Michigan’s Adjusted Line Yards - 2012
I read Seth’s post today and spent this morning running through the every snap video uploaded by DGDestroys. The RPS nature of that one play made me question my harsh take on the OL performance in that game.
This play is well blocked… it just doesn’t comprehend the safety.
Good point Seth. Good job OL. Maybe there are more RPS losses than OL misses in this game (I thought to myself…)
So I downloaded the Game stats for the year; adjusted the rushing output for line yards following for the most part Football Outsiders technique to decipher how much the OL contributed to the rushing stats so far. Hit the link above for the run down on Adjusted Line Yards but here is the short story.
…the Adjusted Line Yards formula takes all running back carries and assigns responsibility to the offensive line based on the following percentages:
Losses: 120% value
0-4 Yards: 100% value
5-10 Yards: 50% value
11+ Yards: 0% value
I started to normalize and pull the RB yards – but it’s not what this morning is going to be for me. Instead I thought I’d post this the rough run down because I need to go and get MSS chores done.
The concept here is that yards over 10 are due to shake and bake and WR blocking while <10 yards are more or less on the OL. A good RB skews this up… a bad one down but the idea is in general a good one.
Stats are from http://www.cfbstats.com/. It looks like sacks are counted as rushing attempts. I didn’t doctor this data – it’s straight from their rushing chart. I could tinker with this but I’m not sure how much value it would add. The comparison is straight up.
Ohio is a significant shift in OL performance in terms of Line Yards.
Fuller
It’s difficult to pull out OL stats. The play that Seth pulled this morning is comprehended in the Stat pull and analysis I did just now as an OL miss – but it wasn’t as I take his analysis. In general, I think the Adjusted Line Yard stat is a good one though to begin to tease this out.
I have intentions of looking at this further (Power Success/RB Yard/2nd Level Yards/Sacks and Sack Rates) with comparisons by and to conference and by coaches… but I’m not sure I will get back to this. In the meantime I think it’s worthy to share this season run down.
Photos from MGoBlog – attributed as posted
November 27th, 2012 at 6:48 PM ^
As you admit, Adjusted Line Yards is a somewhat noisy and imperfect measure of offensive line run-blocking - but it's still informative, and I think this is a really useful exercise. Thanks.
Any chance you could do a time-series comparison at some point (i.e. the data for Michigan over a number of years)? I have to assume the play-by-play data is available going back for at least a few years. I think an exercise like this would help put this year's results in better context, and I suspect it would help validate the measure to some skeptical readers as well (i.e. that the measure has some relation to their subjective assessment of our run-blocking ability each year).
In the long run it would be neat if this could be expanded further - for example, with similar data for our competitors, notably Wisconsin - but this is a good start.
November 27th, 2012 at 9:42 PM ^
once they lost huyge and molk their line couldnt run block at all this whole year. hope next season kalis, bryant,miller can open up blocks cuz hoke is recruiting the best lineman out there
November 27th, 2012 at 10:00 PM ^
but why don't we account for passing yards in this?
After all pocket protection is also on the OL, no?
November 28th, 2012 at 8:51 AM ^
I will do this if I can.
It just takes time...which I don't have. I need to normalize this and account for success wrt down, distance and field position as well.
I included QB runs and disregarded formation which FO takes into account because they care about NFL stuff (boring...). Sacks could/should have been excluded but ... again time is the bite.
Earth calling Mathlete... help.
Maybe this weekend if it doesn't snow enough for the slopes.
Comments