User Pass
Home Sign Up Contact Log In
Forum > USA AA Leagues > USA AA #2 > Computer Rankings After Game #5
Page:
 
Craelath
offline
Link
 
Originally posted by @ftermath
Originally posted by Craelath

Originally posted by krwynn


Originally posted by Ryiotgear




I enjoy the rankings and think it is kind of pointless for everyone to worry too much about them being perfect - he laid out the system he is using and that is what the results will show.

I think to say SOS doesn't matter is incorrect though. The rankings are based upon descriptive statistics of games played to date. The rankings can therefore only describe what has happened to this point. At this time we have not played a round-robin schedule, so the stats and rankings are not based upon a round robin format.

Besides the 1 cross-conference matchup we will eventually come very close to a round-robin schedule, but right now the schedule is pretty far from that so we can't rely on the round robin format making SOS negligible until we have truly played out the format. Once week 16 rolls around I think it would be accurate to say SOS has very little bearing on your rankings.


How about this, when someone can provide an SOS, a true SOS, which would be based on LAST season...nevermind. It's irrelevant. SOS just plainly, simply, does not factor in. Does anyone really think the monkeys will be easier to beat because they've lost a game? And you have other factors, boosting, player's coming on a team, players leaving. To even think you could arrive at an SOS is preposterous. Besides after today I'll more than likely keep them in house. I feel like I have to defend myself for posting stats. lol. I quit doing things like that for no $$ years ago.


That's a shame. I've been watching the couple you've posted and, like I said, it looks like it could be quite a valuable tool. It starts off rather sketchy, but that's due to limited data which you have no control over, but like I said, it gets infinitely more accurate when more data is accumulated.



Well if he added a human voting portion to offset the skewed data and portions of the game that a computer can't simulate he may have something. Wait, someone already does that. Ah yes it's the BCS and it doesn't work.

Anyway, it's one guy putting some time into something that can be fun if you enjoy this sort of thing and ignored if you don't. It's a bit like a certain appreciation thread that I refuse to even look at because I think it's ridiculous. But that's my opinion so I just stay away.


Except, unlike the BCS, this has absolutely no bearing on anything. At most, this allows you to temper your expectations for a given year, not decide if you're in the championship or not. And, also unlike the BCS, everyone plays the same group of teams. That alone makes this a pretty good judge of where you are, statistically, at any point in the season.
 
@ftermath
offline
Link
 
Originally posted by Craelath
Originally posted by @ftermath

Originally posted by Craelath


Originally posted by krwynn



Originally posted by Ryiotgear





I enjoy the rankings and think it is kind of pointless for everyone to worry too much about them being perfect - he laid out the system he is using and that is what the results will show.

I think to say SOS doesn't matter is incorrect though. The rankings are based upon descriptive statistics of games played to date. The rankings can therefore only describe what has happened to this point. At this time we have not played a round-robin schedule, so the stats and rankings are not based upon a round robin format.

Besides the 1 cross-conference matchup we will eventually come very close to a round-robin schedule, but right now the schedule is pretty far from that so we can't rely on the round robin format making SOS negligible until we have truly played out the format. Once week 16 rolls around I think it would be accurate to say SOS has very little bearing on your rankings.


How about this, when someone can provide an SOS, a true SOS, which would be based on LAST season...nevermind. It's irrelevant. SOS just plainly, simply, does not factor in. Does anyone really think the monkeys will be easier to beat because they've lost a game? And you have other factors, boosting, player's coming on a team, players leaving. To even think you could arrive at an SOS is preposterous. Besides after today I'll more than likely keep them in house. I feel like I have to defend myself for posting stats. lol. I quit doing things like that for no $$ years ago.


That's a shame. I've been watching the couple you've posted and, like I said, it looks like it could be quite a valuable tool. It starts off rather sketchy, but that's due to limited data which you have no control over, but like I said, it gets infinitely more accurate when more data is accumulated.



Well if he added a human voting portion to offset the skewed data and portions of the game that a computer can't simulate he may have something. Wait, someone already does that. Ah yes it's the BCS and it doesn't work.

Anyway, it's one guy putting some time into something that can be fun if you enjoy this sort of thing and ignored if you don't. It's a bit like a certain appreciation thread that I refuse to even look at because I think it's ridiculous. But that's my opinion so I just stay away.


Except, unlike the BCS, this has absolutely no bearing on anything. At most, this allows you to temper your expectations for a given year, not decide if you're in the championship or not. And, also unlike the BCS, everyone plays the same group of teams. That alone makes this a pretty good judge of where you are, statistically, at any point in the season.



