In post 146, Almost50 wrote:(Town would have backed off once they realized why I put Umlaut on the list by mistake)
In post 147, MURDERCAT wrote:I'm not backing off because I think choosing the first three people who posted is a really easy way to fake reads
Where in here is it implied that's why I originally voted you?
I voted you because your first statement sounded like you were trying to make fake content. If you weren't I think you were more likely to explain the reasoning for your claim. The fact that you didn't bother to check if umlaut had posted anything while trying make it seem like your reads were based on content when referring to nom made it worse.
I suggest everybody check their last 5 games and see if Scum posted in the the 1st 3. If you get 4 games out of 5 you should probably vote MC here (or at least look closely at them 1st 3 posters(. If you get less than 3 games out of 5 you should ignore me. Exactly 3? Up to you. Just check your own games for funzies
I may have done it in some, yes, but I used the rule more often than I had spoken of it.
But what does it matter? Let's say -hypothetically- that this is the 1st time I speak of it, or even apply it. How does that change the fact that in 10 out 13 games the rule did apply?
In post 160, MURDERCAT wrote:Also I'm not saying your rule doesn't work, I'm saying it works with any 3 players and isn't useful for that reason
You do realize that a mean of "a whole one" is a lot more than the mean on "any random 3 players" (which is 0.692). Right?
And you do realize that the Stantard Deviation here is 1.209 which means the result tends to be close enough to the mean/expected value. Yes? (I do realize your "any random 3" has a SD of 0.961 which is even better, but considering your expected value isn't "1" that's still less than what my rule yields).
I mean, we're in a a Battle for Calculasia, so we're bound to use our mathematical/statistical weapons after all, and I'm applying simple statistics and probability here.
Also IRRELEVANT: Remind me when your Open Draft Mafia is open for sign-ups of you will. I think I replaced into one earlier, but I want to be in the whole thing from the start
There are 135 combinations (of 3) that contain exactly 1 Scum, 30 that contain exactly 2 Scums, and one that contains 3, out of of a total of 286 combinations in a 13-players setup.
That is to say there are 166/286 combinations that include at least one scum in MC's "random" theory. That's like 58% of all combinations.
My proposition gives 10 out of 13 cases which is about 76.9%
My "mean" or "average" would net one scum per set of 3 (first posters), while MC's would net less than 0.7 per set. So, hypothetically if we applied my method to 100 cases and MC's to 100 cases I would get 100 Scums and he will get only 70
The SD is smaller for his proposition, so his results will revolves around the mean (0.7) with less deviation than mine would around a whole one, but given that my mean is higher my proposition still gives better results over time (because mine revolves around one.. give or take, and his revolves around 0.7 give or take, and with a SD of 0.96 is unlikely to ever get much higher).
But if you don't want to work your brain at all, just look at my games to get the gist of it. There's a 76.9% chance there's scum in the 1st 3 posters (or if you choose to trust me it's closer to 80% over a bigger sample).
