User Pass
Home Sign Up Contact Log In
Forum > Lv 26 Minor Leagues > Lv 26 Minor League #2 > YTWINS Minors Lv 24 #2 - Week 9 Rankings
Page:
 
Mr. Me2
offline
Link
 
Originally posted by silenthatred
Originally posted by ArGenTin32

Originally posted by silenthatred


Originally posted by ArGenTin32



all im seeing is some excuses right now...blah blah blah.

the fact remains the same...WE BEAT YOU and there is no reason why you should be ranked above us right now...NONE.

Diggits>Oblongs

case closed.


Following that logic, Florida shouldn't have been ranked above Ole Miss. Because after all, they beat them. No way OU should have been ranked above Texas and played in the Big 12 Championship, because after all, Texas beat them. No way anyone should have been ranked above Utah. After all, they were undefeated.


last time I checked when 2 teams met up and played a game the winner of that game is then considered the BETTER of the two teams right? when has that ever changed? this isn't the sorry ass BCS here.


You are arguing over power rankings which is a formula. The BCS also is a formula. Fail.


BTW, what I said can also be applied to the AP, Harris and Coach's poll. The AP poll has nothing to do with the BCS.
Edited by silenthatred on Jul 7, 2009 19:05:59
 
ytwins
offline
Link
 
Originally posted by tmhouse
These rankings are based on a greasemonkey script. Ytwins, did you write the script? If not, then nobody here (assuming none of your wrote the script either) can say it's right.

http://userscripts.org/scripts/show/51628

I just ran the script and got a different output than this thread shows. I show:

Rnk ... Off ... Def ... SoS ... Team
1 ........ 1 ........ 1 ........ 0.519 ........ Cagliari Outbreak
2 ........ 6 ........ 4 ........ 0.420 ........ Jebediah's Oblong Ball Club
3 ........ 3 ........ 5 ........ 0.370 ........ _ _ The Branded _ _
4 ........ 2 ........ 3 ........ 0.235 ........ Kenya Diggits
5 ........ 5 ........ 6 ........ 0.370 ........ Justice League
6 ........ 4 ........ 2 ........ 0.358 ........ Montevideo Bulls
7 ........ 9 ........ 7 ........ 0.469 ........ Deadwood Destroyers
8 ........ 11 ........ 9 ........ 0.321 ........ Polish Pittbulls
9 ........ 8 ........ 8 ........ 0.469 ........ Umea White Werewolves
10 ........ 10 ........ 11 ........ 0.296 ........ Sanibel Monkey Shockers
11 ........ 12 ........ 12 ........ 0.506 ........ Thunder Bay Raging Endzone Predators
12 ........ 13 ........ 15 ........ 0.309 ........ Lambsheim Lions
13 ........ 7 ........ 13 ........ 0.346 ........ Dallas Patriots
14 ........ 14 ........ 14 ........ 0.580 ........ North Alaska Insomniacs
15 ........ 16 ........ 17 ........ 0.556 ........ CUP Rugby
16 ........ 15 ........ 10 ........ 0.506 ........ Sarasota Bluefish
17 ........ 25 ........ 18 ........ 0.519 ........ Wroclaw Huskies
18 ........ 17 ........ 19 ........ 0.716 ........ San Francisco Gold Rush
19 ........ 18 ........ 21 ........ 0.506 ........ Edmonton Stallions
20 ........ 24 ........ 23 ........ 0.605 ........ And 1 Blaze
21 ........ 21 ........ 20 ........ 0.654 ........ Chile Peppers
22 ........ 19 ........ 16 ........ 0.617 ........ Subspace Extremes
23 ........ 20 ........ 25 ........ 0.506 ........ troy trojans
24 ........ 28 ........ 27 ........ 0.691 ........ Die Harders
25 ........ 26 ........ 24 ........ 0.630 ........ Dublin Irishmen
26 ........ 27 ........ 26 ........ 0.531 ........ São Paulo Vultures
27 ........ 23 ........ 28 ........ 0.395 ........ Bangkok Dragons
28 ........ 22 ........ 22 ........ 0.296 ........ Fredensborg Warriors
29 ........ 31 ........ 31 ........ 0.815 ........ Quito Cañari Miners
30 ........ 32 ........ 32 ........ 0.667 ........ Iguaçu Falls Quilombos
31 ........ 30 ........ 29 ........ 0.691 ........ ARIZONA GRAVEROBBERS
32 ........ 29 ........ 30 ........ 0.531 ........ Sin City Silent Assassins


Yours ranking consider pre-season and regular season game equal; the ranking I use gives regular more weight than pre-season games.
 
