Re-rating recruits based on performance
I've been toying with my spreadsheet with recent recruiting classes and tried to objectively re-rate each recruit signed since 2010. I want to stick to the Scout system of 1*-5*. The idea is to see how many lived up to projected potential, evaluate recruiting success or lack thereof, and ultimately to bash on Sarkisian because that's just fun. Here is the draft rating "system" I came up with:
1*: never made it or flamed out before doing anything
2*: stayed but just a bad player unable to get on the field, or stayed a while and played poorly, then left.
3*: stayed and played but not a good player, played some at a functional level then left, or still here and undetermined
4*: stayed/staying, solid or good player or important contributor whether they should be or not (Mickens for example)
5*: All Conference caliber
There are some weaknesses in my system. It's hard to account for someone like Josh Shirley, who actually gave a couple of decent years but got the boot with eligibility still on the table. He gets 3* on my scale. It's hard to fairly rate young guys, like the bulk of the 2013 class, but that mostly results in the average being dragged toward the middle (3*) since nobody that young is likely to be All Conference at this stage and few are even going to be important contributors yet. I also have a problem with a guy like Mickens, who I believe to be just barely above terrible, being only one point below a great player like Shelton or Kikaha simply because he's been targeted a lot for strange reasons (Sark and his bubble screen obsession and Smith because he has very limited options).
So...the point of this post is to solicit input/suggestions. How would you do it?
Comments
-
Funny that you mention this... @CokeGreaterThanPepsi and I have been debating this over the last year or so. My thought is to rate guys only after they have exhausted their eligibility. We've looked at it a few ways including something similar but everything is so arbitrary based on the person actually evaluating. We thought about creating some sort of equation (like in fantasy football) using stats but that is really hard to rate some defensive players.
I would love to hear solutions. -
A thought on that: Set up a Google Doc and post the link or invite posters here to rate players using those guidelines, then take the average.MrsPetersen said:Funny that you mention this... @CokeGreaterThanPepsi and I have been debating this over the last year or so. My thought is to rate guys only after they have exhausted their eligibility. We've looked at it a few ways including something similar but everything is so arbitrary based on the person actually evaluating. We thought about creating some sort of equation (like in fantasy football) using stats but that is really hard to rate some defensive players.
I would love to hear solutions. -
I don't think stats are the way. They're helpful in some cases, for instance the number of tackles for a linebacker, or QB sacks/pressures for a defensive end. We have to rely on our eyes for most though which means subjectivity but well defined ratings might take some of that out. We know who the all conference guys are (5*). The solid 4* guys are going to be obvious, but it gets real fuzzy on the borderline 3*/4* part of the scale. Some guys that many of us think are very good players but somehow don't get involved (Kendal Taylor is a good example) would be very open to debate. I think he rates a 4* but his (lack of) production (stats) says otherwise. The 1* and 2* categories are actually fairly cut and dry.MrsPetersen said:Funny that you mention this... @CokeGreaterThanPepsi and I have been debating this over the last year or so. My thought is to rate guys only after they have exhausted their eligibility. We've looked at it a few ways including something similar but everything is so arbitrary based on the person actually evaluating. We thought about creating some sort of equation (like in fantasy football) using stats but that is really hard to rate some defensive players.
I would love to hear solutions.
I do agree that rating most players who still have most or part of their career ahead of them is going to result in errors. I looked at my list for the 2013 class and most are stuck at 3* right now because you just don't know. The dropouts are obvious enough, and the ones playing meaningful snaps right now (like Qualls) are obvious, but the rest are unknowns.
Kind of fun anyways. -
Really like that idea grundle! We could find like ten posters on here to help us re-rank. That way we can avoid duck trolls trying to make UW recruiting look bad!
-
And @chuck, we should compare spreadsheets someday, might get new ideas. Although, @MrsPetersen is the one who built the spreadsheet we use.
-
Agree @chuck, using stats makes rating OL virtually impossible. We also thought about assigning point totals for things like All Conference, All American, Position award.... etc. But at the end of the day I think I agree that there isn't a way to take the "human" element out of the equation. So maybe polling 10 people or so to get a good cross-section and hopefully remove some of the bias.
-
Mine is pretty unpolished. I just copied and pasted straight from the commit list pages on Scout then fought through the formatting enough to be able to sort. I've thrown together some summary tables but most of them are no good anymore because of changes that have happened since and a lack of forethought on how to set it up.CokeGreaterThanPepsi said:And @chuck, we should compare spreadsheets someday, might get new ideas. Although, @MrsPetersen is the one who built the spreadsheet we use.
-
This article is way above my pay grade, but does have some ways to quantify offensive line performance.MrsPetersen said:Agree @chuck, using stats makes rating OL virtually impossible. We also thought about assigning point totals for things like All Conference, All American, Position award.... etc. But at the end of the day I think I agree that there isn't a way to take the "human" element out of the equation. So maybe polling 10 people or so to get a good cross-section and hopefully remove some of the bias.
The Toolbox: Offensive Line Stats - Football Study Hall -
We went through the Scout commit pages back to 2002 (when they started). We entered each player in with all of their details (stars, hometown, state, high school, position, commit date, whether they were a consistent starter or not, qualify/no qualify, nfl, nfl draft position). Then we added a "Contributor" field and that is the one we are stuck on. Our original thought was a 1-5 where 0 means they never stepped foot on the field, and 5 is basically a Bishop Sankey.
-
What about adjusting the scale to make 0 the flameout, never got into school grade?
