Started By
Message
My Computer Top 25: Week 5, 2013
Posted on 10/1/13 at 12:05 am
Posted on 10/1/13 at 12:05 am
Week 4 Thread
(open image in new tab for a larger view)
Rank, [AP Rank], Team, (Previous Week)
1. [1]--Alabama (1)
2. [6]--Georgia (11)
3. [3]--Clemson (3)
4. [19]-Michigan (2)
5. [11]-Oklahoma (18)
6. [14]-Miami (Fl) (5)
7. [4]--Ohio State (15)
8. [13]-South Carolina (NR)
9. [24]-Ole Miss (4)
10.[15]-Washington (13)
11.[5]--Stanford (10)
12.[20]-Texas Tech (8)
13.[12]-UCLA (9)
14.[NR]-Virginia Tech (NR)
15.[10]-LSU (6)
16.[23]-Fresno State (19)
17.[NR]-Houston (25)
18.[8]--Florida State (22)
19.[NR]-Missouri (22)
20.[NR]-N. Illinois (NR)
21.[7]--Louisville (16)
21.[16]-Northwestern (16)
23.[25]-Maryland (NR)
24.[9]--Texas A&M (NR)
25.[2]--Oregon (NR)
21/25 of my computer's top 25 are also in the current AP top 25 poll, with Alabama and Clemson perfectly agreed upon.
SEC Rankings: 7/14 are in top 25
1. Alabama
2. Georgia
8. South Carolina
9. Ole Miss
15. LSU
19. Missouri
24. Texas A&M
31. Auburn
36. Florida
45. Tennessee
58. Vanderbilt
66. Arkansas
73. Mississippi State
92. Kentucky
I'll try my best to rationalize your bickering
(open image in new tab for a larger view)
Rank, [AP Rank], Team, (Previous Week)
1. [1]--Alabama (1)
2. [6]--Georgia (11)
3. [3]--Clemson (3)
4. [19]-Michigan (2)
5. [11]-Oklahoma (18)
6. [14]-Miami (Fl) (5)
7. [4]--Ohio State (15)
8. [13]-South Carolina (NR)
9. [24]-Ole Miss (4)
10.[15]-Washington (13)
11.[5]--Stanford (10)
12.[20]-Texas Tech (8)
13.[12]-UCLA (9)
14.[NR]-Virginia Tech (NR)
15.[10]-LSU (6)
16.[23]-Fresno State (19)
17.[NR]-Houston (25)
18.[8]--Florida State (22)
19.[NR]-Missouri (22)
20.[NR]-N. Illinois (NR)
21.[7]--Louisville (16)
21.[16]-Northwestern (16)
23.[25]-Maryland (NR)
24.[9]--Texas A&M (NR)
25.[2]--Oregon (NR)
21/25 of my computer's top 25 are also in the current AP top 25 poll, with Alabama and Clemson perfectly agreed upon.
SEC Rankings: 7/14 are in top 25
1. Alabama
2. Georgia
8. South Carolina
9. Ole Miss
15. LSU
19. Missouri
24. Texas A&M
31. Auburn
36. Florida
45. Tennessee
58. Vanderbilt
66. Arkansas
73. Mississippi State
92. Kentucky
I'll try my best to rationalize your bickering
12
Posted on 10/1/13 at 12:07 am to Harry Caray
quote:
4. [19]-Michigan (2)
quote:
I'll try my best to rationalize your bickering
Begin please.
Posted on 10/1/13 at 12:08 am to Harry Caray
looks pretty good to me
Posted on 10/1/13 at 12:08 am to Harry Caray
quote:I like
19.[NR]-Missouri (22)
Posted on 10/1/13 at 12:11 am to Harry Caray
quote:
14.[NR]-Virginia Tech (NR)
ha
Posted on 10/1/13 at 12:15 am to dbt_Geaux_Tigers_196
Michigan usually picks up the shite first. Well for starters Michigan converted all their possible points (4.618), which is relatively a good bit at this point in the season, but still a far cry from Alabama's 9.375. Michigan's weighted away win (which most ranked teams lack), albeit UConn, and a "big win" over then-ranked Notre Dame gave them great opportunities. I know they only beat Akron by 4 and UConn by 3, but the system does not reward or penalize margin, just outcome. They'll eventually drop a game or three and be phased out.
Posted on 10/1/13 at 12:16 am to dbt_Geaux_Tigers_196
quote:
Begin please.
Sample Size.
That was easy.
Posted on 10/1/13 at 12:18 am to beaver
quote:
14.[NR]-Virginia Tech (NR)
ha
Away wins are key in Virginia Tech's case. If their wins against ECU and GTech are home games, they'd be back in the 30s
Posted on 10/1/13 at 12:20 am to Harry Caray
Looks like its coming together. Looking forward to seeing it in a couple weeks.
Posted on 10/1/13 at 12:20 am to JuiceTerry
this is bullfrickin shite fk ur computer
Posted on 10/1/13 at 12:21 am to Harry Caray
Well, your system is interesting, be curious to see how it pans out as season goes by. Regarding Meeeeeshigan, yeah they beat UConn by 3...a team beaten by Towson (17 pts) and Buffalo (29 pts). A near miss against Akron.....Akron.