ArGenTin32
offline
Link
 
Originally posted by silenthatred
Originally posted by ArGenTin32

Originally posted by silenthatred


Originally posted by ArGenTin32



all im seeing is some excuses right now...blah blah blah.

the fact remains the same...WE BEAT YOU and there is no reason why you should be ranked above us right now...NONE.

Diggits>Oblongs

case closed.


Following that logic, Florida shouldn't have been ranked above Ole Miss. Because after all, they beat them. No way OU should have been ranked above Texas and played in the Big 12 Championship, because after all, Texas beat them. No way anyone should have been ranked above Utah. After all, they were undefeated.


last time I checked when 2 teams met up and played a game the winner of that game is then considered the BETTER of the two teams right? when has that ever changed? this isn't the sorry ass BCS here.


You are arguing over power rankings which is a formula. The BCS also is a formula. Fail.


wtf do you mean fail???

the BCS is a joke and so are these rankings. there is no reason for a team to be "ranked" below a team it has a better record then and has already beaten.
 
tmhouse
offline
Link
 
Originally posted by ytwins
Yours ranking consider pre-season and regular season game equal; the ranking I use gives regular more weight than pre-season games.


No, my settings did not include pre-season at all, only regular season. It also includes team talent, which I assume you do not include in your script settings.
 
ytwins
offline
Link
 
Originally posted by tmhouse
Originally posted by ytwins

Yours ranking consider pre-season and regular season game equal; the ranking I use gives regular more weight than pre-season games.


No, my settings did not include pre-season at all, only regular season. It also includes team talent, which I assume you do not include in your script settings.


No, I select three items;
1. Which part(s) of the season do you want to include? Regular Season (check the box for regular season)
2. Do you want to consider pre-season, regular season, and playoff games to be equal? No
3. Do you want to include team talent? Yes

When I ran the script with yes for Item 2, I got the exact same results that you got.
 
Michael Vick
offline
Link
 
Originally posted by ArGenTin32
Originally posted by ATL_DirtyBird

Originally posted by ArGenTin32


Originally posted by ATL_DirtyBird



Originally posted by ArGenTin32




these rankings are fubar.

no way a team (Jebediahs 8-1) that we (Kenya 9-0) have beaten already should be ranked above us.


Originally posted by ytwins




NOTE: It seems SoS has more weight than head-to-head matchup.


What's the problem?


the problem is that Oblong is indeed NOT better then Kenya.


Did you not read the bold or do you not understand what SoS means?



SoS don't mean a damn thing when we have played them and BEAT them already...lol



Ok so you don't understand what the bold means, got it.

 
ArGenTin32
offline
Link
 
Originally posted by ATL_DirtyBird
Originally posted by ArGenTin32

Originally posted by ATL_DirtyBird


Originally posted by ArGenTin32



Originally posted by ATL_DirtyBird




Originally posted by ArGenTin32





these rankings are fubar.

no way a team (Jebediahs 8-1) that we (Kenya 9-0) have beaten already should be ranked above us.


Originally posted by ytwins





NOTE: It seems SoS has more weight than head-to-head matchup.


What's the problem?


the problem is that Oblong is indeed NOT better then Kenya.


Did you not read the bold or do you not understand what SoS means?



SoS don't mean a damn thing when we have played them and BEAT them already...lol



Ok so you don't understand what the bold means, got it.



Ok so you don't understand that Kenya has beaten Jebediah already, got it.
 
avgbrad
offline
Link
 
Originally posted by ArGenTin32
Originally posted by ATL_DirtyBird

Originally posted by ArGenTin32


Originally posted by ATL_DirtyBird



Originally posted by ArGenTin32




Originally posted by ATL_DirtyBird





Originally posted by ArGenTin32






these rankings are fubar.

no way a team (Jebediahs 8-1) that we (Kenya 9-0) have beaten already should be ranked above us.


Originally posted by ytwins






NOTE: It seems SoS has more weight than head-to-head matchup.


What's the problem?


the problem is that Oblong is indeed NOT better then Kenya.


Did you not read the bold or do you not understand what SoS means?



SoS don't mean a damn thing when we have played them and BEAT them already...lol



Ok so you don't understand what the bold means, got it.



Ok so you don't understand that Kenya has beaten Jebediah already, got it.


you don't understand that the rankings using a series of uber complex formulas can see thru that regular season week 2 game and knows Jebs is better

just sayin
 
tmhouse
offline
Link
 
The only way these rankings are going to be accurate is after the regular season is complete. Even then, there has only been one inter-conference match per team. The ranking should be very accurate within a single conference, and fairly accurate with the two conferences combined. But right now there is still a lot of "what if" in the formula.

There is no way to make the formulas perfect without seeing everyone play everyone. Kenya could have improved our standing if we tried to beat every CPU or low-leveled team 255-0. I'm sure lots of other teams could say the same thing.

