Congrats to the Auburn Bay Oilers for hanging on to the #1 spot again this week!

Week 15 Rankings..
01 Auburn Bay Oilers (E)
02 Edmonton Executors (E)
03 The 300 Spartan Coyotes (+1)
04 Kipton Sabres (+1)
05 Shankers (-2)
06 Las Vegas Hogs (E)
07 Philadelphia Eagles (E)
08 Syosset Braves (+1)
09 Red Gremlins (+1)
10 I Ride The Bus (+1)
11 WYC Wolverines (+1)
12 Toronto Thunder (+3)
13 Syracuse Redmen (+4)
14 ghost riders (E)
15 New Jersey Hitmen (+4)
16 Osler Rebels (-3)
17 NY Stompers (+3)
18 Greenport Storm (-2)
19 Dallas Meteors (-1)
20 Monsters of Midway (NR)

Oh So Close..
Barrow Whalers
Los Angeles Rams

Gone bot..
Fredericksburg Freakshow

Congrats to the Auburn Bay Oilers for hanging on to the #1 spot again this week!

Week 17 Rankings..
01 Auburn Bay Oilers (E)
02 Edmonton Executors (E)
03 The 300 Spartan Coyotes (+1)
04 Kipton Sabres (-1)
05 Shankers (E)
06 Las Vegas Hogs (+1)
07 Philadelphia Eagles (-1)
08 Syosset Braves (E)
09 WYC Wolverines (E)
10 Syracuse Redmen (+1)
11 ghost riders (+1)
12 Red Gremlins (-2)
13 I Ride The Bus (+1)
14 Toronto Thunder (+1)
15 New Jersey Hitmen (E)
16 Osler Rebels (E)
17 NY Stompers (E)
18 Dallas Meteors (E)
19 Greenport Storm (E)
20 Los Angeles Rams (NR)

Oh So Close..
Monsters of Midway
Barrow Whalers

Gone bot..
No One!

labianca2016 wrote: Love to see these every week. Is there room to expand the stats if you had some assistance doing some deep dives on teams each week? Reasons I ask are self-serving but interesting nonetheless. I'll share: I have beaten Red Gremlins twice in league play. Auburn Bay once in friendly, and been smacked around pretty badly by the 300 Spartan Bay team in Cup play. Based on h2h I know my team is comparable to WYC Wolverines as our matches have been pretty competitive when I've not done something stupid like botted a lineup. (and I've done that a couple times in friendly cup play this season and twice against WYC.). So while I'm in the top 15, I know my team is better than that place would suggest, though likely not in the top 5-8. I'm betting that there's some level of granularity that would better rank the teams that is not approachable given the time commitment needed. I'm willing to assist if you have some thoughts. Of course, I can stop botting matches and that will help the cause

There's always been a bit of uncertainty with the rankings. botted games will definitely skew your points, since you may be easily able to defeat a team, but you forgot to set your lineup. The loss itself doesn't count against you, but you don't get the points and bonus' added to your point total, so in a way, it does hurt you. And quite frankly, IMO, if you bot a game, you deserve to have your rankings affected negatively, even if you're the best team in your league/division.

Same with teams in a league populated with several bots. Take Division 3 for instance. Teams in 3.3 and 3.4 have only 2 bots in their league. Richmond Rams and Sith, in 3.8, on the other hand, have 8 bots in their league. They have 12 fewer human vs. human games in a season, and without human teams to play against, all I have to go on is PD. But how is that fair, they have 12 games they can win 80-0 while other teams are facing human teams with resulting closer scores.

I've made tweaks to the formula over the years and it's pretty close now. My goal was that the final rankings in D.III should have the 9 teams promoting to D.II at or near the top. Because of bots and what not, it doesn't always happen, but that's life. For the TD/DII rankings, I hoped to have the 7 teams remaining in the TD and the 3 teams promoting from D.II at or very near the top.

