Final Season Rankings
OK, here's my final rankings for 2008. Unlike during the season, where I try to create snapshots of who is better right that minute, this ranking will use data from the entire season to see who was better all season long.
Again, these rankings are based on analysis of total game scores. I group the scores into sextiles (kind of like quartiles, but with 6 groups not 4). I use sextiles because 6 divides nicely into the 12 teams we have in the league. This means that in an average week 2 teams will score in each sextile. So if you score in the top sextile you should expect only 1 other team to even be in your neighborhood. Of course, it rarely happens that scores are evenly distributed that week. But over time I think this method does a good job of showing which teams did consitently best and how many games they should have won based on expected distributions of scores.
Now that I'm done trying to convince everyone that these rankings are not the ravings of a lunatic, on with the fun stuff.
1) Mimes (pre draft: 11th, post draft: 8th)
My formulas expected Mimes to win 9.5 games -- they won 9. They led the league with the most game scores in the top sextile and second sextile (4 each). They also were one of only 2 teams to not score in the bottom sextile all season long. And with only 2 scores in the 5th sextile, they had a league low 2 scores combined in the bottom two sextiles. Their 9.5 expected wins is the second highest number in the three years that I've made this estimate (Capo had 10.0 expected wins last year). It also represents an improvement of 3.9 expected wins over last year. This is the second best season-to-season increase since Malleus took over Lightning and improved them by 4.3 expected wins last year.
2) Flannel (pre draft: 4th, post draft: 7th)
Flannel graded out with 9.0 expected wins this year and won 9.5 (I'm counting ties as half of a win). Like Mimes, Flannel scored in the bottom two sextiles only twice. They scored in the top two sextiles 7 times.
3-tie) Buffalo (pre draft: 3rd, post draft: 1st)
This might seem like a surprise since John only won 5 games this season. But my calculations suggest his team should have won more like 7.9 games. Buffalo scored in the top two sextiles 6 times (which is fourth most in the league this year) and only scored in the bottom sextile once. However, he lost one of the games when he scored in the top sextile and did not win a single game when he did not score in one of the top two sextiles.
3-tie) Drow (pre draft: 8th, post draft: 4th)
In an interesting contrast with Buffalo, Drow also have an expected win total of 7.9. Unlike Buffalo they won 9 games. Not surprisingly, their score distribution was similar although even more skewed to the ends of the scale. Drow tied Mimes with a league leading 4 games in the top sextile and tied Flannel for the second most games in the top two sextiles combined (7) but also scored in the bottom sextile 3 times and only Brawlers and Hamburglers scored more games in the bottom two sextiles (Drow had 5, they had 7 and 8). Obviously Drow was luckier then Buffalo. Drow won one game when they scored in the 4th sextile and one game where they scored in the 5th sextile.
5-tie) Old No. 7 (pre draft: 12th, post draft: 12th)
Another Middle Division team that underperformed this year. Old No. 7 should have won more like 7.5 games this year but only won 4.5. No other team this year had more then 4 game scores in the same sextile. Old No. 7 had 6 game scores in the 3rd sextile, or just above the mean. But they obviously could not convert them into wins, going 1-4-1 in those games.
5-tie) Capo (pre draft: 1st, post draft: 3rd)
Capo managed 8.5 wins instead of their 7.5 expected wins. There is no real good way to explain their results as compared to Old N0. 7 other then they were not as unlucky as Old No. 7. They did not score in the bottom sextile all year (7 did it twice) but other then that the whiskey bottle generally outperformed them except where it counts.
7) Ramjets (pre draft: 9th, post draft: 10th)
Given the mid-season hype, I think most people expected to see Ramjets ranked higher then this. But I still have them expected to win 7.4 games this year. They won 4.5. Nothing terribly odd about their score distributions, although it is interestingly similar to Beavers...
8) Beavers (pre draft: 5th, post draft: 6th)
Beavers graded out with an expected win total of 7.3. Their score distribution was very similar to Ramjets, although Roger's team had some better results in the middle and bottom of the scale. Yet Beavers got 8 wins this year. Yet, in a bit of irony, both teams are scheduling tee-times instead of setting playoff rosters.