Very true.
 
Jaxx
offline
Link
 
I enjoy looking at stuff like this and even though the monkey's and I do not see eye to eye, to take them for granted because of one loss would be a huge mistake. Austin and Crush are with out a doubt the top two teams in the west division with the Pack and Monkey's a close third. Can anyone of these teams be upset on a given day, yes and who knows when the playoff sims kick in seen some odd things happen.

Please keep up if you have the time with posting this stuff I believe it adds to the game instead of taking anything away.
 
coachviking
offline
Link
 
Originally posted by Jaxx
I enjoy looking at stuff like this and even though the monkey's and I do not see eye to eye, to take them for granted because of one loss would be a huge mistake. Austin and Crush are with out a doubt the top two teams in the west division with the Pack and Monkey's a close third. Can anyone of these teams be upset on a given day, yes and who knows when the playoff sims kick in seen some odd things happen.

Please keep up if you have the time with posting this stuff I believe it adds to the game instead of taking anything away.



Krwynn's numbers are just another way of looking at the strength of the teams in this league. Anything that is well thought out and opens up this much debate is a good thing for the game and this league. IMO
 
RunningMn9
offline
Link
 
Originally posted by krwynn
How about this, when someone can provide an SOS, a true SOS, which would be based on LAST season


You don't need a TRUE SOS based on last season, since this game is unlike reality in that teams can change so much in an offseason. We are 5 games into a season. Let's say that I've played the 5 worst teams in the East, and you've played the 5 best teams in the West.

Explain to me (in terms other than you shouting for me to provide a source) why strength of schedule shouldn't temper any analysis of the statistics that we've both achieved. When I was calculating power ratings last season, my formula to calculate strength of schedule was simple, take the combined w/l record of my opponents and subtract my own w/l record from that (to remove my games from their records). Voila!

You now have a value that you can use to attenuate the statistical performance of each team. Teams that run up stats against an 0-5 team get less credit than teams that pound a 4-1 squad.

It's not complicated. You just have to stop being so goddamn stubborn about everything and listen to other people for once. After Week 15, when all teams in each conference have played a Round Robin schedule, strength of opponents is irrelevant. After Week 5, when no one has played a Round Robin schedule, strength of opponents is absolutely relevant to a statistical comparison. My statistics were achieved against a different set of teams than the Doomsday Sin's statistics.

We only have one common opponent. The Sin's opponents are a combined 14-6 in their non-Sin games. Our opponents are a combined 11-9 in their non-PAIN games. How could that NOT be relevant to the statistics that each of us generated (it makes the Sin's statistics much more impressive).


Last edited Jun 23, 2008 09:00:48
 
The Meal
offline
Link
 
Originally posted by Craelath
(U)nlike the BCS, everyone plays the same group of teams. That alone makes this a pretty good judge of where you are, statistically, at any point in the season.


No, it makes a good judge of where you are after 15 games in the season (and likely fewer, as the quality of opponents you have faced is reduced as a factor as you play more games -- at the other end of the spectrum, the quality of opponents you have played is maximized when you've only played one game so far in the season).

Last edited Jun 23, 2008 10:13:25
 
HootyHoo
offline
Link
 
Originally posted by RunningMn9
Originally posted by krwynn

How about this, when someone can provide an SOS, a true SOS, which would be based on LAST season


You don't need a TRUE SOS based on last season, since this game is unlike reality in that teams can change so much in an offseason. We are 5 games into a season. Let's say that I've played the 5 worst teams in the East, and you've played the 5 best teams in the West.

Explain to me (in terms other than you shouting for me to provide a source) why strength of schedule shouldn't temper any analysis of the statistics that we've both achieved. When I was calculating power ratings last season, my formula to calculate strength of schedule was simple, take the combined w/l record of my opponents and subtract my own w/l record from that (to remove my games from their records). Voila!

You now have a value that you can use to attenuate the statistical performance of each team. Teams that run up stats against an 0-5 team get less credit than teams that pound a 4-1 squad.

It's not complicated. You just have to stop being so goddamn stubborn about everything and listen to other people for once. After Week 15, when all teams in each conference have played a Round Robin schedule, strength of opponents is irrelevant. After Week 5, when no one has played a Round Robin schedule, strength of opponents is absolutely relevant to a statistical comparison. My statistics were achieved against a different set of teams than the Doomsday Sin's statistics.

