View Full Version : The statistics of destruction

28 Jul 03, 19:18
I did a little number-crunching.

Given the worst possible pK > 0 (1%), let's assign that to an M16A2. This means that any single burst fired by an M16A2 has a 1% chance of killing an enemy fire team.

Let's assume for a second that the pH is 100%. Based on my observations for desert terrain, that doesn't seem far off, as pretty much any shot gets at least a suppression result.

Further assuming that an M16A2 can generate 15 bursts per minute on average, we find that a single M16A2 will have about an aggregate 14% chance of killing the enemy fire team for each minute that it is firing. Add the second M16A2, and we find that a given fire team has about a 26% per minute to kill an enemy fire team.

Keep in mind that they are typically firing near the weapon's max effective range. This may seem poor accuracy for a firing range, but combat has been shown consistently to grossly degrade marksmanship, assuming you can train them well enough to reliably fire their weapons at all.

If we give them the weapons they *should* have, lets add a SAW to the mix. At 2% per burst and 20 bursts per minute, the SAW gunner will have an aggregate 33% per minute to kill a fire team by himself. Added to his teammates, that would be an overall 50% chance per firing team per minute.

Now, perhaps this is too high or too low. I'd have to dig up some more combat statistics to verify it. What numbers I have from Korea and Vietnam don't seem to contradict it. A couple of interesting titles on combat performance in Korea can be found at:

Regardless, I'm pretty sure its a lot closer to accuracy than the present situation, which is:

A single M16A2 @ 33% per burst, 15 bursts per minute = 99.75% chance of a kill per minute. In fact, he has an 86% chance to kill in 20 seconds or less. The 2 M16A2s in a given fire team together have >80% chance to kill a fire team in just *two bursts* each, less than 10 seconds, which certainly lends credence to my complaint that targeted fire teams usually die within the first couple of bursts. Put another way, they have about a 2/3 chance to hit a man with any given burst, which is pretty darn good if you ask me. Even cops firing at much shorter ranges don't begin to approach that level of accuracy, generally runing about 4 shots per cop per incident with overall 10-20% hit rate. Machine guns, obviously, would be even worse. With MGs, you'll typically be hitting more than one man per burst on average.

Like I said: for all practical purposes, (fire at infantry) == (kill infantry) in v1.02

--- Kevin

29 Jul 03, 12:00
In case 1% still seems low to you, consider this:

If each 3-round burst has a 1% chance to kill a fire team of 3 men, then that's the same as saying that 1 round has a 1% chance to kill one man. Put differently, it will take an average of 100 rounds to kill one man.

That's *still* substantially better than estimates of hit/kill ratios in modern warfare, which typically run to thousands of rounds per kill.

Even if we factor in pH, we'd have to drop pH to 10% with pK at 1% before we hit 1,000 rounds per kill, which is still a pretty optimistic number. I am quite certain that pH is way, way higher than 10%; probably higher than 75%, based on my playtesting.

33% pK? That's the same as saying one round has a 33% for one man, or on average three rounds per kill. Like I said before, that's significantly better than cops do at an average engagement range of around 20 meters, even though our boys are typically engaging > 250m in the game. If you further consider that those three rounds are really, for all practical purposes, a single shot, you are now in the hallowed land of "one shot/one kill". That's pretty hard to do on a nice, orderly firing range (my personal best is 39 out of 40), much less in the chaos of combat.

Can this kind of kill ratio really be justified from history, or even MILES? I certainly can't find anthing that even comes close to this practically cybernetic effectiveness. Even Hollywood doesn't portray quite this level of superhuman accuracy.

--- Kevin

29 Jul 03, 12:30
I think that your calculations make sense.


31 Jul 03, 19:26
That's some serious number crunching. I feel like I'm in geometry class again.