'Random' moves

Yup. 1 out of 3 people wearing a jacket is still 33.33%. And I can honestly say that 33.33% of people sampled were wearing a jacket. It may not mean anything at that small a sample, but it is still a statistic.

And, by the way, a few hundred is more than enough for most research companies to publish results. In my industry its 75-125. My wife is a microbiologist- they work with cohorts of 10-20 mice and derive publishable stats.

Different institutions have different standards, itā€™s why we sadly canā€™t trust everything published to be accurate. Of course, people can record huge volumes of data then analyse it terribly and sometimes from small sample sizes you can gain some incredibly insightful results that may not be easy to identify with larger samples.

Stillā€¦ hereā€™s some juicy analysis on entrance speed for any of you number geeks like me:


2 Likes

So is the first sheet the entrance time for different speeds?

Yeah, those are some data points I was gathering to help work out the random variance (the stuff in brackets on the right). All are monsters with +0 from potions. I had multiple records for many of those but I wasnā€™t noting down frequency (couldnā€™t be bothered), you can see on some of them I was able to create a range from multiple records.

Btw if any of you feel up to itā€¦ Iā€™d like to hear if you can identify anything from looking at that first image.

Nothing to do with inverse porportion is it?

Time to crack out the graph paper methinks

Ya you guys go do the math, Iā€™ll wait here and finish these beers. Then you tell me what you find out :rofl:

1 Like

Itā€™s a trick question of sortsā€¦ the best answer is thereā€™s not enough data to analyse properly :wink:

But yeah inverse proportion is a good idea. The thing isā€¦ because the final number will be rounded, this can appear as larger variance when the numbers are small. For example, I think I estimated that it doesnā€™t vary more than +2.5% yet it appears to be +3.09% in one of the records because itā€™s been rounded up to 60. So thereā€™s nothing pointing directly to inverse proportion in this data, but itā€™s probably the most likely. I decided to look at other possibilities first.

What I theorised instead is that thereā€™s a Ā±X component as well as a percentage component. I could get something which fitted all the data points but when I included reduction via potions it didnā€™t match up, which is why I need to gather some more data. Itā€™s probably an inverse proportion anyway and maybe it interacts with potions in some way.

One really interesting thing the data shows is how almost everything is +X%, with only a few -X%. This is what made me like the idea of it being Ā±0.6 + [0:2]%.

Anyway Iā€™ll let you guys and girls know when I get around to calculating this stuff :slight_smile: