needs moar usage
Class size vs star rating
Filed under:
There is always a debate about the significance of recruiting rankings when teams don't have the same number of recruits. Here's the question for this year.
Is it fair to rank UM's class higher than OSU's, as a number of the recruiting services have done, with rivals not even putting OSU in the top 25?
Some argue that OSU’s class is only ranked lower because they had a smaller class size. But how does class size affect the impact of the recruiting success of the entire class.
We can get a very crude idea of the impact by asking a purely hypothetical, simplified question. Suppose you have a class the size of UM’s and each recruit has an independent, 50-50 chance of succeeding. Then the likelihood that more than half of the recruits succeed is itself, coincidentally, 50% (see link). By contrast, the chance of success of the same number of recruits in the smaller class for OSU is only 2 %.
But what if we raise the chance of success for OSU's recruits (with an average rank about 6% higher in scout or rivals, as I recall)? Let's be generous and say that each of OSU’s recruits has a 10% greater chance of succeeding than UM’s. Then the chance that their class succeeds as a whole (more than half contribute) is still only 9%, less than a fifth of UM’s.
This oversimplified analysis admittedly ignores possible dependencies, heterogeneities, different degrees of contribution, and the fact that recruiting more players today may lead to more spaces next year, or vice versa. However, it’s uncertain how important the latter is.
UM’s large class this year may also be due to the smaller size of prior years (after attrition) as compared with OSU’s. So, prospectively, it is possible that we will be able to recruit as many as OSU in the coming few years—somebody might check this out. So, possibly, this year, we just made up a lot of ground and evened the playing field in numbers, while also making this year's class a lot more likely to succeed than OSU's.
http://www.stat.tamu.edu/~west/applets/binomialdemo.html
February 3rd, 2010 at 10:31 PM | 
Where is the 2% coming from?
Where is the 2% coming from?
February 3rd, 2010 at 10:54 PM |
use the binomial calculator (link)
use the binomial calculator (link)
to compute the chance of "success" of more than 13 out of 18 players for OSU (assuming the chance is 50%). Then I also caluclated it for OUS assuming the chance is 60%. I calculated for UM the chance of more than 13 out of 27 successes assuming the chance of success was 50%.
You can put in other numbers too if you want to see what happens with different assumptons.
michelin
February 3rd, 2010 at 11:27 PM |
Gotcha... Makes sense now
Gotcha... Makes sense now
Gotcha... Makes sense now
February 4th, 2010 at 8:49 AM |
Note
Note
Link is missing...
Since the scholarship limit is 85, that means a school should average around 21 per year, with more or less slippage and/or wobble each year depending on 5th-years, walk-ons, transfers (in or out), and off-field problems. U-M has had a lot of slippage in recent years, so this is a big class. As such, it is an important class. Rich's future at Michigan is riding on this class, possibly even more so than last year's. These guys will be the juniors and red-shirt sophomores in Rich's fifth season, the heart of the team's depth. The more contributors he has among this group, the stronger that (2012) team will be. And it needs to be dominant -- there needs to be a sense that the future is bright if Rich is going to get a new contract that Fall, and it's this class that will provide that sense.
February 4th, 2010 at 11:08 AM |
I agree
I agree
and the large class size should help him do what he needs, given the inevitable uncertainty about how many highly ranked recruits don't pan out and how many lowly ranked ones do. The large class size removes some of the uncertainty about having enought key contributors.
Sorry about the missing link...I thought I had included it in one of my edits but my IE has been hanging up on this site when when I try to add material. Now, it's posted and here I'll post it again. Thanks for pointing out the omission.
http://www.stat.tamu.edu/~west/applets/binomialdemo.html
michelin
February 3rd, 2010 at 10:32 PM |
Rating classes
Rating classes
The classes aren't really comparable in the traditional sense. OSU had fewer holes to fill than we did, and were thus able to make do with a smaller class size while focusing on quality over quantity. Michigan had a ton of holes to fill, particularly on the defensive side of the ball, and thus really needed to improve depth.
Rivals, Scout, ESPN, etc. all rate recruiting classes in a formulaic matter that is designed to predict the value of a class according to your criteria, that is, the level of contribution in quality and quantity expected out of that class. They do this not because it is always a relevant standard for evaluating classes, but because it is the best standard to compare every class in the country in order to make a ranking system and their consumers (namely us) like to see these rankings. That said, when you measure the value of a class in depth, it is generally a better idea to look at whether a class fills the specific needs of the team in question.
