I thought we were all just riding the recruiting high but 9 to 10 wins seems more realistic when you look at that chart. Go Hoke and Go Blue.
Wins of First Year Coaches 2001-2010
[Ed-M: Bumped to Diaries]
(Click for larger)
I looked at the coaching records of new hires since 2001. There's a pretty wide spread. I looked at returing starters vs. this data but the data I have is incomplete. Hopefully this comes out. My wife is about to pull the plug on me.
I don't have time to play with the sizing here. This has some value maybe. If I come back to this I'll add more detail.
The negative 10 outlier is Stan Parrish following Brady Hoke at Ball State...
[Ed-M again: Adding some thoughts so I can get this to 200 words and bump to Diaries.
The bell curve here suggests the data are pretty solid, but a Delta of +0.5 or 1 wins for a new coach seems really really good. Reason being, look at the third chart, with the coaches who took over 12-win teams. We're not talking about replacing just fired guys, but also those who moved up or moved on after dream seasons. Michigan is in that 7-win area where only two teams improved.
So is there an advantage gained from having a 1st-year coach? This doesn't say, but it's not a massive disadvantage, which is what I kind of expected would happen.
I think this needs to show the difference between coaches replacing a fired guy (which suggests there was a level of suspected incompetence) versus a coach brought in to replace a legend or retiree, who in turn was followed out the door by most of his starters. In other words, it's gotta pull out the noise. There's plenty of that around, since we can expect the circumstances of the team that lost 10 more games from the year before had a lot more going into them than Stan Parrish (meh) taking over for Brady Hoke (RAWR!)].
Maybe I'm reading this wrong, but it looks to me like the chart indicates a 62% shot at 14-0.
The chart represents that only 2 coaches who inherited a team with 7 wins the previous year won more than 7 games the next year, so I'm unsure how you draw that conclusion.
However, I don't know that we can really draw any correlation between these numbers because of how many factors there are, what was the recuiting class like, how many players graduated, who was the coach that left, are there sanctions involved, Lane Kiffin is two of those dots, etc, etc.
EDIT: looking at it further, 13 coaches with 7 or fewer victories won 8 or more games in their first year, that's 6.5%. I think you can also see that the bulk of the positive deltas occur at or below 6 wins (which makes sense to me), but I think that colors the delta chart in a rosy light.
One thing I just noticed is that the trend goes up until year 6 and then drops a tad before it goes back up again at year 9. This is interesting if you think about the real first class a coach gets to recruit from start to finish is about 6 years out. By the time they are seniors one would assume they have bought in to the philosphy and schemes from the regime.
Then the little drop off (Years 7 & 8) are the kids/classes that need to step up and fill those shoes, as they mature they improve even more than their predecessors. Pretty cool, I like stats when one can apply some good old common sense to back it up.
um... where is this discussion coming from? unless i'm missing something, the above data only applies to the first year after a coach is hired. the numbers on the x-axis of the bottom chart are wins from the previous year, not years coached.
LOL Derp, I think I should not post before coffee.
ha, that was an impressive amount of analysis pre-coffee.
dude, that made my day. I am still laughing at how seriously you were analysing that data in entirely the wrong way. I agree with andrew though, that was some impressive analysis for having not had a coffee yet.
Dude...I'm still laughing (though I hope I'm laughing with you at this point and not at you).
finish this season 11-3 (Conference Championship lost and Bowl win included) thanks to hoke "walking" to Ann Arbor, Greg Mattison hiring, and D-Rob on a mission to prove he's a true dual-threat!. Bleed Maize and Blue
Don't take THE KNOWLEDGE's job
What I find most interesting about the data is that the charts seem to show that first-year coaches tend to do as well as or better than their predecessors, and not only when taking over teams with very low win totals, but also when taking over teams in the 7-11 win range. Only the new coaches taking over 12-win teams appear to have consistently done worse, and it's pretty uncommon to post back-to-back 12-win seasons under any circumstances.
That's presumably because coaches tend to get fired for underperforming relative to talent and circumstance.
The slope of the least squares line is about 1. I don't think you can say that "first-year coaches tend to do as well as or better than their predecessors". Their success rate seems to be neither better nor worse. This is also pretty clear from the top graph which looks like a nice spread-out Bell centered around 0.
The observation about taking over 12-win teams is accurate, but there's obviously not much place else to go but down when your team wins 12 games in a season...
What this shows is that new coaches, in their first year, are (on average) no better than their predecessors. This suggests an 8–4 regular season would be pretty good for Michigan, and 9–3 would be outstanding.
I am not suggesting that the fans would be happy with 8–4, only pointing out what is realistic. At 9–3, Hoke would be in the 75th percentile for a first-year coach. At 8–4, he would be above average.
It would be interesting to see the data for Years 2 and 3. Most new head coaches replace someone who failed, so you expect to see improvement. However, the data suggests that it’s hard to make great progress immediately, because you still have mostly the same players, and it takes a while to install a new system.
there are issues with some of the data (the +9 data point is not real - probably a few more of these with interim coaches) but I don't think it changes the distribution too much. I'm not going to be able to revisit this without marital strife.
On par new coaches don't do much their first year.
Exceptions (>2 delta wins) for 6+ win programs are Charlie Weis-Notre Dame-2005 COACH, Jimbo Fisher-Florida State-2010 COACH, Chris Petersen-Boise State-2006 COACH, Gary Crowton-BYU-2001 COACH.
Petersen is a stud but is not instructive - different team (9 wins)/league. Weis is ... Weis. Fisher was coach in waiting (same playbook-same terminology-less Bowden). Crowton looks like the best - "be like this guy" for Hoke.
Last year, we would have guaranteed improvement.