goemaw.com
TITLETOWN - A Decade Long Celebration Of The Greatest Achievement In College Athletics History => Kansas State Football => Topic started by: CHONGS on September 30, 2012, 05:11:10 PM
-
| Team | Effective Off | Effective Def | Power % |
| 1 | FloridaSt. | 0.75 | 0.22 | 1. |
| 2 | Alabama | 0.61 | 0.15 | 0.95 |
| 3 | Georgia | 0.7 | 0.31 | 0.9 |
| 4 | KansasSt. | 0.63 | 0.27 | 0.88 |
| 5 | Texas | 0.61 | 0.29 | 0.86 |
| 6 | SouthCarolina | 0.54 | 0.22 | 0.86 |
| 7 | Oregon | 0.59 | 0.28 | 0.84 |
| 8 | Florida | 0.47 | 0.17 | 0.84 |
| 9 | TCU | 0.47 | 0.19 | 0.83 |
| 10 | LSU | 0.56 | 0.29 | 0.82 |
| 11 | NotreDame | 0.41 | 0.16 | 0.81 |
| 12 | MississippiSt. | 0.52 | 0.29 | 0.8 |
| 13 | WestVirginia | 0.63 | 0.39 | 0.79 |
| 14 | TexasTech | 0.51 | 0.29 | 0.79 |
| 15 | TexasA&M | 0.49 | 0.18 | 0.78 |
| 16 | OhioSt. | 0.46 | 0.27 | 0.77 |
| 17 | LouisianaTech | 0.54 | 0.36 | 0.76 |
| 18 | Nebraska | 0.59 | 0.33 | 0.75 |
| 19 | ArizonaSt. | 0.51 | 0.25 | 0.75 |
| 20 | OklahomaSt. | 0.66 | 0.31 | 0.74 |
| 21 | Cincinnati | 0.4 | 0.26 | 0.73 |
| 22 | Stanford | 0.44 | 0.21 | 0.73 |
| 23 | Oklahoma | 0.49 | 0.27 | 0.72 |
| 24 | SouthernCalifornia | 0.47 | 0.26 | 0.71 |
| 25 | NorthCarolina | 0.55 | 0.24 | 0.71 |
I now have enough data to call the third column the power %.
70% of the Big12 in the top 25!
-
I Love Chingon and I love the ORCHID!
-
An outstanding ranking system by an outstanding poster.
-
THIS IS INCREDIBLE! I LOVE IT AND I LOVE CHINGON!!!
-
Meh, I think it has us a few spots low.
-
now it's starting to look like something
-
Nailed it.
-
Sorry if luked... but where can we see more details? As in what do the numbers mean?
-
Is there anything better than the ORCHID? If it was a food I would eat it for breakfast, lunch, and dinner. EVERYDAY!
-
Sorry if luked... but where can we see more details? As in what do the numbers mean?
:flush:
-
Meh, I think it has us a few spots low.
You better watch your back 8mp. The ORCHID is the best and we will run you out of town if you think otherwise.
-
looks like it took this dipshit a week or two to get his stats straight but better late than never i guess.
-
Sorry if luked... but where can we see more details? As in what do the numbers mean?
If you have to ask you'll never know.
-
FYI who those who don't know:
Effective Off is effective offense, the average number of points that team scores per play (adjusted per game for opponents effective defense).
Effective Def is effective defense, the average number of points that team allows per play (adjusted per game for opponents effective offense).
The power % is the difference between them minus a penalty for # of losses.
It's simple and objective.
-
If you want to know who has the best defense, the number to look at is not yard per game or yards per play or points given up, but points per play given up (adjusted to strength of schedule) which is effective defense.
Alabama is #1 in that category and Florida St is #1 in effective offense.
KSU is #6 in effective offense and #23 in defense at this point in time.
-
Thanks Chingon.
As a math-y person I'm pretty interested in this kind of stuff. Seems simple enough. Would be interested in how you factor in the adjustments, but I won't ask you to just give away your entire formula for rankings. Good job. :thumbs:
-
:eek:
Looks like the Orchid is really hitting its stride!
