User Pass
Home Sign Up Contact Log In
Forum > USA A Leagues > USA A #3 > The Nokes Power Poll - Eastern Conference
Page:
 
mattynokes
offline
Link
 
First off, here's what I posted in the other Power Poll suggestion thread.

Originally posted by mattynokes
Here's an idea of how to get a power poll going with some simple numbers. I don't think yards gained or yards allowed should be used. In the whole scheme of things all that matters in the "W" or "L" and the points you put on the score board. Your yards gained can easily manipulate your Power Rank number if you're constantly driving 70 yards to only turn it over inside the Red Zone.

I give you the Nokes Power Poll formula...

Wins = A
Losses = B
Points Scored Per Game = C
Points Allowed Per Game = D
Last 4 Game Wins = E
Last 4 Game Losses = F
Current Streak = G
Total Games Played = H

Step 1: (A+C+E) - (B+D+F) = Value 1
Step 2: Value 1 + (G) = Value 2
Step 3: Value 2 / H = Power Rank Value

EXAMPLE

After 7 games Team Alpha is 4-3. They've scored an average of 21 points and have allowed an average of 17 points. They're 2-2 in their last 4 games and are on a 1 game losing streak.

A = 4 Wins
B = 3 Losses
C = 21 Points Scored Average
D = 17 Points Allowed Average
E = 2 Last 4 Games Wins
F = 2 Last 4 Games Losses
G = -1 Streak
H = 7 Total Games

Step 1: (4+21+2) - (3+17+2) = 5
Step 2: 5 + (-1) = 4
Step 3: 4 / 7 = 0.571

Power Rank Number = 0.571

This might look difficult, but it's not and takes very little time of looking through the numbers to figure out what to use. And yes there is a shortcut if you're wondering. I just wrote it all out so people can see how the whole process works.


From the above formula I decided to make wins and losses have more weight. The formula now looks like this:

Power Poll Formula

[ (Wins x 4) + (Points Scored Per Game + Last 4 Wins) - (Losses x 4) - (Points Allowed Per Game + Last 4 Losses) + (Streak) ] / Games Played
 
mattynokes
offline
Link
 
So...The Pre-Season Polls...

I did some slight manipulating of the stats as I don't quite think those BBB teams are going to come up here and dominant with 200 points per game in A #3. Also for the fact that Season 7 A teams should improve on their mark from last season. Here's the adjustments that I made...

BBB Teams: Points For and Wins Divided by 3 / Points Allowed and Losses Times 3
A Teams: Wins Times 1.25 / Losses Divided by 1.25
AA Teams: Points For and Wins Times 1.5 / Points Allowed and Losses Divided by 1.5

Note - If a team now came up as say 6-8 their winning percentage was taken for a 16 game schedule. Thus making that team 7-9. Essentially all BBB teams turned up 8-8.

A #3 Eastern Conference Pre-Season Polls

1. Tricity: +5.395
2. East St. Louis: +5.089
3. Chicago Comm: +4.483
4. College Station: +3.082
5. Nevada: +2.450
6. Kansas: +1.476
7. Windy City: +1.181
8. Chicago Storm: +0.706
9. New Jersey: +0.644
10. Carolina: +0.150
11.Youngstown: -0.094
12. Kentucky: -0.431
13. El Paso: -5.194
14. Chicago TCA: -8.024
15. Flathead: Gutted
16. Hilo: CPU
Last edited Feb 22, 2009 16:55:54
 
billsman2
offline
Link
 
I am not sure I understand the ranking system. Last season, Tri-City finished behind EST and Chicago Commandos in wins and had more losses. No biggie as I am sure all have made changes, and rankings are pretty close, just not sure I understand it's point.
Last edited Feb 22, 2009 13:33:45
 
Sam Bolster
offline
Link
 
LOL at Flathead being number 5... how?
 
billsman2
offline
Link
 
Yes, this system seems very flawed. Especially after teams gutted.
Last edited Feb 22, 2009 15:41:44
 
mattynokes
offline
Link
 
This Power Poll isn't biased towards an opinion on a team's performance it's straight up fact of how they've performed. The reason Tricity is ahead of both the Commandos and East STL is because Tricity has a much higher margin of victory which suggests they should play better. Tricity went 11-5 in the regular season while East and CHI both went 12-4, it's not like Tricity played much worse than the others.

Tri beat Chi
STL beat Tri
Chi beat STL

So take that as you will, but I think that erases the argument of that East and CHI should be ahead of Tri. It's just how the numbers work their way out in the formula.

And I forgot about Flathead gutting and rebuilding. They'd been moved to the #15 spot as a result. This is a very good system, just might need some tweaks for football. I used this in a baseball mogul league and once the season got a good deal under way it showed who's hot and who's not very well. The problem is that "W's" and "L's" are a lot more abundant than points scored vs points allowed in baseball (runs rather) and weigh more. In football your scoring is much higher than your record. I'm thinking possibly I need to try to have the record weigh even more than just times by 4.
Last edited Feb 22, 2009 17:15:41
 
robinhoodnik
offline
Link
 
We (Kentucky) ranked the lowest of all of the teams that came up with us?