By this standard, Michigan and OSU performed equally well. Michigan addressed a lot of issues, shoring up the defensive line (minus a nose tackle), linebackers, secondary, and QB depth. However, we only picked up one lineman and we didn't get a true nose tackle, which leaves issues for next year's class to solve. OSU performed similarly, shoring up the defensive side of the ball and the QB position, while leaving a problem at DT and the offensive line depth. That said, OSU had fewer depth issues than Michigan and thus didn't have to sign as many players this year.
One thing to note is that because OSU didn't have the depth issues that Michigan does, they were free (well...freer than Michigan anyway) to go after elite prospects because they didn't need to be sure that they were picking up players to fill a bunch of holes. For the most part, they failed to get any, so by that standard their class is a disappointment, while Michigan's really isn't. However, all that really means is that OSU had more potential for this class than Michigan did but failed to fulfill that potential, which isn't really relevant when comparing the value of the classes.
February 3rd, 2010 at 10:50 PM |
my analysis is just for illustrative purposes
my analysis is just for illustrative purposes
I did not discuss the "balance" aspect. The whole point was to take a look at the relative effects of size and rating. I actually think that you would need a pretty complicated "human resources" model to get at the issues you raise. I doubt that Scout or any other services really provide the level of detail that would be needed for that.
michelin
February 4th, 2010 at 2:15 AM |
very true
very true
+1 to seth. as i have been saying to my friends. i'd love to see an after signing day grading scale kinda like what the NFL has done. they evaluate the draft based on need and pick-ups. we needed a lot and got a lot.while i think ours was a very good class based on need and lack of depth. i will take exception to the fact that osu had a good class also. let's look at a few facts, osu has beaten us for quite a few years, been co or outright big 10 champs the last 4 or so yrs, just won the rose bowl, and this was the best they could do? we on the other hand have gone 3-9 and 5-7 and have pulled this kind of class, needs or no needs. if we can manage a 8-4 season with a good bowl game(win or lose) the recruiting of RR will soon put the big 10 championship in our hands more years than not.
dundee
February 3rd, 2010 at 10:35 PM |
Double Post
Double Post
Please delete.
February 4th, 2010 at 12:48 PM |
Couple things
Couple things
I don't even worry about where UM's recruiting class is ranked. I look more at what the average star/player is of the class and where the players where ranked for their position group.
The one question I have is what is your definition of a recruit 'succeeding'? Do they have to start at least one year to be considered a success? Do they need to be named to an All-Big Ten team to be considered a success?
There's also various levels of success. Hypothetically speaking now - It could happen that maybe half of UM's class ends up being starters when they are seniors but none are really standouts. For OSU, when they are seniors only 4 of this years class ends up being starters. However, 2 are 1st team All-Americans and the other two get 1st team All Big Ten.
Really, the final analysis of how well a recruiting class is/was can't be determined until 4-5 years from now when we see how they all panned out.
February 4th, 2010 at 2:51 PM |
the dynamics of recruiting
the dynamics of recruiting
1. It does not matter what definition of success we agree on. The post shows the effect of numbers (vis a vis star ratings) regardless of the criterion we choose to define success.
2. I already acknowledged that the analysis does not consider varying degrees of success. But, I believe that the same general conclusions will apply to multiple success categories.
3. WHEN the players succeed in making a contribution does complicate matters. But undoubtedly, the most important question is how class characteristics affect the flow of high quality recruits into positions of need, depending on the chances of needs arising.
Thus, if you have a deficiency in a position, say DB, then recruit a lot of them, you may end up with a surplus of at that position in 4 years at the cost of creating deficiencies in other positions of need eg LBs. Similarly, some businesses order a lot of products when there is a demand but then by the time the products are produced and ready to be sold, the demand has decreased and they go bankrupt.
In our haste to fix the team in the short term, we can create more long term problems.
michelin
February 4th, 2010 at 2:58 PM |
the role of the AD
the role of the AD
Perhaps the AD, or his representative, should play a role in recruiting, since there is a need for long-term planning at different positions, which may have an impact beyond the tenure of a coach (as when we knew LC would retire soon).
That essentially happens in the NFL already, since coaches may take a shorter term view that minimizes effort (when they know they'll retire) or which preserves their jobs (when they feel threatened).
In the past, we did not have an AD with enough expertise to get personally involved in recruiting decisions. But now maybe we do.
michelin
Where is the 2% coming from?