I agree - I don't like "points against" either as a final determinant (nor do I like points for)
Forum > Suggestions > Playoff Seeding Error
-Phaytle-
offline
offline
Originally posted by bhall43
You mean use the only other record in use for a tie breaker and you win? (points for) All those other team statistics you mention would never be used in any tiebreaker ever known.
Hi, I like taking things out of context.
Also, you don't have to only use one statistic. PF-PA works, a combination of others may. I'm not saying it to just to go down the line and use the 9 statistics (whatever#) we are beating them at to not use the one we are behind by 2 in. PA would never be used at any level of real football so your argument is null.
You're arguing for the sake of arguing and not actually trying to fix or improve anything.
You mean use the only other record in use for a tie breaker and you win? (points for) All those other team statistics you mention would never be used in any tiebreaker ever known.
Hi, I like taking things out of context.
Also, you don't have to only use one statistic. PF-PA works, a combination of others may. I'm not saying it to just to go down the line and use the 9 statistics (whatever#) we are beating them at to not use the one we are behind by 2 in. PA would never be used at any level of real football so your argument is null.
You're arguing for the sake of arguing and not actually trying to fix or improve anything.
-Phaytle-
offline
offline
Originally posted by briansimoneau
This suggestion is interesting but Phaytle is not going to like it. We beat Chicago early in the season 14-0 and then Chicago beat us 20-10 last game.
I'd actually rather lose because of this than PA. Not perfect, but it does sit better I think to have it come down to how well you did against the team you are tied with.
This suggestion is interesting but Phaytle is not going to like it. We beat Chicago early in the season 14-0 and then Chicago beat us 20-10 last game.
I'd actually rather lose because of this than PA. Not perfect, but it does sit better I think to have it come down to how well you did against the team you are tied with.
-Phaytle-
offline
offline
Originally posted by NiborRis
No, it isn't.
The teams that go to the playoffs are the teams that performed best in their division/league that season. Not which team is better, not which team is the best at the end of the season - which teams performed better along the way. The tiebreaker is there to help decide which team had a better season in that league.
Using ladder ranks instead of some statistics involved in the league's play that season is terrible because it has little to no relation to how a team performed in the league that season.
This is why the ladder rank is a poor choice of tiebreaker and should not be used.
-- this does not mean that PA is the best choice of tiebreaker, but ladder rank is horribly bad.
I think most would agree that for division title, the tie breakers should go H2H -> division record, but after that it gets murky.
PA was the GLB1 answer, because PF was a crapshoot anytime there were cpu/guts (score 300 or 500? basically a roll of the dice, and that dominated your points for total). So you couldn't use margin of victory either.
GLB2 for now just has the GLB1 defaults, which probably isn't the best solution anymore.
Margin of victory (PF-PA) seems okay and is probably better than PA then PF, assuming we don't end up with massive blowouts in a couple of seasons; if massive blowouts return (100+ points), then PF needs to just be thrown away. You could even lift from the NFL and put "total touchdowns" after that before going to coin flip. Don't use the "sum of your rank of points scored and points allowed", because summing rankings is dumb. Don't use "strength of victory", because that's pretty obnoxious to explain and calculate for the user just viewing the standings. Don't use the ladder because it uses things that aren't part of the season's play.
I think this is where we disagree. I view the playoffs as a battle of the best teams in the league for the league championship. Just because they played some teams outside of their league doesn't mean those wins don't count - they are against quality opponents. If you look at it like out of conference games then it makes even more sense.
No, it isn't.
The teams that go to the playoffs are the teams that performed best in their division/league that season. Not which team is better, not which team is the best at the end of the season - which teams performed better along the way. The tiebreaker is there to help decide which team had a better season in that league.
Using ladder ranks instead of some statistics involved in the league's play that season is terrible because it has little to no relation to how a team performed in the league that season.
This is why the ladder rank is a poor choice of tiebreaker and should not be used.
-- this does not mean that PA is the best choice of tiebreaker, but ladder rank is horribly bad.
I think most would agree that for division title, the tie breakers should go H2H -> division record, but after that it gets murky.
PA was the GLB1 answer, because PF was a crapshoot anytime there were cpu/guts (score 300 or 500? basically a roll of the dice, and that dominated your points for total). So you couldn't use margin of victory either.
GLB2 for now just has the GLB1 defaults, which probably isn't the best solution anymore.
Margin of victory (PF-PA) seems okay and is probably better than PA then PF, assuming we don't end up with massive blowouts in a couple of seasons; if massive blowouts return (100+ points), then PF needs to just be thrown away. You could even lift from the NFL and put "total touchdowns" after that before going to coin flip. Don't use the "sum of your rank of points scored and points allowed", because summing rankings is dumb. Don't use "strength of victory", because that's pretty obnoxious to explain and calculate for the user just viewing the standings. Don't use the ladder because it uses things that aren't part of the season's play.
