On the Fragility of Life

[AlexanderHowl]

GURPS does a fairly good job of showing the fragility of life. A skilled combatant in unarmed combat can leisurely kill a less skilled combatant in a minute just by punching and kicking them to death. While the less skilled combatant might struggle and get in a few good attacks, they are likely not going to survive. With melee weapons and stealth, a more skilled combatant can kill a less skilled combated quite quickly, potentially with one strike from stealth, allowing them to eliminate an enemy within a moment.

Imagine then a 250 CP mundane character with ST 12, DX 14, IQ 12, HT 14, Combat Reflexes, Night Vision 5, Stalker 4, Knife-20, and Stealth-20 (along with other unimportant traits for this exercise). If they came upon an enemy camp with one hundred soldiers in the middle of the night, how many of them could they kill in a night with a very fine long knife? Could they kill them all before they were discovered?

[Bengt]

Seems quite unlikely if the assumption is that all the soldiers are up and about since our attacker will have to succeed well enough on each encounter to stop the victim from raising the alarm. He only have to fail once to fail the entire challenge. I.e. even if he have 95% chance to take out one guy silently the chance is only .95^100 ~ 0.0059 to succeed 100 times in a row.

If he have to hide bodies along the way the odds are even worse since that will be more rolls.

If most soldiers are sleeping and he only have to take out a couple of sentries he will have much better chances of course.

[tbone]

Originally Posted by AlexanderHowl

Imagine then a 250 CP mundane character with ST 12, DX 14, IQ 12, HT 14, Combat Reflexes, Night Vision 5, Stalker 4, Knife-20, and Stealth-20 (along with other unimportant traits for this exercise). If they came upon an enemy camp with one hundred soldiers in the middle of the night, how many of them could they kill in a night with a very fine long knife?

It's a perfectly playable scenario. So there's a good way to find out the answer... : )

(I'll add that whether the marauder's Move is higher than the soldiers' should be pretty important. If it's significantly higher, he can keep try to keep from being surrounded once the soldiers sound the alarm.)

[coronatiger]

I did some calculating, using these simple assumptions:

• Soldiers have Per 10.
• Any critical failure on the assassin's part causes the alarm to be raised and the killing spree to stop.
• Killing an awake guard requires the assassin to win a quick contest of Stealth vs Per, and then rolling Knife until he succeeds. (A missed attack, i.e. 17 on the roll, has no consequence other than that the assassin has to try again.)
• The assassin wins the quick contest if he critically succeeds (since his margin of success would be greater than the guard's), or if his margin of success beats that of the guard while the guard doesn't critically succeed.
• If the assassin loses or ties the quick contest, the alarm is raised.
• Killing a sleeping soldier requires a not-critically-failed Stealth roll, then rolling Knife until success. (A missed attack allows retry.)

I get these preliminary results:

• Winning a quick contest has the probability of 96.69%.
• Succeeding before rolling a critical failure in the Knife skill has the probability of 99.53%.
• The probability of killing an awake guard without raising the alarm is the product of the two above, which is equal to 96.23%.
• The probability of killing a sleeping soldier without raising the alarm is equal to 99.07%.

In the extreme case of all the soldiers being awake, the probability of getting them all is 2.15%.

In the extreme case of all the soldiers being asleep, the probability of getting them all is 39.27%.

A few cases in between (awake/sleeping - probability of getting all 100):

• 5/95 - 33.97%
• 10/90 - 29.38%
• 15/85 - 25.41%
• 20/80 - 21.97%
• 30/70 - 16.44%
• 40/60 - 12.29%
• 50/50 - 9.2%

These probabilities aren't that bad (ca 1/3 probability of getting them all if there only are 5 soldiers on watch), but this assumes that any hit is an unnoticed kill. With the additional (and admittedly very harsh) assumption that upon being struck, a soldier who succeeds on a HT (=10) roll will raise the alarm (a miss still goes unnoticed, and any attack is lethal if the soldier fails the HT roll), the probability of not killing anyone is 48.12%, even if all the soldiers were asleep.

If I was the GM in this scenario, I would impose these additional requirements:

• The assassin has to roll damage; even a cutting attack to the neck can deal as little as 6 damage (1d+1, times 3).
• Whenever an "alarm is raised" roll occurs (using all the assumptions above), I would have nearby soldiers roll IQ to awaken and realize what's going on. I'd make a single roll for the soldiers as a group, with a penalty for being asleep, a bonus for being a group, and an increasing bonus based on the noise made so far.
• If the IQ roll succeeds, I would wake up one or more soldiers (based on margin of success), but unless it was a critical success, I would treat them as "awake guards" as described in the assumptions at the top for a few seconds, so the assassin would get the chance of taking them out before they can wake everyone.
• It would take the guards a little time to realize what's going on and to get their weapons, so the assassin would most likely get away if he doesn't try anything stupid.

