a time series impact like Michigan always does poorly in the middle of the schedule or we have been on a slide all year long.

LSAClassOf2000

November 12th, 2013 at 7:16 PM ^

Here is the entire Borges era - rushing and passing. I denoted the splits in the seasons by placing the year on the first opponent in each case. The most interesting shift occurs in the 2012 Nebraska game, if you look at it this way (click on the image to go to Photobucket if you have issues reading this):

 photo PassRushTrends_zps9fc164a0.png

restive neb

November 12th, 2013 at 4:37 PM ^

I think you really should normalize for pace. If you don't want to do yards per play, you can also factor total yards by the percentage of the plays run divided by plays run in an average game.

Gandalf the Maize

November 12th, 2013 at 4:40 PM ^

You may want to include overall defensive efficiency rank (something like defensive FEI) in order to test whether we tend to over/under perform against good/bad defenses. For example, is putting up 751 yards against Indiana (overperforming against a bad defense) and putting up -48 rushing yards against MSU (underperforming against a good defense) a trend? Or is it anomalous? Could be interesting to look at over a multi-year span.

markusr2007

November 12th, 2013 at 5:16 PM ^

Would there be a source for data as to the number of starts an OL player has had as a determinant value of their so called "experience level"?

I'm not sure where to find such a a data source, but I'm thinking Michigan would predictably score out "experience rich" at tackle (Lewan, Schofield) in 2013, but quite "experience poor" from both guards down to center, and dare I say QB  in terms of starts compared to other BIG10 football teams.

Looking at Michigan's OL roster 2011-2012-2013, one could determine the following:

2011 (11-2) - 6 upperclassmen, 9 underclassmen

2 seniors (Molk, Huyge)

4 juniors (Omameh, Koury, E.Mealer, Barnum)

3 sophomores (Gunderson, Lewan, Schofield)

6 freshman (Bryant, J. Miller, Mateus, Glasgow, Burzynski, Dallas Williams)

2012 (8-5) - 6 upperclassmen, 11 underclassmen

3 seniors (Omameh, E. Mealer, Barnum)

3 juniors (Gunderson, Lewan, Schofield)

2 sophomore (Burzynski, Mateus)

9 freshman (Bars, Braden, Bryant, Gibbs, Glasgow, Kalis, Magnuson, J. Miller, Pliska)

2013 (6-3 ytd) - 5 upperclassmen, 14 underclassmen

3 seniors (Gunderson, Lewan, Schofield)

2 juniors (Burzynski, Mateus)

3 sophomores (Bryant, Glasgow, Miller)

11 freshman (Bars, Bosch, Braden, Dawson, Fox, Kalis, Kugler, Magnusson, Pliska, Samuelson, Tulley-Tillman)

Contrast this to Michigan 2003 Rose Bowl team: 11 upperclassmen, 7 underclassmen

7 seniors (Baas, Pape, Pearson, D. Solomon, C. Morgan, J. Gaston, A. Christopfel)

4 juniors (Lentz, Stenavich, Henige, Schoonover)

5 sophomores (Bihl, Kolodziej, R. Riley, Berishaj)

2 freshman (Jake Long, P. Sharrow)

In terms of OL experience, Michigan seems to be upside down. It's unlikely Michigan remains upside down much longer.