I think this is where we disagree. I view the playoffs as a battle of the best teams in the league for the league championship. Just because they played some teams outside of their league doesn't mean those wins don't count - they are against quality opponents. If you look at it like out of conference games then it makes even more sense.
bhall43
offline
offline
Originally posted by briansimoneau
This suggestion is interesting but Phaytle is not going to like it. We beat Chicago early in the season 14-0 and then Chicago beat us 20-10 last game.
lol
This suggestion is interesting but Phaytle is not going to like it. We beat Chicago early in the season 14-0 and then Chicago beat us 20-10 last game.
lol
bhall43
offline
offline
Originally posted by -Phaytle-
Hi, I like taking things out of context.
Also, you don't have to only use one statistic. PF-PA works, a combination of others may. I'm not saying it to just to go down the line and use the 9 statistics (whatever#) we are beating them at to not use the one we are behind by 2 in. PA would never be used at any level of real football so your argument is null.
You're arguing for the sake of arguing and not actually trying to fix or improve anything.
I am not trying to argue "for the sake of arguing". Points for is a terrible logic in games like this because of blowouts and it gives better credit to teams that are just fun and gun rather than teams that move the chains. I like the Head to Head PF-PA because at least there is higher logic to it. But using all the league games is just kinda silly in a game like this.
Hi, I like taking things out of context.
Also, you don't have to only use one statistic. PF-PA works, a combination of others may. I'm not saying it to just to go down the line and use the 9 statistics (whatever#) we are beating them at to not use the one we are behind by 2 in. PA would never be used at any level of real football so your argument is null.
You're arguing for the sake of arguing and not actually trying to fix or improve anything.
I am not trying to argue "for the sake of arguing". Points for is a terrible logic in games like this because of blowouts and it gives better credit to teams that are just fun and gun rather than teams that move the chains. I like the Head to Head PF-PA because at least there is higher logic to it. But using all the league games is just kinda silly in a game like this.
bhall43
offline
offline
Originally posted by -Phaytle-
I think this is where we disagree. I view the playoffs as a battle of the best teams in the league for the league championship. Just because they played some teams outside of their league doesn't mean those wins don't count - they are against quality opponents. If you look at it like out of conference games then it makes even more sense.
Team A goes 8-3 in ladder games. Plays teams anywhere from 30 to 200.
Team B goes 7-4 in ladder games. Plays teams anywhere from 1 to 30.
Which team goes in?
I think this is where we disagree. I view the playoffs as a battle of the best teams in the league for the league championship. Just because they played some teams outside of their league doesn't mean those wins don't count - they are against quality opponents. If you look at it like out of conference games then it makes even more sense.
Team A goes 8-3 in ladder games. Plays teams anywhere from 30 to 200.
Team B goes 7-4 in ladder games. Plays teams anywhere from 1 to 30.
Which team goes in?
The whole ladder thing would be insanity - let's just drop it!
As a case in point there is a thread right now somewhere where a ladder match up is off by 100+ ranks - and that is not a bug - it is just the way it came out this time...
Ladder games can have no part in division or league ties - for that reason and a DOZEN other good reasons that have already been mentioned
As a case in point there is a thread right now somewhere where a ladder match up is off by 100+ ranks - and that is not a bug - it is just the way it came out this time...
Ladder games can have no part in division or league ties - for that reason and a DOZEN other good reasons that have already been mentioned
Time Trial
offline
offline
Originally posted by TxSteve
The whole ladder thing would be insanity - let's just drop it!
As a case in point there is a thread right now somewhere where a ladder match up is off by 100+ ranks - and that is not a bug - it is just the way it came out this time...
Ladder games can have no part in division or league ties - for that reason and a DOZEN other good reasons that have already been mentioned
Since it is VERY easy to force all games to be within 10 ladder ranks, the fact that the code allows games from that far away is a bug imo.
X = A Random number between 1 and 10.
Y = The team's current rank
#1 rank plays Y+X
If #2 is not scheduled then plays Y+X
If #3 is not scheduled then plays Y+X
...
Where X = 1, Each odd numbered team on the ladder would play the even numbered team below them (1 v 2, 3 v 4, etc.)
Where X = 10, (1 v. 11, 10 v. 20, 21 v. 31, 41 v. 51, etc.)
Each game in the ladder would be of "equal" difficulty (approximately), so the results would have less basis in the randomness of the sim sometimes scheduling games that are 40 or 100 ranks apart.
The whole ladder thing would be insanity - let's just drop it!
As a case in point there is a thread right now somewhere where a ladder match up is off by 100+ ranks - and that is not a bug - it is just the way it came out this time...
