Started By
Message
0
Posted on 2/13/13 at 11:20 am to engie
Elaborate on your numbers, methodology, etc. I'm genuinely interested at how you did this.
Posted on 2/13/13 at 11:40 am to DCRebel
I started with an individual game score. Say 41-24 in the Egg Bowl. I then compared it to what those teams score or give up on average to determine which unit was "better" in that game. In the Egg Bowl, you averaged scoring 28ppg in conference and we gave you 41, so your offense is 13 point above average in the game. While, you averaged giving up 29.9ppg in conference play and we scored 24, so your defense was 5.9ppg above average in the game. In this case, it's clear that the offense is what won you the game... And it was the "better unit" by a total of 7.1 points in this game.
The Theory is that you eliminate SOS as a factor(allowing you to include all D1A games although I'm sticking with in-conference only first, then going back and adding the rest to see how they skew the numbers). The idea is that you should hold (bad) teams to the same amount below their scoring average that you score above their average points allowed and viceversa. This ends up telling you quantitatively how good your units were against each other and against the mean...And once all the data is compiled, against everyone else WITHOUT SOS being such a deciding factor like it is in general football statistics...
Edit for dumbass math...
The Theory is that you eliminate SOS as a factor(allowing you to include all D1A games although I'm sticking with in-conference only first, then going back and adding the rest to see how they skew the numbers). The idea is that you should hold (bad) teams to the same amount below their scoring average that you score above their average points allowed and viceversa. This ends up telling you quantitatively how good your units were against each other and against the mean...And once all the data is compiled, against everyone else WITHOUT SOS being such a deciding factor like it is in general football statistics...
Edit for dumbass math...
This post was edited on 2/13/13 at 11:50 am
Posted on 2/13/13 at 11:47 am to engie
Would a number closer to 0 then mean that the game pretty much went as one would have expected it to based on other games that season?
Posted on 2/13/13 at 12:10 pm to DCRebel
quote:
Would a number closer to 0 then mean that the game pretty much went as one would have expected it to based on other games that season?
Yes. The closer it is to zero on each individual side(Odiff and Ddiff columns), the more "average" your units played should perhaps be the terminology. In the Superiority column, a zero simply means the units played equally well/bad...while a positive number means the defense played better than the offense, and a negative number means the inverse...
It's an interesting concept that I still don't fully have a grasp on to be honest with you. I just formulated it and started messing around with it last night, and it got me hooked because of all the things that these numbers could potentially tell us by simply taking things ONE step further(by the theory of eliminating SOS as a constraint, which plays such a blindingly huge role in college football statistically).
I haven't honestly crunched enough numbers yet to be able to say with 100% certainty that my model is mathematically viable... But it certainly passes the initial sniff test...
This post was edited on 2/13/13 at 12:14 pm
Posted on 2/13/13 at 1:06 pm to engie
Fascinating. Seriously
Posted on 2/13/13 at 1:16 pm to engie
I like that calculation. It's similar to one I've been thinking about, but don't have the capability of putting together. Did you put that thread on the SECR or another site?
Posted on 2/13/13 at 1:24 pm to ProjectP2294
quote:
I like that calculation. It's similar to one I've been thinking about, but don't have the capability of putting together. Did you put that thread on the SECR or another site?
It's simply in an argument on SPS at the moment...
Once I complete the calculations for everyone, I'll bring it to the Rant...
This post was edited on 2/13/13 at 1:25 pm
Posted on 2/13/13 at 1:30 pm to engie
Good deal, anxious to see. Also, I don't know what SPS is.
Posted on 2/13/13 at 1:58 pm to engie
But for real it was pretty good all the other SPS posters are horrible
Posted on 2/13/13 at 2:37 pm to ProjectP2294
quote:
Good deal, anxious to see. Also, I don't know what SPS is.
Just finished LSU's...
Offense was 11 points below average(not nearly as bad as I was expecting)...while defense was 63.8 points above average(we're talking over 8 games here). LSU's defense was the better unit by 9.35ppg. The only defensive "failure" you had where a team actually scored above their average was against Ole Miss -- which was 7 points above. Meanwhile, you had offensive failures against Auburn, Florida, and Arkansas...
This post was edited on 2/13/13 at 2:39 pm
Posted on 2/13/13 at 2:59 pm to engie
Could this be used to calculate the odds on betting? If so...I need your phone and email for next season.
Posted on 2/13/13 at 3:02 pm to engie
quote:
Meanwhile, you had offensive failures against Auburn, Florida
No shite.
Posted on 2/13/13 at 3:28 pm to engie
quote:
And I have no idea where #1 Rod Taylor goes. Seems to be leaning toward Bama right now, but who knows? Inner city Jackson really doesn't have a "lean" toward any school traditionally.
Rod told Simmons that he'd have a hard time leaving the state. OM lock
:froze:Posted on 2/13/13 at 3:48 pm to Whereisomaha
OLE MISS STACKED
Posted on 2/13/13 at 8:17 pm to Landsharks
Posted on 2/13/13 at 8:22 pm to engie
I can't wait....
Posted on 2/13/13 at 9:24 pm to engie
You think Dear will "transfer" to a big time HS team or will he stick it out in the JPS system???
Posted on 2/13/13 at 9:31 pm to engie
He's ours
Posted on 2/13/13 at 10:42 pm to Whereisomaha
quote:
He's ours
We'll see.
Page 5 of 8
Popular
Back to top
Follow SECRant for SEC Football News