SIAP: Steve Lorenz calls b/s on offensive line graph

Submitted by EastCoast on September 21st, 2017 at 11:51 AM

I feel somewhat obligated to post this because I posted the offensive line graph that purported to show Owenu and Ulizio being god awful:

 

According to Lorenz, the chart is b/s. Given Steve's reputation, I think I was too quick to post. The article is paywalled, but here's the link: http://michigan.247sports.com/Board/59424/Contents/Quick-VIP-Note-OL-graphicstuff-107670674

Comments

ScooterTooter

September 21st, 2017 at 12:07 PM ^

Yeah, like I posted in the original, the site this comes from is trash. They gave no explanation as to how they came up with those numbers and in fact this wasn't even able to be found on their site. 

Farnn

September 21st, 2017 at 12:57 PM ^

I think there's a really interesting discussion around this idea.  Without knowing the exact play call, how accurate can these game charting systems be?  And coaches will never release the actual play call.  So coaches can call these out and say their internal charting shows very different results, and fans can't really object because they don't know the play call. However, I'm sure coaches will lie to cover their players or create more confusion among opponents.  So the best we get is flawed data.

CLion

September 21st, 2017 at 2:28 PM ^

Right, even if it were Jon Gruden was breaking down the film for these, he still can't know exactly what assignments were, not to mention that there's goiing to always be some subjectivity plus inherent issues in breaking plays down into +/- points on a player by player basis.

What Brian has going for him is he knows the team well and consistency due to him being the only one doing it. With a site like PFF, even if they know the game better, there task is much more difficult.

OwenGoBlue

September 21st, 2017 at 1:18 PM ^

I think the hubbub is mostly because PFF has blown up and the subjective grades are being elevated/presented as gospel by ESPN, game announcers, writers, etc. 

PFF is useful (more information!) but a bit scattershot, particularly with line play. Like any attempts to quantify football, if you don't know the assignments and how things are coached, you're always going to miss on a fair amount of plusses and minuses. 

It's not just sour grapes: I know TJ Lang has been vocal about PFF being wrong and he's a PFF fave rave. 

taistreetsmyhero

September 21st, 2017 at 12:35 PM ^

and worst in all of college football.

Would be curious to see where Ulizio and Onwenu actually fall. Obviously, it's somewhat pointless to compare across all teams this early in the season, as differences in opponent quality are so vast. But it should still be relatively easy to do this analysis. Wonder where these guys went wrong...

Btown Wolverine

September 21st, 2017 at 12:52 PM ^

Below average on this graph is <50.

Onwenu and Ulizio's scores on the graph indicate that Onwenu is in the bottom ~15ish in all of FBS while Ulizio is in the bottom ~3ish in all of FBS (I didn't look up the exact number of teams, I think it's around 130 though, so 260 starters at Guard and Tackle each). That's a LOT worse than just below average.

Farnn

September 21st, 2017 at 1:04 PM ^

Methodology is likely poor and Michigan's internal charting shows much different results.  These sites don't know the playcall, and I doubt they spend enough time on each game if they are actually charting every player on every play.  Look at how long Brian takes to do UFR and that is just for 1 game a week, plus even he admits that his charting isn't perfect due to info he doesn't have from the coaches.  So you need to take these numbers with a big grain of salt and if something doesn't pass the smell test, it's likely an error.

gobluedave

September 21st, 2017 at 1:31 PM ^

I think the 247 crew should stick with recruiting. sure the graph isn't 100% accurate but it's reflective. would love to see it for the entire team, esp. our dl and bush. not sure I'd want to see sp8.