PDA

View Full Version : My Brain Hurts!!!



jztemple
05 Jan 09, 14:17
I am writing an article about assaults and in there I tried to do some math about how I assure that my assault force will cause at least one casualty, and how if it's less than this perfect number the chance is some probability percentage. Trouble is, when I took this math Nixon was president and my mind seems to be unable to come up with a clear answer. I blame Nixon of course :D.

So, here it is: Using the description of Combat Results in the Users manual, and assuming a LCV of X and a HCV of Y, what combat value Z will give me a 90% chance of getting *at least* one kill? For simplicity you can assume that X is one-tenth of Y, since that's a truism in the database no matter whether you're firing, attack assaulting or defending from assault.

Oh, and you will need to show your work, teacher is an inquisitive type!

jztemple
06 Jan 09, 10:26
OK, so I wasn't able to beat the problem with an elegant math solution. So instead I hammered it with some computer number crunching, using a Visual Basic Script running in IE7. Here are the results for defender casualties in Assault:


LCV is 12 HCV is 120, which shows defender casualties in Assault, PASSES is 50000, which shows total number of tries for each CV

Numbers below are Combat Value, then Average Casualities, then percent chance that at least one casualty occurred, then ditto for two

40, 0.27, 26.62%, 0.00%
80, 0.51, 50.73%, 0.00%
120, 0.80, 71.18%, 8.48%
160, 1.06, 82.18%, 24.18%
200, 1.32, 87.72%, 39.86%
240, 1.59, 91.09%, 52.40%
280, 1.85, 93.81%, 60.82%
320, 2.11, 95.44%, 67.29%
360, 2.37, 96.37%, 72.26%
400, 2.64, 97.06%, 76.25%
440, 2.91, 97.87%, 80.27%
480, 3.15, 98.53%, 82.38%
520, 3.43, 98.81%, 84.74%
560, 3.71, 99.22%, 87.10%
600, 3.93, 99.49%, 88.02%
640, 4.25, 99.77%, 89.77%
680, 4.48, 99.77%, 91.30%
720, 4.75, 99.83%, 92.46%
760, 5.02, 99.97%, 93.16%
800, 5.30, 99.96%, 94.15%
840, 5.52, 100.00%, 94.76%
880, 5.82, 100.00%, 95.55%
920, 6.04, 100.00%, 96.33%
960, 6.35, 100.00%, 96.66%
1000, 6.61, 100.00%, 97.33%

Based on fifty thousand tries, which I figure hey, that should be enough, we see that to assure at least a 90% chance of getting a minimum of one casualty against a defender in assault combat, you need a combat value of about 240. That's 12 unarmed men, or about 10 with M1893 rifles, or about 8 with Garands.

TheBigRedOne
06 Jan 09, 13:47
OK, so I wasn't able to beat the problem with an elegant math solution. So instead I hammered it with some computer number crunching, using a Visual Basic Script running in IE7. Here are the results for defender casualties in Assault:



Based on fifty thousand tries, which I figure hey, that should be enough, we see that to assure at least a 90% chance of getting a minimum of one casualty against a defender in assault combat, you need a combat value of about 240. That's 12 unarmed men, or about 10 with M1893 rifles, or about 8 with Garands.

A good experiment along these lines would be 25-30 men (or say 3 fold attacker vs defender) against a normal, disrupted, pinned and demoralized unit. 3-1 is a typical assault ratio players I know use. It would show the difference in combat states. Add a leader to one side, then the other and then both to see if a leader has an effect on the results. One could then do the same experiment using clear, trench, bunker to see if any of those things have an effect....

Just some ideas

I think the bottom line is assaulting to 'kill' versus to occupy a hex isn't worth the potential ramifications of a failed assault, at least to me.

TheBigRedOne
06 Jan 09, 13:50
Never mind, your combat value ratings would simulate about what would happen when you've got a higher ratio of attacker to defender. I didn't see it going all the way up to 1000, which would be about 32 men with Garands roughly taking on 10 defenders based on your calculations. The different state experiment as well as the leaders would still be interesting...

It basically shows what we all know, the more men/power you have, the higher the likelihood of success in an assault.

jztemple
06 Jan 09, 22:13
I think these are the two important items to take from this whole analysis:


You have to inflict a minimum of one casualty on the defenders in order to have a chance of causing them to retreat. A simple ratio isn't sufficient, you need to have enough Combat Value in men and weapons to cause that casualty to happen.
Morale is critical to causing the enemy to retreat. No matter how poor the enemy morale is, if the attacker morale is less than "A" you risk failing to cause the enemy to retreat, and getting yourself pinned.


These are two points that a newbie can use in their gameplay.