These are the stats I track:
Wins
Losses (only for the end of season summary)
Pts For
Pts Against
Wins vs Humans (League)
Wins vs Humans (Cup)
Cup Win over Higher Div Opponent (1 pt for D.III win / 2 pts for D.II win / 3 pts for a TD win)
Rounds in Cup (Everyone gets at least 1 for the first round)

I have certain points weighted depending on which division you're in.
(X is a multiplication factor that I use to try and balance things and varies by stat and division)

League calculations....
((Wins * X) + (Human wins * X) + ((Pts For - Pts Against) * X))
+
Cup Calculations....
((Human Cup wins * 200) + (Upper Div. Human Cup win * X) + (Rounds advanced in Cup))
+
Div. Bonus (500 for TD, -150 for D.II, -300 for D.III, -400 for D.IV)


TD teams get bonus' for points scored, league wins, cup wins and for being in the TD. So when a D.2 teams breaks into the top 7, you know they're pretty good, because they're there without the TD bonus'. I also lowered the TD multiplier for PD.. just based on an eyeball test. I tried it a couple of different ways and this just seemed to "look" right.

Reaus wrote: labianca2016 wrote: Love to see these every week. Is there room to expand the stats if you had some assistance doing some deep dives on teams each week? Reasons I ask are self-serving but interesting nonetheless. I'll share: I have beaten Red Gremlins twice in league play. Auburn Bay once in friendly, and been smacked around pretty badly by the 300 Spartan Bay team in Cup play. Based on h2h I know my team is comparable to WYC Wolverines as our matches have been pretty competitive when I've not done something stupid like botted a lineup. (and I've done that a couple times in friendly cup play this season and twice against WYC.). So while I'm in the top 15, I know my team is better than that place would suggest, though likely not in the top 5-8. I'm betting that there's some level of granularity that would better rank the teams that is not approachable given the time commitment needed. I'm willing to assist if you have some thoughts. Of course, I can stop botting matches and that will help the cause There's always been a bit of uncertainty with the rankings. botted games will definitely skew your points, since you may be easily able to defeat a team, but you forgot to set your lineup. The loss itself doesn't count against you, but you don't get the points and bonus' added to your point total, so in a way, it does hurt you. And quite frankly, IMO, if you bot a game, you deserve to have your rankings affected negatively, even if you're the best team in your league/division. Same with teams in a league populated with several bots. Take Division 3 for instance. Teams in 3.3 and 3.4 have only 2 bots in their league. Richmond Rams and Sith, in 3.8, on the other hand, have 8 bots in their league. They have 12 fewer human vs. human games in a season, and without human teams to play against, all I have to go on is PD. But how is that fair, they have 12 games they can win 80-0 while other teams are facing human teams with resulting closer scores. I've made tweaks to the formula over the years and it's pretty close now. My goal was that the final rankings in D.III should have the 9 teams promoting to D.II at or near the top. Because of bots and what not, it doesn't always happen, but that's life. For the TD/DII rankings, I hoped to have the 7 teams remaining in the TD and the 3 teams promoting from D.II at or very near the top. These are the stats I track: Wins Losses (only for the end of season summary) Pts For Pts Against Wins vs Humans (League) Wins vs Humans (Cup) Cup Win over Higher Div Opponent (1 pt for D.III win / 2 pts for D.II win / 3 pts for a TD win) Rounds in Cup (Everyone gets at least 1 for the first round) I have certain points weighted depending on which division you're in. (X is a multiplication factor that I use to try and balance things and varies by stat and division) League calculations.... ((Wins * X) + (Human wins * X) + ((Pts For - Pts Against) * X)) + Cup Calculations.... ((Human Cup wins * 200) + (Upper Div. Human Cup win * X) + (Rounds advanced in Cup)) + Div. Bonus (500 for TD, -150 for D.II, -300 for D.III, -400 for D.IV) TD teams get bonus' for points scored, league wins, cup wins and for being in the TD. So when a D.2 teams breaks into the top 7, you know they're pretty good, because they're there without the TD bonus'. I also lowered the TD multiplier for PD.. just based on an eyeball test. I tried it a couple of different ways and this just seemed to "look" right.

