|
1EyedJacks -> RE: How to Choose Leaders (7/13/2016 1:19:47 AM)
|
quote:
ORIGINAL: wwengr
quote:
ORIGINAL: Smeulders
No offense, but 5 replications seems a bit low to really draw conclusions from. Not that this result would really surprise me, I'm guessing whether or not the DD spot the sub is of a much greater importance.
You are both correct and incorrect. I won't go into the math, as it can be fairly tedious for all but mathematicians, engineers, and statisticians. Five replications is enough to establish statistical significance. Significance is simply a measure that says that a difference is unlikely to have occured by chance. It says nothing about the relative importance of the result.
In the case of my test, the difference was small and the test of significance says that it was likely a random result. Based on that, If I ran many more test runs, I would expect the results for both sets to converge. That would be the most probable result.
of course, one of the most unsatisfactory things about real life (for non-engineers and such) is that you can't reach absolute certainty, but you can get close enough for all practical purposes. Five test runs leaves a relatively large chance that the test just failed to produce the difference.
I won't go into how I set up my hypothesis, but I'll provide the data set for anyone who wants. The engagements were day time and I set crew experience high and high constant leader chaaracteristics for the ship captains.
TF Commander agression = 30 (2,1,1,0,0); TF Commander agression = 70 (1,1,1,0,0). The numbers are how many destroyers enagaged the sub in each test run.
If I were to look at it and not understand, I would actually conclude that there is an inverse relationship between aggression and the probability of an engagement. That would be wrong.
If you want to improve the test, then add more ships and run more test runs. perhaps test agression = 1 against agression = 99. The test definately could use some improvement. I am not a big experimentalist. To me, that is tedious. Give me the data and I will play...
I think that it's DL that affects the attack but aggression that increases the odds of a successful attack. The Aggressive commander will take his sub to the surface and use the guns. The Aggressive commander will fire a full spread or try to get in multiple attacks. Or at least that's my 2 cents... (smile).
|
|
|
|
