URK 2011: Week 13 Power Rankings
What a crazy regular season! We had two dominant teams, one for each half (the Z’s and the Records), one historically bad team (the first 1 win team, the Minions), the first six-win team to ever miss the playoffs (the SharkLions), and one heck of a wild playoff race.
And it’s not over yet! The postseason start tonight, and it’s as much of a tossup from 1-8 as I can ever remember. Probably won’t have any time to do much recap and analysis of these upcoming games, but I’ll see what I can do. If nothing else appears, look for the postseason awards ballot in about three weeks, and good luck snagging the ultimate crown!
All right, let’s get a bit nerdy for a second (or more than a second, if you want to get technical, which, come to think of it is even more nerdy. Whatever.). I’ve added one more category to the final stats: expected wins. Including it was inspired by some back and forth with Jaron, who is adamant (and incorrect, at least this year) that points against is a better predictor of how you do than points scored – in other words, that you have little control over your fantasy football fate. This is quite easily refuted by comparing charts (a) and (b), which show the linear regression of points scored and points against; pay attention to the r-squared value. R-squared indicates the predictive power of a certain variable; 1 is a perfect predictor, while 0 tells you nothing. Note that r-squared is significantly higher for points scored than points against. QED. However, points scored isn’t a perfect predictor…a slightly better one is something called expected wins, which is shown in (c), and takes into account both points scored and points against.
A couple of notes about the expected wins: I calculated them using Hollinger’s Pythagorean expected wins, which were developed for basketball. And, they’re not perfect – for one thing, unlike real teams, we have no defensive control over how well our opponents do, so points against doesn’t add a whole lot of extra value. Basically, all these stats can do is give you some idea of is how lucky or unlucky a team has been – below, positive numbers relative to actual wins mean the team has been lucky while negative means unlucky. Well, that turned out a bit longer than I intended. Let’s get to the Power Rankings charts, and the final Power Rankings standings.
12. Minions (Chris)
W-L: 1-12
Last Week: 12
Average Ballot Position: 12.0
Expected Wins: 0.2 (+0.8)
Ouch. They shouldn’t have won at all.
11. Jehovah's Jockstrap (Brett)
W-L: 4-9
Last Week: 10
Average Ballot Position: 11.0
Expected Wins: 1.2 (+2.8)
Well, you do have to be pretty lucky to protect God’s nether regions.
10. SayHi2YourMother4Me (Kenan)
W-L: 5-8
Last Week: 11
Average Ballot Position: 9.4
Expected Wins: 6.7 (-1.7)
Weeks 5 and 13 really inflated these numbers for the Mothers.
9. The Dyin' SharkLions (Danny)
W-L: 6-7
Last Week: 6
Average Ballot Position: 8.4
Expected Wins: 7.3 (-1.3)
A borderline playoff team that was simply more unlucky than teams around them.
8. Obese Geese (Steve)
W-L: 7-6
Last Week: 5
Average Ballot Position: 6.7
Expected Wins: 2.9 (+4.1)
As we all suspected, the Geese have been unbelievably lucky in fantasy football as well as in life!
7. Stars (Dan)
W-L: 6-7
Last Week: 9
Average Ballot Position: 6.4
Expected Wins: 9.3 (-3.3)
Would you like to sleep on the Stars? I wouldn’t.
6. A Sheep at the Wheel (Jaron)
W-L: 6-7
Last Week: 7
Average Ballot Position: 6.0
Expected Wins: 5.8 (+0.2)
They had the hardest schedule, so that’s why the Sheep’s number is so low.
5. Cupcakes of DOOM! (Adam)
W-L: 7-6
Last Week: 8
Average Ballot Position: 5.9
Expected Wins: 9.7 (-2.7)
Second best team in the league by this measure…a surprise! Kinda like cupcakes…that kill you…
4. Bacon Cheeseburgers (Andy K.)
W-L: 9-4
Last Week: 3
Average Ballot Position: 3.9
Expected Wins: 7.9 (1.1)
One of the favorites most the year is fading at the end…too much grease?
3. The Schmuckernators (Jason)
W-L: 8-5
Last Week: 4
Average Ballot Position: 3.9
Expected Wins: 9.1 (-1.1)
The Schmucks have been making a strong push of late and it shows up here.
2. Z Power (Andy G.)
W-L: 9-4
Last Week: 2
Average Ballot Position: 3.3 (1)
Expected Wins: 7.5 (+1.5)
The late season fade suggests the Z’s were more than a bit lucky at the beginning of the year.
1. Rostrum Records (Luke)
W-L: 10-3
Last Week: 1
Average Ballot Position: 1.1 (6)
Expected Wins: 12.5 (-2.5)
Wow, these guys have actually underperformed on the year. Scary.
