mvnForum Homepage

Print at Jun 2, 2020 3:44:56 AM

Posted by Gillie017 at May 9, 2009 9:01:42 PM
Re: Conference of the Monarch (Thoughts)
1. You used only one point in time. It is unlikely that this is representative of the activities of all pirates. You should have sampled multiple times over different times of the day and days of the week.

I may do that, but I don't think that saturday evening is that out of whack. You don't see a lot of people commenting on how unusual that is.

You "think" it is no different but you don't know for sure. I "think" it could be quite different but I don't know for sure. The multiple sampling I suggested would eliminate any potential bias from choosing just one point in time.

2. You sampled only pirates on ships. The stats of the pirates in your sample are therefore biased towards the type of voyages out at the time.

Again, that was deliberate because people were explicitly saying that you can't sample the whole population because that doesn't represent jobbers. So, I sampled jobbers. Again, if we want to know the entire population, well, we have that. Able is 50% of the whole population.

People were saying that you cannot simply sample all pirates who have been active over the past 10 days because this give equal weighting to greenies who log on and never come back, labour alts and addicts who are on all the time. This is the flaw with the numbers generated by Lizthegrey. Sampling from those online at a particular point in time overcomes those flaws. You sampled jobbers only and biased your results towards those on ships which are in no way representative of the "average" pirate.

3. You sampled from one archipelago only. The activities in different archipelagos can be vastly different at any given time.

On other oceans, yeah, but not so much for Viridian/Cobalt because the archs are very large (only three on the ocean) and laid out to be symmetrical.

The uniformity of the arches is irrelevant. You sampled people who were in the arch at the time, not those who call it their home. If a popular SMH chart had one or two frigs on the board in a different arch to where you did your sample then there were a lot of pirates who could not have been counted by you. Similarly if the arch you were in had a lot of greenie WB pillages out, you over-represented them in your sample.

4. You got bored and stopped collecting data half way through a frig. If this had been a SMH your results would have been very different had you completed the entire ship.

As I said earlier, I intended to sample about 100 pirates because you really don't need a much larger sample than that to get good results. It was well over 10% of the online population and big enough based on things like the law of large numbers and the central limit theorem. Yes, it surprises a lot of people that you can get accurate measurements from when you poll only a few thousand registered voters for a country the size of the USA, but it works.

It only works if the sampling frame is not biased.
The only time you considered potential bias, you were wrong. No bias can be introduced by checking ships in alphabetical order since ships are named randomly.

I was wrong? What? I didn't think there would be a bias based off the names, and you are agreeing with me, right? And as I have said a couple of times now, I considered arch bias when I chose Viridian. (That, and Gothmog wanted a green ocean.)

I said before the sampling scheme is extremely important. Your sampling failed in a number of important ways, this is why your results are meaningless.
If a bird looks like a duck, swims like a duck and quacks like a duck, then it's probably a Huntard .

Puzzle Pirates™ © 2001-2016 Grey Havens, LLC All Rights Reserved.   Terms · Privacy · Affiliates