PDA

View Full Version : What we need is an algorithm


Romyrick
12-03-2008, 07:08 PM
Instead of scouting or watching the games, we should just create an algorithm that can measure speed per mass. It would be something that takes into account the mass of a player (height/weight) and speed (40/cone/?).

This is sort of a joke but theres a real point here. A lot of us look at the 40 times, cone drills, height and weight as one part of the scouting process. There will always be a debate about how much value should be given to this data. However, the amount of inference between us simply looking at the 40 times then the weight/height, cone, shuttle etc etc, I think could be cleared up by a simple algorithm. If we are going to look at measurables, we should clear up our interpretation of these measurables... Any math majors out there?

Race for the Heisman
12-03-2008, 07:13 PM
Instead of scouting or watching the games, we should just create an algorithm that can measure speed per mass. It would be something that takes into account the mass of a player (height/weight) and speed (40/cone/?).

This is sort of a joke but theres a real point here. A lot of us look at the 40 times, cone drills, height and weight as one part of the scouting process. There will always be a debate about how much value should be given to this data. However, the amount of inference between us simply looking at the 40 times then the weight/height, cone, shuttle etc etc, I think could be cleared up by a simple algorithm. If we are going to look at measurables, we should clear up our interpretation of these measurables... Any math majors out there?

SPARQ rating?

Don Vito
12-03-2008, 07:14 PM
Isn't that what SPARQ is for kids coming out of HS? It doesn't facor in size though.

Race for the Heisman
12-03-2008, 07:24 PM
Isn't that what SPARQ is for kids coming out of HS? It doesn't facor in size though.

Actually I think it does a bit (weight, anyway).

Paranoidmoonduck
12-03-2008, 07:27 PM
Realistically, it would probably be best to create a height vs. weight algorithm and then measure that against a 40 yard dash vs. short cone algorithm. Not that it would tell us a whole lot, to be honest. You can tell all you want about player right now, and any number is going to allow people to overhype its importance.

draftguru151
12-03-2008, 07:32 PM
There was a height/weight vs speed thing for RBs done last year, can't remember where it was though. Like anything like that it wasn't perfect but it was kind of telling (and a bit obvious as well).

Paranoidmoonduck
12-03-2008, 07:38 PM
There was a height/weight vs speed thing for RBs done last year, can't remember where it was though. Like anything like that it wasn't perfect but it was kind of telling (and a bit obvious as well).

I'm pretty sure that it was from Football Outsiders. I pulled my copy of the 2008 Prospectus off my shelf, but I can't find it in there. It might have been one of their smaller articles from the site.

I was sort of thinking something along those lines, but measuring the difference between a players 40-time and their short shuttle is usually a great barometer of general quickness (thought not always; as I recall Patrick Willis 40 and SS were almost the exact same number).

TitanHope
12-03-2008, 07:49 PM
I'm more of an english/history guy, so the math thing doesn't appeal to me as much aside from statistics. Though, whether you're math or english driven, a guy who runs a 4.55 40 at 275 lbs. is impressive. Plus, the human aspect of scouting and such is what gives draftniking life. Opinion is a powerful driving force, and without it things become boring. We're all annoyed by the posters who say, "I don't like him. He doesn't impress me when I watch him," and then leave it at that. But, I'll take that over, "Him having 23.45 lm^2/8.7 yards per second worries me," anyday.

I'd look more towards physics for possible equations, if I wanted to form one.

Hawk
12-03-2008, 07:51 PM
I don't see how it would help determine how good a player is.

Paranoidmoonduck
12-03-2008, 07:56 PM
Well, guys, it wouldn't replace film study. Nothing can. But it might stem the amount of importance of placed on individual workout numbers by combining them into a more telling number. Sure, it might not be wholly important to know that Gaines Adams putting up particular numbers at 265 lbs is less impressive than Williams putting up the same numbers at 295 lbs, but having something by which to cross reference that might ultimately useful.

TitanHope
12-03-2008, 08:00 PM
But dude...I so hate math...like, you don't even know.

If math were a puppy. Wait, not just a puppy. The cutest puppy ever. Like a labrador or golden retriever with a red bow around it's neck under the Christmas tree cute...

I'd punt it just on principle.

illmatic74
12-03-2008, 08:04 PM
Check out Football Outsider's speed score.

draftguru151
12-03-2008, 08:12 PM
We need to get T24L in here to figure out the W/Hp40ydTSs formula.

Bruce Banner
12-03-2008, 08:15 PM
We need to get T24L in here to figure out the W/Hp40ydTSs formula.

He's a copy pasta chef.

bored of education
12-03-2008, 08:16 PM
He's a copy pasta chef.


BAMMM KICK IT UP A VORP!

illmatic74
12-03-2008, 08:19 PM
Football Outsider's speed score- Speed Score – A method for projecting the NFL success of a rookie running back which combines a player's weight with his 40-yard dash time from the NFL scouting combine, thus accounting for the fact that a larger back is more likely to be a good NFL player than a small back with the same speed. Speed Scores generally range from 80 to 120, with 100 as the average.

giantsfan
12-03-2008, 09:33 PM
Football Outsider's speed score- Speed Score A method for projecting the NFL success of a rookie running back which combines a player's weight with his 40-yard dash time from the NFL scouting combine, thus accounting for the fact that a larger back is more likely to be a good NFL player than a small back with the same speed. Speed Scores generally range from 80 to 120, with 100 as the average.

That sounds like the Heather school of scouting. Straightline spedd + Size = BOOM! Awesome player... cone and short shuffle numbers are way more important to me than 40 time unless we're talking about corners where make up speed is somewhat important.

Mr. Stiller
12-03-2008, 11:16 PM
It would almost be a QB Rating type formula for Combine Drills vs. Size/speed.

do able, but you have to decide you're max mins and what the constants applied to each variable.

I guess it would be best to look at the athletic freaks because obviously a guy like Calvin Johnson should have scored a damn near perfect.

The only issue is, then you have to normalize it by position.