And it’s not over yet! The postseason start tonight, and it’s as much of a tossup from 1-8 as I can ever remember. Probably won’t have any time to do much recap and analysis of these upcoming games, but I’ll see what I can do. If nothing else appears, look for the postseason awards ballot in about three weeks, and good luck snagging the ultimate crown!
All right, let’s get a bit nerdy for a second (or more than a second, if you want to get technical, which, come to think of it is even more nerdy. Whatever.). I’ve added one more category to the final stats: expected wins. Including it was inspired by some back and forth with Jaron, who is adamant (and incorrect, at least this year) that points against is a better predictor of how you do than points scored – in other words, that you have little control over your fantasy football fate. This is quite easily refuted by comparing charts (a) and (b), which show the linear regression of points scored and points against; pay attention to the r-squared value. R-squared indicates the predictive power of a certain variable; 1 is a perfect predictor, while 0 tells you nothing. Note that r-squared is significantly higher for points scored than points against. QED. However, points scored isn’t a perfect predictor…a slightly better one is something called expected wins, which is shown in (c), and takes into account both points scored and points against.
A couple of notes about the expected wins: I calculated them using Hollinger’s Pythagorean expected wins, which were developed for basketball. And, they’re not perfect – for one thing, unlike real teams, we have no defensive control over how well our opponents do, so points against doesn’t add a whole lot of extra value. Basically, all these stats can do is give you some idea of is how lucky or unlucky a team has been – below, positive numbers relative to actual wins mean the team has been lucky while negative means unlucky. Well, that turned out a bit longer than I intended. Let’s get to the Power Rankings charts, and the final Power Rankings standings.
12. Minions (Chris)
W-L: 1-12
Last Week: 12
Average Ballot Position: 12.0
Expected Wins: 0.2 (+0.8)
Ouch. They shouldn’t have won at all.
11. Jehovah's Jockstrap (Brett)
W-L: 4-9
Last Week: 10
Average Ballot Position: 11.0
Expected Wins: 1.2 (+2.8)
Well, you do have to be pretty lucky to protect God’s nether regions.
10. SayHi2YourMother4Me (Kenan)
W-L: 5-8
Last Week: 11
Average Ballot Position: 9.4
Expected Wins: 6.7 (-1.7)
Weeks 5 and 13 really inflated these numbers for the Mothers.
9. The Dyin' SharkLions (Danny)
W-L: 6-7
Last Week: 6
Average Ballot Position: 8.4
Expected Wins: 7.3 (-1.3)
A borderline playoff team that was simply more unlucky than teams around them.
8. Obese Geese (Steve)
W-L: 7-6
Last Week: 5
Average Ballot Position: 6.7
Expected Wins: 2.9 (+4.1)
As we all suspected, the Geese have been unbelievably lucky in fantasy football as well as in life!
7. Stars (Dan)
W-L: 6-7
Last Week: 9
Average Ballot Position: 6.4
Expected Wins: 9.3 (-3.3)
Would you like to sleep on the Stars? I wouldn’t.
6. A Sheep at the Wheel (Jaron)
W-L: 6-7
Last Week: 7
Average Ballot Position: 6.0
Expected Wins: 5.8 (+0.2)
They had the hardest schedule, so that’s why the Sheep’s number is so low.
5. Cupcakes of DOOM! (Adam)
W-L: 7-6
Last Week: 8
Average Ballot Position: 5.9
Expected Wins: 9.7 (-2.7)
Second best team in the league by this measure…a surprise! Kinda like cupcakes…that kill you…
4. Bacon Cheeseburgers (Andy K.)
W-L: 9-4
Last Week: 3
Average Ballot Position: 3.9
Expected Wins: 7.9 (1.1)
One of the favorites most the year is fading at the end…too much grease?
3. The Schmuckernators (Jason)
W-L: 8-5
Last Week: 4
Average Ballot Position: 3.9
Expected Wins: 9.1 (-1.1)
The Schmucks have been making a strong push of late and it shows up here.
2. Z Power (Andy G.)
W-L: 9-4
Last Week: 2
Average Ballot Position: 3.3 (1)
Expected Wins: 7.5 (+1.5)
The late season fade suggests the Z’s were more than a bit lucky at the beginning of the year.
1. Rostrum Records (Luke)
W-L: 10-3
Last Week: 1
Average Ballot Position: 1.1 (6)
Expected Wins: 12.5 (-2.5)
Wow, these guys have actually underperformed on the year. Scary.
4 Comments:
I can volunteer to do some playoff analysis week to week if you're going to be overtaxed!
Week 7 is funny. Chris voted a bunch of times and screwed up the results!
@ Jason - Nope, 1 vote per person!
Well, there was a conspiracy, but I'm not naming names!
Post a Comment
<< Home