Re: Galaxy help needed

From: amead@s.psych.uiuc.edu (Alan Mead)

Date: Sat, 09 Jan 1993 00:12:50 +0000

My analysis, was, I think, in error:

>enemy fire: let's say that you have one cruiser, the enemy has 30
>escorts.  Your cruiser automatically kills any escort it hits.  The
>escorts have a 1% chance of killing your cruiser.  What is the chance
>that your cruiser survives?  That's really really hard to say.  A
>simulation is the best way to answer that question.  But consider that
>over an infinite repitition, you would expect the escorts to fire 30
                            ^[of battles]
>times to the cruisers once.   Then 29 times to the cruiser's once.

15.5 times to the cruiser's once, then 15:1, then 14.5:1, ...

>then 28 times to once.  etc.  If each escort has a 1% chance to kill,
>it has a 99% chance not to.  The chance of NOT killing on n shots is
>(.99)^n.  On the above three "rounds", the escorts take 30+29+28 shots
>and have a 0.417 chance of NOT killing and a .582 chance of killing the
>cruiser.  *AND* combat continues for 27! more shots.
                                      ^^^ this is clearly an error

So the expected (the mean for an infinite sample) number of shots the
escorts get is 15.5+15+14.5+14+... which is, I think, about 225.
(.99)^225 = .104, so about 10.4% of the time, the cruiser is expected
to survive the encounter.  I have no idea how to calculate the expected
number of casulties amongst the escorts.  (Except by simulation).

I *think* this is right.  The point is that combat is, for me, not
entirely intuitive and a numerical advatage can be wicked.

Also, I don't know how to solve for the ratio of shield to attack that
would produce a P(kill)=0.01, but an attack of 1.0 is pitifully small
when you talk about cruisers of mass 30 or larger, so this particular
example is a bit contrived.

-alan

References

Re: Galaxy help needed
amead@s.psych.uiuc.edu (Alan Mead)
Fri, 08 Jan 1993 21:58:42 +0000
Galaxy help needed
ma_blase@csd.uwe.ac.uk (Fox)
Thu, 07 Jan 1993 19:23:37 +0000
Re: Galaxy help needed
M22367@mwvm.mitre.org
Fri, 08 Jan 1993 14:06:08 +0000
Re: Galaxy help needed
curt@oasys.dt.navy.mil (Curt Welch)
Fri, 08 Jan 1993 19:24:44 +0000

Referenced By

Re: Galaxy help needed
amead@s.psych.uiuc.edu (Alan Mead)
Fri, 08 Jan 1993 21:58:42 +0000
Galaxy help needed
ma_blase@csd.uwe.ac.uk (Fox)
Thu, 07 Jan 1993 19:23:37 +0000
Re: Galaxy help needed
M22367@mwvm.mitre.org
Fri, 08 Jan 1993 14:06:08 +0000
Re: Galaxy help needed
curt@oasys.dt.navy.mil (Curt Welch)
Fri, 08 Jan 1993 19:24:44 +0000