User Pass
Home Sign Up Contact Log In
jamz
offline
Link
 
Very few close scores.

Looks like it's gunna take 2-3 seasons for this cloud idea to shake out.
 
Duke 36
offline
Link
 
Does anyone reckon we'll get 2 or 3 seasons to let it level out?
 
Enkidu98
offline
Link
 
In my opinion it will take far more than that. I personally think Bort should have just added up everyones median and mean levels for their teams and their number of players. On top of these numbers add or subtract a value based on if you went to the playoffss and how far in you went or if you lost a lot of games, you lose points.

Then, order this list, and every 32 teams, start a new league.

He should also shake up the leagues 2-3 days BEFORE the new preseason and do a huffle where he pushes CPU teams and teams that gut after promotion/relegation down to the bottom again

Example:

Lose 15 or 16 Games = -10 points.
Lose 13 or 14 Games = -8 points
Lose 11 or 12 Games = -6 points
Lose 9 or 10 Games = - 4 points
Lose 7 or 8 games and NOT make the playoffs = -2 point.
Make playoffs but eliminated in first round +2 point.
Make playoffs but eliminated in second round +4 points
Make Playoffs but eliminated in 3rd round = +6 points
Make Playoffs but eliminated in 4th Round = +8 points
Make Playoffs and win it all +10 points.

So lets look at our team. San Dimas went 20-0, winning their league championship so thats good for 10 points.
Their Median level is 37
Their Mean level is 39.09
Thy have 55 players.

Sum of those numbers is 10+37+39.09+55 = 141.09

The team we played our first game is a quality team, just not in the same league as us and so they got screwed by Bort's promotion system that forced them up tot he uncapped minors where they could very well be facing teams even worse than ours. Thats no fun for their players/owner.

lets add them up....

They were eliminated in the third round so that is +6
Median Level is 31
mean is 32.49
Total Player Count: 55

So their score is: 124.49

A 20 point difference doesn't seem like a lot, but if you compared the other teams as well you can see how a very well built team like the Hippos are still going to be dominated by another well built team like us, who have more levels/higher averages etc. I think if the Hippos had been closer to our levels, this would have been a tight game/coin flip to determine who won.

We should be in a league playing with teams with similar scores. If you're going to shuffle them like he is, it shouldn't be based on level caps, but on 'scores' as this will more likely put closer teams together in a league. I know some teams are going to be bad no matter what the levels on their team are. But I think this is covered somewhat by taking away points from teams that are big losers. They may be dropped down to that they are aligned with teams who may have lower levels but better strategies/builds so that they are closer to even.


Even then, He should likely do promotions and demotions twice each off season. Once about 2 days after end of the season, then give another few days and then stop boosting for that season, after then, hand out the CPU/New teams etc and then a few more days after that, to give people chances to recruit, one more promotion/demotion period to take care of folks who gutted after the first round of promotions/demotions.

Since you would be using a fomula based on attributes that can be culled from the database fairly simply and then a simple ordering and take 32 teams, chop and form a league, take 32 teams chop and form a league, etc, the leagues would end up with much much more parity.

Also, to get more movement at the top I think he should maybe 4 demotions instead of just 2.
 
Mr Sinister
offline
Link
 
OK, by your standards, we went 5-11 last season, so that's a -6 points.
We also have 49 players, which is 49 points. (why this is added in, I don't know, cause we get penalized for not having 55 players)
Our median is 37.65 (if this is the average level of all your players) (also I don't know if you round up)
Our mean is ? I have no clue what that is

Could you finish this for me, and tell me what Athens should be at, by your math.

-6 +49 +37.65 + x = Y (so far we are only an 80.65)

http://goallineblitz.com/game/team.pl?team_id=6686

Wow, I had us at 5-10, it was 5-11.
Edited by Mr Sinister on Jun 15, 2009 08:52:38
 
Jayce Jones
offline
Link
 
Originally posted by Enkidu98
In my opinion it will take far more than that. I personally think Bort should have just added up everyones median and mean levels for their teams and their number of players. On top of these numbers add or subtract a value based on if you went to the playoffss and how far in you went or if you lost a lot of games, you lose points.

......
The team we played our first game is a quality team, just not in the same league as us and so they got screwed by Bort's promotion system that forced them up tot he uncapped minors where they could very well be facing teams even worse than ours. Thats no fun for their players/owner.

lets add them up....

They were eliminated in the third round so that is +6
Median Level is 31
mean is 32.49
Total Player Count: 55

So their score is: 124.49

A 20 point difference doesn't seem like a lot, but if you compared the other teams as well you can see how a very well built team like the Hippos are still going to be dominated by another well built team like us, who have more levels/higher averages etc. I think if the Hippos had been closer to our levels, this would have been a tight game/coin flip to determine who won.

We should be in a league playing with teams with similar scores.....


Thats a pretty good analysis of who the Hippos are (my team). Being a teacher I choose to carry my non-boosting students and that will hurt us until we can figure out our true LVL. We made a huge jump from 34 cap to uncapped, I would have preferred to stay at 34 or gone to 38 but thats water under the bridge now. Did I even have that option? I don't think so.

Anyway, I hope we are more competitive than we showed in the first preseason game.

