Crimson's picture


MEMBER SINCE   July 11, 2012


  • COLLEGE FOOTBALL PLAYER: Dane Sanzenbacher
  • NFL TEAM: Detroit Lions

Recent Activity

Comment 04 May 2019

but he'll have to sit out the 2019 season unless an unlikely waiver comes from the NCAA.

He boasted an impressive offer list in high school that included Louisville, LSU, Nebraska, Virginia Tech and Wisconsin.

 Agree to disagree.

Comment 10 Apr 2019

As long as we're being snarky, here's my take.

How dare the coaches script things to see who can do what, as if it's a real game, expecting people to want to come see players who are new or improved!!!  Why should the coaches try to improve the team as they see fit before the season!!!  It's insanity!!!  In my day, we practiced mental toughness by playing Russian Roulette before a game!!!!!

Comment 02 Apr 2019

You're getting a bit argumentative, but whatever . . .

No one here has data to say whether something is a trend or an outlier.  If context matters, how does it help your prediction?  You don't know, and I don't either.  If you don't like the answer, then go collect the data (yeah, I already said that).

None of us have much data.  No need to get on a high horse about it.

It's not a hot take.  That's just your inference.  We missed on Donovan Peoples-Jones.  We took Michael Jordan (2016), Mike Weber, Joshua Alabi (2015), and Damon Webb (2014).  We missed on Malik McDowell too.  But the Coombs effect was real.

Comment 02 Apr 2019

I've told you how I aggregate what I see, in my head.  It's probably not too far off from the "it's accurate at signing day" comment, because you never know when someone will switch commitments for one reason or another (I think they only define accuracy as predictions for the final commitment) or change where they think they want to go if uncommitted.

If someone wanted to do the research, 247 probably has better data available to academics (maybe anyone).  Twitter keeps their full history, and there are ways to download parts of it.  All big companies bring in academics to help on certain parts that will also get publications.  I just don't know who has the time and resources to do it.

Without such analysis, my answers are: I think CB predictions should be viewed cautiously.  Many things change: did I get an offer I didn't have before, will OSU now accept my commitment, did I catch a coach lying, etc.  The final predictions are pretty accurate, but my guess is that early ones are far less so (pending good analysis).
I view NHB's comment as a valid counterpoint.  In the last few years, if we really want a Michigan player, we can take them most of the time.  I would guess that if any of these guys we have early interest in are actually at the top of our board later on, the CBs will shift, and we have a great shot to get them.  From what I've read, I would point more toward Cameron Martinez than McGregor & Edwards.

Comment 01 Apr 2019

CBs are a dance.  Harrison and Friday were widely predicted to Michigan.  Harrison started as OSU, then moved far way, then moved back in the days before he signed.  Friday was a last minute interest of OSU and trended accordingly, if I recall.

The problem with these accuracy metrics is that they're ONLY locked in when the recruit commits (and only to the school they commit to).  News seems to always leak out where players are going within a few days before, everyone switches their crystal ball, and then they commit and predictions are locked in.  This is why crystal balls aren't reliable: not because they aren't, but because they are when it doesn't matter (two days before they commit and after).

If you'd like to evaluate this on a large sample, feel free to pull all crystal ball predictions (not just final ones) and see what the world looks like.