Page 1 of 4
Another "conspiracy" question
Posted:
Tue Mar 22, 2022 1:30 pm
by egvrich
Does anyone think/believe that the distribution in divisions is totally random?
Because it seems to me that managers with higher winning percentages seem to get lumped into the same divisions more often than not.
Maybe I'm wrong, but that could be one quick and
easy way for Strat to make it easier for newbies to make the playoffs.
Example: A league just drafted an hour ago, the four highest winning % managers are all in the same division.
https://365.strat-o-matic.com/league/459540It could be bad luck or
it could be something else.Feel free to express your thoughts? Or just call me a tin foil hat wearing quack.
Re: Another "conspiracy" question
Posted:
Tue Mar 22, 2022 1:35 pm
by Salty
egvrich wrote:Does anyone think that the distribution in divisions is totally random? Because it seems to me that managers with higher winning percentages seem to get lumped into the same divisions more often than not. Maybe I'm wrong, but that could be one quick and easy way for Strat to make it easier for newbies to make the playoffs.
Example: A league just drafted an hour ago, the four highest winning % managers are all in the same division. It could be blind luck or it could be something else.
Feel free to express your thoughts? Or just call me a tin foil hat wearing quack.
Haven't payed as much attention to this-
given all of the other things they don't disclose its definitely possible.
Re: Another "conspiracy" question
Posted:
Tue Mar 22, 2022 1:36 pm
by Palmtana
You could look at more leagues and see if the pattern holds.
Re: Another "conspiracy" question
Posted:
Tue Mar 22, 2022 2:13 pm
by MochaDog
I looked at a number of past autoleagues I've been in and found no such pattern - sure, in a couple of them four of the higher win percentage coaches might be in one division, but more often it appears totally random . . . and this kind of variability of grouping is exactly what's to be expected of a random distribution system.
Re: Another "conspiracy" question
Posted:
Tue Mar 22, 2022 3:00 pm
by MtheB
In order to establish the validity of that, a wide sample of data is needed, from owners of different skill levels so the info as to league makeup is not just coming from top owners that notice that.
that being said, I have noticed that phenomena upon occasion (one division having a predominance of the best owners in the league in one division) and I have wondered about it. and yes, that will occur with a true random sample, but that is why a wide sample of data is needed.
I will add it to my data accumulation to go along with our on-going data collection of run differential effects.
Re: Another "conspiracy" question
Posted:
Tue Mar 22, 2022 4:49 pm
by goffchile
Probably worth a look but I have not noticed it. With some of these things, there is the possibility of heuristic bias--I only notice my division mates when they seem particularly tough therefore it seems like it happens more than it does.
It would seem like it would be an unnecessary step (more work) than really necessary for SOM if it was true and it certainly would have no guarantee of working--giving newbies more championships.
Re: Another "conspiracy" question
Posted:
Tue Mar 22, 2022 5:25 pm
by mighty moose
I've created hundreds of Barnstormers leagues and have never observed anything suspicious. The sheer "randomness" of RANDOM means that once in a while a division may be overly strong, but never so many times that it raises eyebrows.
I would lean towards shooting down this one. IMHO.
Re: Another "conspiracy" question
Posted:
Tue Mar 22, 2022 5:55 pm
by MtheB
goffchile
thank you, that's a very important point.
I just called a very good friend of mine who is a math professor, and proposed the probability question to him.
here is his response:
The chance of the three best owners, randomly placed in a league, ending up in the same division is 5.45%, so that is basically a 1 in 20 chance. So it is not common for it to happen, but it can happen (approx once in 20 leagues)
He explained that in statistic analysis there is what is called The P value, which is the point at which if a result is above the P value, there is high confidence that it is a random result. If it is below the P value, then the result is suspect. 5 % is considered established P value, so a 5.45% is basically borderline, but still above the P value.
So unless this happens as a regular recurring pattern, then it most likely is a random result.
Re: Another "conspiracy" question
Posted:
Tue Mar 22, 2022 7:00 pm
by goffchile
MtheB wrote:The chance of the three best teams, randomly placed in a league, ending up in the same division is 5.45%, so that is basically a 1 in 20 chance.
As a person who plays Dungeons and Dragons as well as Strat--1 in 20
SEEMS to come up a lot more than you think.
Re: Another "conspiracy" question
Posted:
Tue Mar 22, 2022 7:41 pm
by MtheB
goffchile--
you just got my daughter, who is an avid D&D girl, to literally laugh out loud.....