We only have one common opponent. The Sin's opponents are a combined 14-6 in their non-Sin games. Our opponents are a combined 11-9 in their non-PAIN games. How could that NOT be relevant to the statistics that each of us generated (it makes the Sin's statistics much more impressive).




You're outnumbered. It's meaningless. So what, you pound an 0-4 team. New ownership comes in changes players, boosts, then what. It's nuts to even think such a thing. One thing I do agree on is that its not "real", so things you would do with "real" teams, such as an SOS might be relevant. However this isn't the case. To many things can change not only week to week but day to day that make an SOS ridicluous to even consider. Starting to wonder if you actually know the reasoning or why an SOS was developed in the first place. You're clearly wrong about this and to discuss it any farther is pointless. Good luck.
Last edited Jun 23, 2008 10:34:27
 
RunningMn9
offline
Link
 
Originally posted by krwynn
Starting to wonder if you actually know the reasoning or why an SOS was developed in the first place.


Are you really? Is it time to link me to your spreadsheet again yet?

There are plenty of problems with evaluating SOS in GLB as the season progresses. It being hard to calculate isn't the same as your earlier assertion that SOS is irrelevant. It's clearly relevant (to everyone but you apparently). Calculating it accurately may be more difficult (although the system that I employed was simple and would be accurate enough for most cases).
 
HootyHoo
offline
Link
 
Originally posted by RunningMn9
Originally posted by krwynn

Starting to wonder if you actually know the reasoning or why an SOS was developed in the first place.


Are you really? Is it time to link me to your spreadsheet again yet?

There are plenty of problems with evaluating SOS in GLB as the season progresses. It being hard to calculate isn't the same as your earlier assertion that SOS is irrelevant. It's clearly relevant (to everyone but you apparently). Calculating it accurately may be more difficult (although the system that I employed was simple and would be accurate enough for most cases).


She's all yours kid. I think we'll all be expecting your first posting after todays game. Have fun.
Last edited Jun 23, 2008 11:16:35
 
Drago
offline
Link
 
Thanks for posting the stats Krwynn.
 
RunningMn9
offline
Link
 
Originally posted by krwynn
She's all yours kid. I think we'll all be expecting your first posting after todays game. Have fun.


No thanks gramps. I'll be sure to stay off your lawn in the future, don't fret.
 
HootyHoo
offline
Link
 
I'm pretty sure 100% of the people here knew you'd rather bitch and whine then to actually have the balls to post anything. You know those who can do, those who can't bitch.
 
RunningMn9
offline
Link
 
Originally posted by krwynn
I'm pretty sure 100% of the people here knew you'd rather bitch and whine then to actually have the balls to post anything. You know those who can do, those who can't bitch.


I would guess that you're correct, since even I know that I wouldn't waste my time coming up with arbitrary statistical rankings that half the people would complain about. Look at the grief it gets you.

Of course, that doesn't make your claims regarding SOS true, it just means that you are free to disregard them in your rankings for whatever reason you'd like (including wrong reasons).
 
Link
 
Originally posted by krwynn

I don't think you're getting the idea behind the rankings. Its only showing who is doing better statistically.


Ummm... We (the Fancy Boys) are doing better statistically than the Monkeys. Statistically, we are undefeated, and they have one loss. Statistically, we scored more points than they did when we played them. Statistically, We beat them. Statistically, we are better than them.

Originally posted by krwynn

You're reading way to much into it. If you really want to get down to it, it shows that the Monkeys played a bad game, had turnovers, or maybe bad tactics. As per their stats they played below what they're capable of. It also doesn't take into affect that the pack boosted before that game. When you play more games as a boosted team I'm sure you'll bounce back up. Maybe.


The Monkeys played a bad game? Yeah, they played a bad game because we have a great defense, and we outplayed them. We are the better team, physically and statistically. And what does the Pack have to do w/ us?
Last edited Jun 23, 2008 13:11:48
 
HootyHoo
offline
Link
 
Originally posted by RunningMn9
Originally posted by krwynn

I'm pretty sure 100% of the people here knew you'd rather bitch and whine then to actually have the balls to post anything. You know those who can do, those who can't bitch.


I would guess that you're correct, since even I know that I wouldn't waste my time coming up with arbitrary statistical rankings that half the people would complain about. Look at the grief it gets you.

Of course, that doesn't make your claims regarding SOS true, it just means that you are free to disregard them in your rankings for whatever reason you'd like (including wrong reasons).


yawn, you're the only one complaining. 23 thank you's today. And like I said. NO one thought you would post anything. you're a back row whiner. You know it, I know it, now everyone else knows it.
 
Page:
 


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