Posted on 10/1/13 at 12:22 am to Harry Caray
Do you recalculate points each time you do pod swaps?
That seems to be the one flaw I can find. If, lets say LSU beat TCU, who at the time was tier 1 in week one, and TCU ends up going 5-7. LSU would end up having a tier 1 win for that game even though TCU could have fallen to tier 6 or 7.
That seems to be the one flaw I can find. If, lets say LSU beat TCU, who at the time was tier 1 in week one, and TCU ends up going 5-7. LSU would end up having a tier 1 win for that game even though TCU could have fallen to tier 6 or 7.
Posted on 10/1/13 at 12:27 am to Cap Crunch
quote:
That seems to be the one flaw I can find. If, lets say LSU beat TCU, who at the time was tier 1 in week one, and TCU ends up going 5-7. LSU would end up having a tier 1 win for that game even though TCU could have fallen to tier 6 or 7.
And that's one thing that's been pestering me this whole time. If I did do that, it'd be incredibly tedious and would cause looping that I doubt I could keep up with.
I'm tossing around the idea for next year of weighing games lighter at the beginning of the season and heavier as the season goes on. But that could minimize a big week 1 win.
So I decided to be lazy and award points as they currently stand.
ETA: plus I figure it won't matter that much. I just recalculated LSU's ranking and they'd be 16th instead of 15th. Because teams initially ranked very high shouldn't drop to the lower five pods (unless we're talking preseason #1 2012 USC), and the already low-ranked teams rewards wouldn't vary significantly enough for any major changes.
This post was edited on 10/1/13 at 12:32 am
Posted on 10/1/13 at 12:42 am to Harry Caray
Yeah TCU was just one I thought off. I'm sure there could be much better examples
Posted on 10/1/13 at 12:53 am to Harry Caray
Interesting and intriguing.. UT @ 45 throws me off.
Posted on 10/1/13 at 12:55 am to Harry Caray
Interesting metric you are using for the rankings. I do have three questions.
1. When you calculate the scores are they recalculated based on every team performance that week then added to their total? Or do you start from 0, use your metric and change the values to reflect wins or losses to spit out the numbers and which pod teams are in.
2. When you factor in wins or losses on the road, to which value do you multiply or divide the 1.5?
3. If all of Michigan opponents have a negative value, mathematically, how are they number 4? Not sure if I'm following the procedure corectly but this could be answered by commenting on the first questions.
Again good job and please continue this every week if time allows.
1. When you calculate the scores are they recalculated based on every team performance that week then added to their total? Or do you start from 0, use your metric and change the values to reflect wins or losses to spit out the numbers and which pod teams are in.
2. When you factor in wins or losses on the road, to which value do you multiply or divide the 1.5?
3. If all of Michigan opponents have a negative value, mathematically, how are they number 4? Not sure if I'm following the procedure corectly but this could be answered by commenting on the first questions.
Again good job and please continue this every week if time allows.
Posted on 10/1/13 at 1:00 am to dbt_Geaux_Tigers_196
Im not sure how much value is put in their win against ND and the teams they've played but I would expect them to drop when they lose and if ND sucks this year.
Posted on 10/1/13 at 1:04 am to Buttermilk Pancakes
quote:
1. When you calculate the scores are they recalculated based on every team performance that week then added to their total? Or do you start from 0, use your metric and change the values to reflect wins or losses to spit out the numbers and which pod teams are in.
The current points value each team has is a summation of the points earned/lost by the result against all opponents' strength at the time of the game. Adjusting every week would be really fricky.
quote:
2. When you factor in wins or losses on the road, to which value do you multiply or divide the 1.5?
If the away team wins, the points earned are multiplied by 1.5. Example: If a team beats Rutgers (Pod 4) on the road next week, they will earn 3/(4) points, multiplied by 1.5, to earn 1.125 points for the week.
On the other side, if an away team loses, the points the team would lose is divided by 1.5. Losses are usually equivalent to the opponent's Pod #. Take LSU this past week, lost to Pod 1 Georgia on the road. That's -1/1.5 for a total of -.667 points for the week.
quote:
3. If all of Michigan opponents have a negative value, mathematically, how are they number 4? Not sure if I'm following the procedure corectly but this could be answered by commenting on the first questions.
Like I said, the position of Notre Dame at the time of the game and Michigan's away win helps their case.
Posted on 10/1/13 at 1:06 am to Buttermilk Pancakes
quote:
Im not sure how much value is put in their win against ND and the teams they've played but I would expect them to drop when they lose and if ND sucks this year.
If they lose at home to Minnesota next week, Michigan will drop to 27 against the current rankings.
Posted on 10/1/13 at 1:17 am to Harry Caray
Interesting forumla. It's good to see you value teams who play good competiton but don't penalize them too much if the lose.
One last question. When you assigned teams to pods before the season, what method or poll did you use?
One last question. When you assigned teams to pods before the season, what method or poll did you use?
This post was edited on 10/1/13 at 1:20 am
Page 1 of 2
Popular
Back to top
Follow SECRant for SEC Football News