There's no point in arguing about the rankings, everyone should know by know that they're not perfect. It should be obvious that the top ten teams are much better than the bottom ten teams. Comparing #5 to #6 doesn't bring much value and either could be right depending on what you consider important. Kenya beat Jed's 30-10; Kenya was better in that game for sure. Kenya and Jebediah's may get another shot in the playoffs - and then we'll see if Kenya is still better.
 
Shardaddy
offline
Link
 
Hey, Newsflash......

rankings mean NOTHING.

Getting worked up over arbitrary rankings? Just prove it in the playoffs after everyone has boosted and put in their good D
 
Ace79603
offline
Link
 
After next week both kenyas and the bulls SOS will Improve
 
avgbrad
offline
Link
 
successful trolling is successful

c'mon, you seriously think me or any of the other guys that came in here gave one crap about the rankings?


 
rutgar
offline
Link
 
I spent a little time digging through the source code. It looks like the rankings script takes the results of actual games and uses those results to simulate other games. I only spent about 5 minutes looking at the code, but I'm not even sure that it uses the actual game results in the rankings. At one point it compares simulated results to actual results.

It's entirely possible that the rankings are based on the the overall real results of a team so far put through a simulation. The script takes the information the teams and simulates a season. Based on the results it looks like the simulation has JOBC going 16-0, Montevideo 15-1, Kenya 14-2. Again, I'm not positive on this but this is what it looks like at a glance.



Another thing I'm not sure the script takes into consideration is turnovers. Consider the Kenya vs JOBC game. If you compare offenses with defenses, the game was pretty even. The game wasn't a blowout, 30-10. Kenya scored 6 points on 2x FGs that came when JOBC turned the ball over in their own end. Kenya scored another 7 points on an interception return.

So what if the code uses yardage numbers from across the season to determine the results of its simulation? If that's the case, the Kenya vs JOBC game is pretty much tie. Both teams gained in the low 200s. The difference was JOBC's turnovers which led directly to 13 of Kenya's points.

The author of the script states that the SOS number doesn't factor into the final rankings and that is correct. It's just a nice number that gets tossed out for people to talk about, not used in any calculations. Still, putting up good numbers against good teams while help when the script goes to simulate unplayed games against weaker opponents while big numbers against weak opponents doesn't necessarily mean simulated wins against good competition.


I used to do rankings based on both points and yardage differential with the final numerical ranking coming from matrix solution. That definitely takes into account head to head games. I was having to maintain the matrix for each league I did by hand though and that got old really quick. I wrote the code to do the matrix calculation in Java though. I'd love to put it all together into a greasemonkey script, but I don't know greasemonkey all that well =/
 
Michael Vick
offline
Link
 
Originally posted by rutgar
I spent a little time digging through the source code. It looks like the rankings script takes the results of actual games and uses those results to simulate other games. I only spent about 5 minutes looking at the code, but I'm not even sure that it uses the actual game results in the rankings. At one point it compares simulated results to actual results.

It's entirely possible that the rankings are based on the the overall real results of a team so far put through a simulation. The script takes the information the teams and simulates a season. Based on the results it looks like the simulation has JOBC going 16-0, Montevideo 15-1, Kenya 14-2. Again, I'm not positive on this but this is what it looks like at a glance.



Another thing I'm not sure the script takes into consideration is turnovers. Consider the Kenya vs JOBC game. If you compare offenses with defenses, the game was pretty even. The game wasn't a blowout, 30-10. Kenya scored 6 points on 2x FGs that came when JOBC turned the ball over in their own end. Kenya scored another 7 points on an interception return.

So what if the code uses yardage numbers from across the season to determine the results of its simulation? If that's the case, the Kenya vs JOBC game is pretty much tie. Both teams gained in the low 200s. The difference was JOBC's turnovers which led directly to 13 of Kenya's points.

The author of the script states that the SOS number doesn't factor into the final rankings and that is correct. It's just a nice number that gets tossed out for people to talk about, not used in any calculations. Still, putting up good numbers against good teams while help when the script goes to simulate unplayed games against weaker opponents while big numbers against weak opponents doesn't necessarily mean simulated wins against good competition.


I used to do rankings based on both points and yardage differential with the final numerical ranking coming from matrix solution. That definitely takes into account head to head games. I was having to maintain the matrix for each league I did by hand though and that got old really quick. I wrote the code to do the matrix calculation in Java though. I'd love to put it all together into a greasemonkey script, but I don't know greasemonkey all that well =/


tl;dr wall of text

cliffs plz
 
Mr. Me2
offline
Link
 
We just wanted to say we were right. Never question us again.

Sincerely,

The Power Rankings
 
Page:
 


You are not logged in. Please log in if you want to post a reply.