9) Malleus (pre draft: 10th, post draft: 9th)
Early in the year I commented that Malleus was like the embodyment of the mean. Turns out my formulas agree and expected 7.0 wins from them this year. A couple really interesting things about this revolving around the fact that is exactly how many wins expected out of Malleus last year. Last year, 7.0 expected wins was good enough to get you ranked 6th, this year... 9th. That means 8 out of 12 teams scored mostly above the mean this year, while some of that slack was taken up by our last two teams this year... this is something odd I might look into more later. Also, this year Malleus managed 9 wins while last year he managed only 5.5. This one might correspond to how Malleus was "average" this year as opposed to last year. Last year, Malleus mostly scored around the mean -- 8 of his 14 games scored in the 3rd or 4th sextile. This year, he scored in the middle third only 4 times and scored in the top two and bottom two sextiles more frequently then last year.
10) Bill (pre draft: 7th, post draft: 5th)
Yes, a playoff team this low. But Bill graded out with just 6.8 expected wins. Although note that this really isn't terrible and is also not even close to the worst expected win total for a playoff team in the last 3 years (07 Beavers & Brawlers = 6.8, 06 Drow = 6.3, 06 Flannel = 5.4, 07 Old No. 7 = 5.3, and 06 Old No. 7 = 4.3). Also note that for a team ranked this far down the scale, Bill did score in the top sextile 3 times, which is more then any other team not ranked in the top 4.
11) Brawlers (pre draft: 6th, post draft: 11th)
Brawlers late season surge helped keep them out of the basement at the end of the season. Their expected win total of 4.6 is better then 6 others over the last 3 years. But they were consistently bad. They tied Hamburglers with 4 scores in the bottom sextile and their 7 total scores in the bottom 2 sextiles was only 1 short of Hamburglers. They only scored in the top 3 sextiles 3 times all season.
12) Hamburglers (pre draft: 2nd, post draft: 2nd)
Expected wins for Hamburglers was 3.9. They only scored in the top 3 sextiles twice all year and led the league in any combination of scores you can think of involving the bottom 3 sextiles. Scoring 4 times in each of them.
Again, these rankings are based on analysis of total game scores. I group the scores into sextiles (kind of like quartiles, but with 6 groups not 4). I use sextiles because 6 divides nicely into the 12 teams we have in the league. This means that in an average week 2 teams will score in each sextile. So if you score in the top sextile you should expect only 1 other team to even be in your neighborhood. Of course, it rarely happens that scores are evenly distributed that week. But over time I think this method does a good job of showing which teams did consitently best and how many games they should have won based on expected distributions of scores.
Now that I'm done trying to convince everyone that these rankings are not the ravings of a lunatic, on with the fun stuff.
1) Mimes (pre draft: 11th, post draft: 8th)
My formulas expected Mimes to win 9.5 games -- they won 9. They led the league with the most game scores in the top sextile and second sextile (4 each). They also were one of only 2 teams to not score in the bottom sextile all season long. And with only 2 scores in the 5th sextile, they had a league low 2 scores combined in the bottom two sextiles. Their 9.5 expected wins is the second highest number in the three years that I've made this estimate (Capo had 10.0 expected wins last year). It also represents an improvement of 3.9 expected wins over last year. This is the second best season-to-season increase since Malleus took over Lightning and improved them by 4.3 expected wins last year.
2) Flannel (pre draft: 4th, post draft: 7th)
Flannel graded out with 9.0 expected wins this year and won 9.5 (I'm counting ties as half of a win). Like Mimes, Flannel scored in the bottom two sextiles only twice. They scored in the top two sextiles 7 times.
3-tie) Buffalo (pre draft: 3rd, post draft: 1st)
This might seem like a surprise since John only won 5 games this season. But my calculations suggest his team should have won more like 7.9 games. Buffalo scored in the top two sextiles 6 times (which is fourth most in the league this year) and only scored in the bottom sextile once. However, he lost one of the games when he scored in the top sextile and did not win a single game when he did not score in one of the top two sextiles.
