New YPPT Method

Quote Originally Posted by VGPOP:



Interesting approach you have here. I love systems, even the crazy ones.

The most important factor here is tha YPPT (Yards Per Points), and it appears you use Sagarin's strenght of schedule (SOS) to determing if to lower or increase the the YPPT.

What are you going to do when more games have been played?

Are you going to continue to use overall stats?

Teams get hot, teams get cold, and that needs to be considered.

Again, great stuff you have here. I may tweaked the system myself to see how it does.

I run a statistical model much like this for baseball.  It is dealing w run per out. 

The way I combat the above questions is to run the model for

1.) overall stats

2.) home/away stats - depending on if the team is home or away

3.) pitchers last 3 starts / hitting teams last 10 games

A line is creeated for all 3 using the same forula, but the stats oviously differ.  An average of all 3 is taken and that is my overall line.  The way in which i weight all 3 of the differnt lines is constantly changing in the way i feel it is most accurate.  It is not an exact science is weighting these.  I have an overall approach in how i like to weight them but like i said you can subjective change the weighting process how you feel has the most importance for that game. 

I need to make a correction from my earlier post. LA Monroe's final YPPT should be 15.9 not 15.2 ( I forgot to multiply by Ark.St.'s SOS multiplier 15.2 x 1.046 = 15.89). This fix changes the projected score total to.

Arkansas State            35.4  (35)

LA Monroe                    26.8  (27)

PLAY: Arkansas State + 2.5

GL

(Sorry I was'nt responding to posts this afternoon, I have Epilepsy & sometimes need to lay down for a few hours to avoid seizing.)

Here is all of the work for you people with excel. I will be trying  to learn how to use said program in the near future.

Ark.St.     SOS (68.32)  (1.046 Multiplier)

LA Monroe SOS (62.35)  (.954)                                                                   (68.32 + 62.35 = 130.67/2 = 65.34) (62.35/65.34 = .954)

Ark.St. (O)    162.7 (R)  (-22.4)    180.0 (P)  (-2.9)  11.6 YPPT   (-1.6)

LA Monroe (D) 110.2 (R) (+6.3)    268.2 (P)  (36.9) 12.2 YPPT   (-1.7)

Ark.St. (162.7 + 110.2 = 272.9/2 = 136.5)

            (-22.4 + 6.3 = -16.1/2 = -8.1)

            136.5 - 8.1 = 128.4 x 1.046 SOS = 134.3 (R)

           (180.0 + 268.2 = 448.2/2 = 224.1)

           (-2.9 + 36.9 = 34/2 = 17.0)

           224.1 + 17.0 = 241.1 x 1.046 = 252.2 (P)

           (11.6 + 12.1 = 23.7/2 = 11.9)

           ( 1.6 - 1.7 = -0.1/ 2 = -0.05  I left this alone)

            11.9 x .954 (LA Monroe's SOS) = 11.35 YPPT.

Ark.St.    134 (R)     252 (P)   386 TYDS    11.4 YPPT  ( 33.9)

REPEAT STEPS FOR LA MONROE & THERE YOU ARE.

GL

Runs per out?  Total runs/27?

skan 

With double positives & double negatives I split the difference meaning (-1.7 - -1.6 = -0.1).

You multiply offensive teams stats using offenses SOS. & multiply YPPT using defenses SOS.

Hope this clears things up a little.

GL