 
Enkidu98
offline
Link
 
Originally posted by Mr Sinister
OK, by your standards, we went 5-11 last season, so that's a -6 points.
We also have 49 players, which is 49 points. (why this is added in, I don't know, cause we get penalized for not having 55 players)
Our median is 37.65 (if this is the average level of all your players) (also I don't know if you round up)
Our mean is ? I have no clue what that is

Could you finish this for me, and tell me what Athens should be at, by your math.

-6 +49 +37.65 + x = Y (so far we are only an 80.65)

http://goallineblitz.com/game/team.pl?team_id=6686

Wow, I had us at 5-10, it was 5-11.


Median = Half the players are over thie level and half are under it.
Mean = Average

The reason I add the number of players a team has is because it means they have additional depth. Depth means their players play at a higher level of efficiency for the game, generally, then do teams who have fewer players so they use more energy per play. You can counteract this with stamina, but this means other attributes may be lower... so it balances out to some extent.

The Playoff/Season record thing was just a way to make sure teams who may have say, all level 50's players, but who don;t have a good OC/DC etc or their builds are subpar get dropped a little bit so they are facing teams they may be more competitive with.

For your team the numbers would be:Median Level is 37; mean is 38.57

-6 + 49 + 37 + 38.57 = 118.57 Just based on that, my guess is that if you were to play the Pygmy Hippos you would lose, though the game would likely be close. Barring a streak of turnovers from either team which are somewhat unpredictable.
 
Enkidu98
offline
Link
 
Adding: The team you played, the Belfast Brawlers, has a rating of:

Median Level is 40.5; mean is 40.60
Total Player Count: 40

-6 for last seasons record

40.5 + 40.6 + 40 - 6 = 115.1

Since the Athens Nighthawks ranked at 118.57 I would have (and did) predict a win. (Though the formula I was using for that was more complex but used some of the same elements)

You can see though that since your ratings are close to one another, the game was closer then my own teams matchup with the Hippos where there was nearly a 17point difference.

Given equal skill level of the players/builders/coaches and absent a fumble/turnover streak the higher number will win most often. The larger the disparity in the number, the better the chance the blowout will be complete. CPU teams fare even worse but this is expected.

So the aim really should be to rank all the teams by a formula like this, then just cut the list every 32 team to make a league. Then the scores will be closer and the competition much more similar.
 
Mr Sinister
offline
Link
 
Now don't get me wrong, I like the work you've done, and what you say with it, but I also think it doesn't account for 3 factors, if people don't know them, then they should figure them out ( I think you know what they are Enkidu98, you hit on 2 of the 3). Keep this stuff up Enkidu98 [I'm calling you Ek from now on (Red Sox reference)], I am just always skeptical about other peoples math and breakdowns of things, andI'm sure you will figure that out for yourself from this post.

Our game was close based on 3 things I did as OC, which I will not point out. Anyone can look at the game film and figure these things out, maybe. Our next game will be close (I know it shouldn't, and we should crush them), but again because of what I am doing. Hell I can put it out there, as they are a CPU team. It will either be a blow out by us, or a really close game based on what I am trying to find out.

See it's all good in theory, is what I am trying to prove out. We get hurt because of our roster, and how many games we lost last season. Now comes the but, BUT if we had a full 55 man roster(+6), we would have been a 124.57, and if they had a full roster (+15), they would have been a 130.1, I still think they would have lost, if they kept to the means average (40.60). You just take into account the median and means numbers only, we are only an 75.57, and should have lost to an 80.60 team. So we ended up getting 9 extra points based on roster size.

I would like a game against the Hippos to prove this out one way or the other. They are higher rated than us 124.49 - 118.57, so based on that they should win. I think we will, I can wait till week 7.

I talk too much...
 
Enkidu98
offline
Link
 
No worries, you're certainly free to agree or disagree.

I just believe the method Bort has used to do promotions and demotions and shuffle leagues has been proven a dismal failure each time he has done it.

I believe there are better methods. Mines not perfect, I would prefer to use a more complete set of numbers but some of these numbers can be gamed by teams who are smart, so using them wouldn't always have closer to optimal results. So I narrowed my sets to numbers that can not be gamed as much. I know they can still be gamed somewhat, but not too much.

My aim is not to create a perfect league where its likely for every team to be at 8-8 and then they start needing to use tie breakers to determine playoffs seeding. Instead my aim is just to create leagues that should be more competitive. There will still be variation with the quality of the teams and their ability to game plan etc, but it should not be as disparate as the Western Conference ended up for instance.
 
Jayce Jones
offline
Link
 
And then there are teams like mine that carry poor depth (for various reasons) that throw things off. I do wish that I would have had the option to stay in the 34 cap (and let people go) or only move up to the 38.
 
Mr Sinister
offline
Link
 
Oh, don't get me wrong. I do agree with you 1000% that Borts way or doing things sucks, and it sucks bad. Now I would use half of what you do using only the means, and median to decide which teams move up, and which move down. Then I would also take teams who are say in a capped league, and move them up 2, but only as high as a capped league. If I had enough teams in say a 41 range (32 teams) I would make a capped league for that.

Also I need to show you that some teams are still messing around to try and get an easy trophy. This team should either be moved up, or have all players more than 3 levels above there cap removed. No way in hell can you tell me a capped 28 league should have a team with level 33 and up players on it.

http://goallineblitz.com/game/team.pl?team_id=7039

 


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