Inbox! Questioning the Rocky Top Talk Jars System

by Joel Hollingsworth on Jul 2, 2007 7:06 AM EDT in Tennessee Volunteer Football

On Friday, I got an email from a reader in California who I've talked with a few times about the Getting to Know You Series. I thought that now might be a good time to post it, especially in light of the fact that this morning I named Gerald Jones, a four-star guy, No. 7 in the class over No. 8 Chris Donald, who had five star ratings from both Rivals and Scout. So here it is:

You know that I deeply enjoy your "Getting to know the Draft Class" Series. But today's feature on Chris Donald had me a bit confused. I know that you have your own ranking system, and you explain it from time to time with regards to a recruit's standing.

I was wondering why Chris Donald was at #8 on your list when he was, I believe, in the Rivals top 10. Shouldn't he have been higher in this class? I know that we got some whoppers this year, and I thought Donald was going to be probably in the top 3.

-- John

My response:

Thanks, John. Yeah, there's going to be a lot of room for argument when we're all done, which will be at least half of the fun. Everybody above Donald received offers from Florida, and three of them were offered by Southern Cal, too. Had Donald's top three offers been Florida, Notre Dame, and Miami, he would have been No. 4 instead of No. 8. Had his top three offers been Florida, Southern Cal, and Notre Dame, he would have been No. 1.

Is Donald underrated in the Rocky Top Talk Jars system? Probably. As we've learned, local kids who commit early to play for their in-state school generally are. Donald didn't commit early, but he was an in-state prospect, and that might have had something to do with not getting interest from schools like Florida and Southern Cal.

It's probably time I post the explanation of the system again, since I haven't done so since the first player way back on April 9:

A player is awarded Rocky Top Talk Jars based on four components: the view of the player by the major recruiting rankers (Scout, Rivals, and ESPN), all converted to a 100-point scale, and our own "offer rating." The offer rating awards players up to 100 points for up to three offers from other schools. For instance, if a player received an offer from Florida, which had the best recruiting class last year, he would receive 100 points. If he also had other offers from, say, Southern Cal and Texas, he would get another 96 points for the Southern Cal offer and another 88 points for the Texas offer. A player would get points for his top three offers only, and the schools ranked from 100 points for a Florida offer down to two points for a West Virginia offer. Yes, there are problems with this notion. For instance, some schools didn't offer a guy who committed early and was obviously not going to change his mind. Guys like that should be held in high esteem by all Vol fans, but loyalty is difficult to measure.
The data on the players was often different, depending on the source. In such a case, I either averaged them or took the less impressive number, depending on my mood at the time. Get over it.
I wanted to rank the guys on a ten-point scale instead of a five-point scale so that we could better differentiate between them all. Keep in mind, though, that Tennessee's class was ranked No. 3 or No. 4 in the nation this year, so even the guys that get two out of 10 Jars are quite good. All Tennessee recruits, by definition, are at the top of the bell curve.
You can reanimate the profile thingy by clicking on the stopwatch, the football player, or, for those that have weight-lifting information, the barbell.

So does the Rocky Top Talk Jars system overvalue offers and undervalue stars? I don't know. Shouldn't we put more stock in what a coach from a major college football program with a limited number of scholarships to offer thinks about a player than what a recruiting service with an unlimited supply of stars thinks of him? The stars are weighed, but the offers weigh more. But I could be wrong.

What do y'all think?

Stay tuned! Nos. 6-1 coming up.

Tweet Comment 1 comment | 0 recs |

Do you like this story?

Comments

I think

I think you ought to do away with the subjective human input and let computers determine the rankings. Do Huh? You mean that's been tried before and miserably failed so now we're back to relying on human input?! Sheesh!

The truth is that there is no perfect ranking system. Whether you're ranking player potential, teams in the BCS, or our Granny's Apple Pies, there will always be holes in the logic...human or artificial. Things that one individual may value most, another may not value at all. As long as all entities are evaluated to the same criteria/standards, not much can be done other than to forthcomingly (is that even a word) disclose known flaws in the system.

With that said, I really like the "Jars" system. However, I think you need points for either/both in-state and/or early commits to help offset the "offer rating". Perhaps 25 for in-state and 50 for an early commit?

by Aerobab on Jul 2, 2007 9:31 AM EDT reply actions