-
Feels good to be undefeated vs ORCHID Top 25 teams.
-
Is it possible to account for SOS in these rankings? Maybe multiply power percentage by SOS or something. Or do you already account for it in your adjustments?
-
KSU is #6 in effective offense
This is crazy considering it seems like we run soooo many plays for 3.3 yards per play, so one would think it takes us more plays to score points. But the stats don't lie. Great work Chingon.
-
Is it possible to account for SOS in these rankings? Maybe multiply power percentage by SOS or something. Or do you already account for it in your adjustments?
already accounted for. The stats from every game this season so far are adjusted each week. So even though KSU was idle our numbers changed.
-
Is it possible to account for SOS in these rankings? Maybe multiply power percentage by SOS or something. Or do you already account for it in your adjustments?
already accounted for. The stats from every game this season so far are adjusted each week. So even though KSU was idle our numbers changed.
:thumbsup:
-
how does D/ST scoring affect the numbers, if at all?
-
how does D/ST scoring affect the numbers, if at all?
effective offense is number of points scored per play (regardless of how)
-
how does D/ST scoring affect the numbers, if at all?
effective offense is number of points scored per play (regardless of how)
:surprised: that's why KSU is #6!
-
how does D/ST scoring affect the numbers, if at all?
effective offense is number of points scored per play (regardless of how)
:surprised: that's why KSU is #6!
bingo!
:fingertonose:
and because we rarely go backwards. we chug forward at a good pace albeit we let the clock run. We have fewer offensive plays than a lot, but we make them count.
-
WHAT NOW, BITCHES! :emawkid:
-
Chings, after the season, you need to go back through your rankings and figure out how accurately they predict winners.
-
Why isn't effective offense based on points per possession instead of points per play? Haven't looked at real data for correlation but it seems to me like it would be less easily skewed.
4 drives from offense A:
- 3 and out
- 2 plays, touchdown
- 1 play, touchdown
- 4 plays, fumble
4 from offense B:
- 10 plays, TD
- 9 plays, TD
- 11 plays, FG
- 10 plays, TD
Team A pts/play: 1.4
Team B pts/play: .6
Team A pts/poss: 3.5
Team B pts/poss: 6
Just seems like pts/play favors quick strike offenses too much. Maybe there is a happy medium? Maybe large sample sizes make the difference negligible? I dunno, just my thoughts.
-
Those were good thoughts
-
I also like points/possession more than point/play. At the end of the day, it doesn't matter if you use a ton of plays or very few, if you find the endzone at the end of the possession, it was a success. If you keep the opposing offense out of the endzone, it was a success.
-
Why isn't effective offense based on points per possession instead of points per play? Haven't looked at real data for correlation but it seems to me like it would be less easily skewed.
4 drives from offense A:
- 3 and out
- 2 plays, touchdown
- 1 play, touchdown
- 4 plays, fumble
4 from offense B:
- 10 plays, TD
- 9 plays, TD
- 11 plays, FG
- 10 plays, TD
Team A pts/play: 1.4
Team B pts/play: .6
Team A pts/poss: 3.5
Team B pts/poss: 6
Just seems like pts/play favors quick strike offenses too much. Maybe there is a happy medium? Maybe large sample sizes make the difference negligible? I dunno, just my thoughts.
I agree in principle, but the number of possessions are not an easy stat to obtain.
I still think all things equal the team that scores in fewer possessions is more efficient.
I will post more on this, but for now here is the data for all games so far this season:
This is the distribution of raw points per play so far this year (the mean is 0.42 and the standard deviation of the data is 0.23)
[attachment deleted by admin]
-
Also the avg # of offensive plays per game so far is about 70.
-
Possible correlation?
(but does not mean causation people!)
[attachment deleted by admin]
-
I like points per play because I think field position and special teams have a greater impact on points per play than points per possession. I don't think the rankings would look a whole lot different either way, though.