Okay, sounds good to me.
 
robinhoodnik
offline
Link
 
Originally posted by mattynokes
I used this in a baseball mogul league .


Sportsmogul Baseball Mogul? Mogul Online?
 
masierra
offline
Link
 
The system works, but only for stats that have been accumulated THIS season with the teams playing each other. All the stats used for the formula, as of now, are from last season and it paints a very different picture.

Mattynokes, rightly so, gave an edge to the teams that were in USA A3 last season. I say after week 3 or 4 we'll have a much better idea of how things might shake out.
 
mattynokes
offline
Link
 
Originally posted by robinhoodnik
We (Kentucky) ranked the lowest of all of the teams that came up with us?

Okay, sounds good to me.


Yeah, that's because of allowing the most points per game. That's likely to do with the fact that you guys moved the furthest in the playoffs and played better teams. I did use playoffs in this. Maybe I shouldn't. I'll see about going back and re-doing them without the playoffs being a factor and maybe adding in a +1 for the streak if the team made the playoffs.

Originally posted by masierra
The system works, but only for stats that have been accumulated THIS season with the teams playing each other. All the stats used for the formula, as of now, are from last season and it paints a very different picture.

Mattynokes, rightly so, gave an edge to the teams that were in USA A3 last season. I say after week 3 or 4 we'll have a much better idea of how things might shake out.

Exactly. I'm not even going to do this again until after Week 4. Let a few games be played because if an average team plays 2 gutted teams to start off they're going to look very good while a good team may play the eventually A #3 Championship teams and look horrid. Just no point in doing this until Week 4, but I thought it'd be neat to see how teams stack up going into the season.
Last edited Feb 22, 2009 17:30:46
 
mattynokes
offline
Link
 
Did some manipulating of the stats that would take too much to explain. Basically know that I tried to incorporate the improvement that should happen for previous A teams and the learning curve that BBB teams should expect as well.

A #3 Eastern Conference Pre-Season Polls

1. Tricity: 7.582
2. East St. Louis: 6.859
3. Chicago Comm: 5.973
4. College Station: 4.188
5. Nevada: 2.574
6. Kansas: 2.426
7. Windy City: 1.684
8. Youngstown: 1.094
9. Chicago Storm: 0.582
10. New Jersey: -0.012
11. Carolina: -0.102
12. Kentucky: -0.973
13. El Paso: -5.566
14. Chicago TCA: -5.816
15. Flathead: Gutted
16. Hilo: CPU

Power Poll Formula

[ (Wins x 5) + (Points Scored Per Game) + (Last 4 Wins x 5) - (Losses x 5) - (Points Allowed Per Game) - (Last 4 Losses x 5) + (Streak) ] / Games Played
Last edited Feb 24, 2009 13:12:54
 
billsman2
offline
Link
 
is that total points scored per game of last season? Or last 4 games of last season points scored?

 
mattynokes
offline
Link
 
Originally posted by billsman2
is that total points scored per game of last season? Or last 4 games of last season points scored?



Total points per game. Oh, and also Streak and Last 4 Games weren't used for the opening formula. Momentum doesn't carry over from season to season so no need to include it.
 
billsman2
offline
Link
 
Originally posted by mattynokes
Originally posted by billsman2

is that total points scored per game of last season? Or last 4 games of last season points scored?



Total points per game. Oh, and also Streak and Last 4 Games weren't used for the opening formula. Momentum doesn't carry over from season to season so no need to include it.


Without going back and looking, I thought the Firebirds scored more points than Tri-City. That must not be the case then. We won one more game than they did, as did the Commando's. TriCity is a great team, just trying to figure out how your formula put them above both of these teams. No biggie, Im just a nerd like that I guess.
 
mattynokes
offline
Link
 
Originally posted by billsman2

Without going back and looking, I thought the Firebirds scored more points than Tri-City. That must not be the case then. We won one more game than they did, as did the Commando's. TriCity is a great team, just trying to figure out how your formula put them above both of these teams. No biggie, Im just a nerd like that I guess.


I just went back and double checked and no, Tricity outscored you guys 92.1 ppg to 74.8 ppg. Like I said, their margin of victory is the likely cause of why they came out higher. A good example of a worse record jumping a better one is Nevada and Kansas. Nevada not only out scored Kansas, but also allowed few points as well. Kansas seems to play up (or down) to their opponent. Maybe it's because they play backups against gutted teams, but whatever the reason is I cannot alter the formula for those cases as it would be putting a bias in the formula. Kansas playing up/down to their opponents suggests they are more likely to lose one of their close games as a result. Nevada having a much higher margin of victory suggests that they can and probably will over come some of those losses.

I'm also not exactly sure of what the relation is between the ratings like what a 0.500 gap means vs a 1.250 gap. My hypothesis is as follows:

2.000+ = Team is far superior to the team below them and can beat them easily
1.000 - 1.999 = Team is superior to the team below them
0.750 = Team is clearly better than the team below them, but a loss wouldn't surprise.
0.500 = Team is better than the team below them, but needs to have a good game to win
0.250 = Team isn't too much better than the team below them. Should get an even matchup.
Last edited Feb 23, 2009 14:55:57
 
Page:
 


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