3-tie) Drow (pre draft: 8th, post draft: 4th)
In an interesting contrast with Buffalo, Drow also have an expected win total of 7.9. Unlike Buffalo they won 9 games. Not surprisingly, their score distribution was similar although even more skewed to the ends of the scale. Drow tied Mimes with a league leading 4 games in the top sextile and tied Flannel for the second most games in the top two sextiles combined (7) but also scored in the bottom sextile 3 times and only Brawlers and Hamburglers scored more games in the bottom two sextiles (Drow had 5, they had 7 and 8). Obviously Drow was luckier then Buffalo. Drow won one game when they scored in the 4th sextile and one game where they scored in the 5th sextile.
5-tie) Old No. 7 (pre draft: 12th, post draft: 12th)
Another Middle Division team that underperformed this year. Old No. 7 should have won more like 7.5 games this year but only won 4.5. No other team this year had more then 4 game scores in the same sextile. Old No. 7 had 6 game scores in the 3rd sextile, or just above the mean. But they obviously could not convert them into wins, going 1-4-1 in those games.
5-tie) Capo (pre draft: 1st, post draft: 3rd)
Capo managed 8.5 wins instead of their 7.5 expected wins. There is no real good way to explain their results as compared to Old N0. 7 other then they were not as unlucky as Old No. 7. They did not score in the bottom sextile all year (7 did it twice) but other then that the whiskey bottle generally outperformed them except where it counts.
7) Ramjets (pre draft: 9th, post draft: 10th)
Given the mid-season hype, I think most people expected to see Ramjets ranked higher then this. But I still have them expected to win 7.4 games this year. They won 4.5. Nothing terribly odd about their score distributions, although it is interestingly similar to Beavers...
8) Beavers (pre draft: 5th, post draft: 6th)
Beavers graded out with an expected win total of 7.3. Their score distribution was very similar to Ramjets, although Roger's team had some better results in the middle and bottom of the scale. Yet Beavers got 8 wins this year. Yet, in a bit of irony, both teams are scheduling tee-times instead of setting playoff rosters.
9) Malleus (pre draft: 10th, post draft: 9th)
Early in the year I commented that Malleus was like the embodyment of the mean. Turns out my formulas agree and expected 7.0 wins from them this year. A couple really interesting things about this revolving around the fact that is exactly how many wins expected out of Malleus last year. Last year, 7.0 expected wins was good enough to get you ranked 6th, this year... 9th. That means 8 out of 12 teams scored mostly above the mean this year, while some of that slack was taken up by our last two teams this year... this is something odd I might look into more later. Also, this year Malleus managed 9 wins while last year he managed only 5.5. This one might correspond to how Malleus was "average" this year as opposed to last year. Last year, Malleus mostly scored around the mean -- 8 of his 14 games scored in the 3rd or 4th sextile. This year, he scored in the middle third only 4 times and scored in the top two and bottom two sextiles more frequently then last year.
10) Bill (pre draft: 7th, post draft: 5th)
Yes, a playoff team this low. But Bill graded out with just 6.8 expected wins. Although note that this really isn't terrible and is also not even close to the worst expected win total for a playoff team in the last 3 years (07 Beavers & Brawlers = 6.8, 06 Drow = 6.3, 06 Flannel = 5.4, 07 Old No. 7 = 5.3, and 06 Old No. 7 = 4.3). Also note that for a team ranked this far down the scale, Bill did score in the top sextile 3 times, which is more then any other team not ranked in the top 4.
11) Brawlers (pre draft: 6th, post draft: 11th)
Brawlers late season surge helped keep them out of the basement at the end of the season. Their expected win total of 4.6 is better then 6 others over the last 3 years. But they were consistently bad. They tied Hamburglers with 4 scores in the bottom sextile and their 7 total scores in the bottom 2 sextiles was only 1 short of Hamburglers. They only scored in the top 3 sextiles 3 times all season.
12) Hamburglers (pre draft: 2nd, post draft: 2nd)
Expected wins for Hamburglers was 3.9. They only scored in the top 3 sextiles twice all year and led the league in any combination of scores you can think of involving the bottom 3 sextiles. Scoring 4 times in each of them.
Labels: 2008
0 Comments:
Post a Comment
Subscribe to Post Comments [Atom]
<< Home