All makes sense. The only thing I might add to it is normalization of division for number of human games played across leagues, eliminating the human skew from the equation.

Unfortunately, that would mean that in your current example with one D3 league containing two human players that their results would either be multiplied to normalize against the league with 8 human players or other leagues would only submit their top results to compare to the human interaction of the gimped league.

I'll play with it and see what I come up with based on this last season's results.

labianca2016 wrote: Reaus wrote: labianca2016 wrote: Love to see these every week. Is there room to expand the stats if you had some assistance doing some deep dives on teams each week? Reasons I ask are self-serving but interesting nonetheless. I'll share: I have beaten Red Gremlins twice in league play. Auburn Bay once in friendly, and been smacked around pretty badly by the 300 Spartan Bay team in Cup play. Based on h2h I know my team is comparable to WYC Wolverines as our matches have been pretty competitive when I've not done something stupid like botted a lineup. (and I've done that a couple times in friendly cup play this season and twice against WYC.). So while I'm in the top 15, I know my team is better than that place would suggest, though likely not in the top 5-8. I'm betting that there's some level of granularity that would better rank the teams that is not approachable given the time commitment needed. I'm willing to assist if you have some thoughts. Of course, I can stop botting matches and that will help the cause There's always been a bit of uncertainty with the rankings. botted games will definitely skew your points, since you may be easily able to defeat a team, but you forgot to set your lineup. The loss itself doesn't count against you, but you don't get the points and bonus' added to your point total, so in a way, it does hurt you. And quite frankly, IMO, if you bot a game, you deserve to have your rankings affected negatively, even if you're the best team in your league/division. Same with teams in a league populated with several bots. Take Division 3 for instance. Teams in 3.3 and 3.4 have only 2 bots in their league. Richmond Rams and Sith, in 3.8, on the other hand, have 8 bots in their league. They have 12 fewer human vs. human games in a season, and without human teams to play against, all I have to go on is PD. But how is that fair, they have 12 games they can win 80-0 while other teams are facing human teams with resulting closer scores. I've made tweaks to the formula over the years and it's pretty close now. My goal was that the final rankings in D.III should have the 9 teams promoting to D.II at or near the top. Because of bots and what not, it doesn't always happen, but that's life. For the TD/DII rankings, I hoped to have the 7 teams remaining in the TD and the 3 teams promoting from D.II at or very near the top. These are the stats I track: Wins Losses (only for the end of season summary) Pts For Pts Against Wins vs Humans (League) Wins vs Humans (Cup) Cup Win over Higher Div Opponent (1 pt for D.III win / 2 pts for D.II win / 3 pts for a TD win) Rounds in Cup (Everyone gets at least 1 for the first round) I have certain points weighted depending on which division you're in. (X is a multiplication factor that I use to try and balance things and varies by stat and division) League calculations.... ((Wins * X) + (Human wins * X) + ((Pts For - Pts Against) * X)) + Cup Calculations.... ((Human Cup wins * 200) + (Upper Div. Human Cup win * X) + (Rounds advanced in Cup)) + Div. Bonus (500 for TD, -150 for D.II, -300 for D.III, -400 for D.IV) TD teams get bonus' for points scored, league wins, cup wins and for being in the TD. So when a D.2 teams breaks into the top 7, you know they're pretty good, because they're there without the TD bonus'. I also lowered the TD multiplier for PD.. just based on an eyeball test. I tried it a couple of different ways and this just seemed to "look" right. All makes sense. The only thing I might add to it is normalization of division for number of human games played across leagues, eliminating the human skew from the equation. Unfortunately, that would mean that in your current example with one D3 league containing two human players that their results would either be multiplied to normalize against the league with 8 human players or other leagues would only submit their top results to compare to the human interaction of the gimped league. I'll play with it and see what I come up with based on this last season's results.

If there is an easier way of updating these I'd like in also, as it's a pain to sort out each week