Page 3 of 4

PostPosted: Sat Jun 11, 2005 12:53 am
by GotowarMissAgnes
In a 16 team 6x6 league

24.47856205


It might be better to calculate the coefficient of variation, since leagues of different sizes will have different stdev just because of the number of teams and cats. Divide the s.d by the mean point total. For my league it is

0.24

which means that the deviation is 24% of the mean score.

PostPosted: Sat Jun 11, 2005 1:16 am
by DCNats
SHOCKandAWE wrote:I am dumb. I got the stdev but hwo do I apply this to my league and what am I looking at? Basically what in the heck does it mean?



My league is 5x5 13 teams

= 13.59534234


It's just a way to describe the dispersion, or variance, in points among teams in your league (the st. dev. is the square root of the variance). The lower the st. dev. the higher the parity in your league, though comparing leagues of different sizes and/or with different numbers of cats is problematic as the poster above points out. Your league appears to be relatively competitive.

PostPosted: Sat Jun 11, 2005 1:42 am
by Half Massed
i'm confused a bit. i get some of it, except of course the important part. how do you find a team's points? people are posting points next to the teams in their league, how do you find those? and just to clarify, for excel in the range part of the stdev what do you put? is it the games back people in the league are or what?

PostPosted: Sat Jun 11, 2005 1:50 am
by DCNats
Half Massed wrote:i'm confused a bit. i get some of it, except of course the important part. how do you find a team's points? people are posting points next to the teams in their league, how do you find those? and just to clarify, for excel in the range part of the stdev what do you put? is it the games back people in the league are or what?


I think most (or all) who've posted are using roto leagues. You could calculate it for head to head using team winning percentages, but that would not be comparable at all to the st. devs. others are posting here.

PostPosted: Sat Jun 11, 2005 1:57 am
by Half Massed
that would make a lot more sense then. i saw a few for h2h so i had to ask.

PostPosted: Sat Jun 11, 2005 2:06 am
by BobbyRoberto
In my 12-team Yahoo Public League: 21.77

The top 2 teams (I'm one of them) are at 103 and 107, then it drops down to 72 for the 3rd-place team. Last place is at 30.5.

In a 10-team private TQSTATS league: 14.4

Top-4 teams: 84, 84, 81, 76.5
Next-5 teams: 64, 61.5, 59, 54.5, 51
Bottom team: 44.5

PostPosted: Sat Jun 11, 2005 2:09 am
by shortsavage
Half Massed wrote:that would make a lot more sense then. i saw a few for h2h so i had to ask.


It would be pretty tough to translate point system stats into the roto mold that we have established here. If you were to say look at total points, you would get something, but odds are you would have a really high standard deviation.

PostPosted: Sat Jun 11, 2005 2:14 am
by shortsavage
It is interesting to see the levels of competition we deal with, especially when people's most competitive leagues are public.

I wonder how close typical twelve team 5x5 "expert" leagues tend to be.

PostPosted: Wed May 31, 2006 4:58 pm
by reiser
I always though this was a neat topic, so I'm giving it a bump!
my $$ league this year is insanely competitive. Std. Dev of 9.58 for a 12 team.

PostPosted: Wed May 31, 2006 5:23 pm
by roninmedia
14.72

6x5 10 team league.

Tier 1 - 2 Teams, 1.77
Tier 2 - 3 Teams, 1.5
Tier 3 - 4 Teams, 4.37
Tier 4 (worst team) - 1 team, 0