Ladder games can have no part in division or league ties - for that reason and a DOZEN other good reasons that have already been mentioned
Since it is VERY easy to force all games to be within 10 ladder ranks, the fact that the code allows games from that far away is a bug imo.
X = A Random number between 1 and 10.
Y = The team's current rank
#1 rank plays Y+X
If #2 is not scheduled then plays Y+X
If #3 is not scheduled then plays Y+X
...
Where X = 1, Each odd numbered team on the ladder would play the even numbered team below them (1 v 2, 3 v 4, etc.)
Where X = 10, (1 v. 11, 10 v. 20, 21 v. 31, 41 v. 51, etc.)
Each game in the ladder would be of "equal" difficulty (approximately), so the results would have less basis in the randomness of the sim sometimes scheduling games that are 40 or 100 ranks apart.
bhall43
offline
offline
Regardless there are still a ton of issues with using Ladder ranks for tiebreakers. Issues that should never even need to be touched when you can do so much with league.
Time Trial
offline
offline
Originally posted by bhall43
Regardless there are still a ton of issues with using Ladder ranks for tiebreakers. Issues that should never even need to be touched when you can do so much with league.
I agree. That being said, my post was quoting the issues in the ladder, not that the ladder should be used as a tie-breaker.
Regardless there are still a ton of issues with using Ladder ranks for tiebreakers. Issues that should never even need to be touched when you can do so much with league.
I agree. That being said, my post was quoting the issues in the ladder, not that the ladder should be used as a tie-breaker.
-Phaytle-
offline
offline
Originally posted by bhall43
Team A goes 8-3 in ladder games. Plays teams anywhere from 30 to 200.
Team B goes 7-4 in ladder games. Plays teams anywhere from 1 to 30.
Which team goes in?
The team with the higher rank in my opinion because it can be looked at as strength of schedule. There is a reason the better teams go up the ladder and vice versa. Some near the top have a loss or two more than the middle, but they are beating better opponents.
I guess I see it as the playoffs should be the best 4 teams. The ladder is an ongoing system of which teams are better, so it will naturally send the better team if used in a tiebreaker. But it seems like people don't think the tiebreaker should send the better of two tied teams. Might as well just go to the coin flip instead of PA then.
Team A goes 8-3 in ladder games. Plays teams anywhere from 30 to 200.
Team B goes 7-4 in ladder games. Plays teams anywhere from 1 to 30.
Which team goes in?
The team with the higher rank in my opinion because it can be looked at as strength of schedule. There is a reason the better teams go up the ladder and vice versa. Some near the top have a loss or two more than the middle, but they are beating better opponents.
I guess I see it as the playoffs should be the best 4 teams. The ladder is an ongoing system of which teams are better, so it will naturally send the better team if used in a tiebreaker. But it seems like people don't think the tiebreaker should send the better of two tied teams. Might as well just go to the coin flip instead of PA then.
-Phaytle-
offline
offline
Originally posted by Stobie
Thread moved by moderator.
Moved due to the fact that it is not a bug as the process works as described. This is a suggestion in wanting to change how its done.
Thanks.
Thread moved by moderator.
Moved due to the fact that it is not a bug as the process works as described. This is a suggestion in wanting to change how its done.
Thanks.
Ladder doesn't work.
My team is (I think) the highest ladder rank in my league. We are 7-1 in ladder games and ranked at 74.
If I dump my team this offseason and it gets a new owner - they will still be in the 70's -- they might go 4-4 in ladder games next year and STILL be ahead of a team that goes 7-1 in ladder games next season...
If the ladder started fresh every single season (impossible) - ladder rank might work for a tie breaker...but there is no scenario under which ladder rank can or should mean something for a league or division tie...
I don't know why I'm even trying here as people more eloquent than I am have already been ignored by you!
Honestly - I'm starting to wonder if you're just trolling..."But it seems like people don't think the tiebreaker should send the better of two tied teams" Ok - I'm convinced.
My team is (I think) the highest ladder rank in my league. We are 7-1 in ladder games and ranked at 74.
If I dump my team this offseason and it gets a new owner - they will still be in the 70's -- they might go 4-4 in ladder games next year and STILL be ahead of a team that goes 7-1 in ladder games next season...
If the ladder started fresh every single season (impossible) - ladder rank might work for a tie breaker...but there is no scenario under which ladder rank can or should mean something for a league or division tie...
I don't know why I'm even trying here as people more eloquent than I am have already been ignored by you!
Honestly - I'm starting to wonder if you're just trolling..."But it seems like people don't think the tiebreaker should send the better of two tied teams" Ok - I'm convinced.
You are not logged in. Please log in if you want to post a reply.