The additional requirements would make it easier for the assassin than the harsh HT assumption would do. If he tries to keep fighting after all the soldiers are awake and armed, I would say he's doomed (but I'd still play it out, of course). Even if he managed to kill as many as 20 soldiers before the grand battle, he'd still have 80 to deal with. Even if a strategy of repeated Move-and-Attack actions so that only one of the soldiers can attack each round is viable, it would fail once that sequence of soldiers got in a few critical hits. And that's not counting all the guys with bows; the probability of 40 soldiers all rolling above 4 on their attack rolls is only 47.35%; give them another shot, and the probability drops to 22.42% of not a single critical hit.

Edit: By the way, by taking a -5 on the attack rolls to strike the neck, the assassin will critically fail on a roll of 17 (instead of just missing). This was not included in my calculations.
Edit 2: He can of course overcome that by making All-Out Attacks (determined), for +4 to the rolls. At least until he's discovered and the soldiers start hitting back.

[johndallman]

Originally Posted by AlexanderHowl

Imagine then a 250 CP mundane character with ST 12, DX 14, IQ 12, HT 14, Combat Reflexes, Night Vision 5, Stalker 4, Knife-20, and Stealth-20 (along with other unimportant traits for this exercise). If they came upon an enemy camp with one hundred soldiers in the middle of the night, how many of them could they kill in a night with a very fine long knife? Could they kill them all before they were discovered?

I've done a little of this in-game, and it depends on details not specified above, including the number of active sentries, and their organisation. Things I learned:

• Evaluate is really useful for surprise attacks without immediate time pressure.
• Realistically, don't try to cut throats. People with cut throats are very noisy.
• The best way with the GURPS combat system seems to be a grapple to the jaw and a stab to the vitals. This requires Extra Attack or All-Out Attack (Double). ST13, for the extra point of impaling damage helps a lot, because it makes your minimum injury 9, which is a fairly guaranteed major wound, in a way that 6 isn't.
• Having Luck, to re-roll critical failures, will help a lot.

[Icelander]

Not that it's all important for the purposes of this exercise, but innate Night Vision 5 is not possible for a human in a realistic setting. As per GURPS Tactical Shooting, a maximum of Night Vision 2 can be justified as realistic for a human.

More generally, I suspect that the fact that it generally takes a long time to kill a single person with a knife and there will usually be sounds of struggle and incidental sounds while the victim bleeds out will lead to even a very skilled infiltrator being discovered before he kills very many.

This is going to come down to GM adjudication, but even a person brought to negative HP and failing a death check or consciousness check will not necessarily expire noiselessly. A lot of people sleeping in the rough will be huddled close for body warmth and even if we assume somehow that there was no system for keeping watch, there might be people having trouble sleeping and/or having to get up to attend to bodily functions.

Even with a grapple of the face area to prevent screaming of a knife victim, there might be struggling, even sounds of air escaping from wounds (especially loud for cut throats, but possibly noticeable even with a stab to the vitals) that could be enough to alert someone sleeping right next to them. Hell, even just the smell of blood or bowel contents might awaken someone sleeping close, not to mention how someone would almost certainly awaken if the person sleeping next to them suddenly spilled the contents of their body under their joint covers. Even if all the soldiers are really private and shy about their sleeping arrangements, blood from arteries, like the ones in the neck or vitals area, can spray far enough to hit a person sleeping far enough away to avoid all physical contact.

So, basically, in game terms, killing more than one of the sleeping soldiers requires more than skill and abilities, it also requires luck about factors that the infiltrator has no real control over. And in any sort of realistic setting, that luck will eventually run out.

[AlexanderHowl]

I would disagree with that assessment, as you can develop the equivalent of Night Vision through operating regularly at night. If we assume that -5 vision penalty is the brightness of a crescent moon (since -7 is starlight), I learned to be quite functional in that light while I was a child because I had to feed and care for animals on a farm without artificial light (I found that flashlights were fairly useless for such activities because they ruined my night vision). Even though it was a routine activity for me, I found that many of my friends were hopelessly incapable of functioning in that level of light. Of course, that changed as I developed vision issues, but I was quite found of the night for a time.

From a technical point of view, humans actually have quite good night vision, as we can maintain visual acuity down to what GURPS would classify as a -9 (overcast moonless night). What would define advantageous night vision, would be rapid adaptation to differences in light, as it can take hours for people to achieve complete night vision. With the plethora of artifical lights though, most people have not been forced to operate in darkness during their childhood long enough to develop the proper visual pathways, so I would not be surprised if most people raised in urban environments suffer from a quirk level of night blindness.

[Icelander]

Originally Posted by AlexanderHowl

I would disagree with that assessment, as you can develop the equivalent of Night Vision through operating regularly at night. If we assume that -5 vision penalty is the brightness of a crescent moon (since -7 is starlight), I learned to be quite functional in that light while I was a child because I had to feed and care for animals on a farm without artificial light (I found that flashlights were fairly useless for such activities because they ruined my night vision). Even though it was a routine activity for me, I found that many of my friends were hopelessly incapable of functioning in that level of light. Of course, that changed as I developed vision issues, but I was quite found of the night for a time.

From a technical point of view, humans actually have quite good night vision, as we can maintain visual acuity down to what GURPS would classify as a -9 (overcast moonless night). What would define advantageous night vision, would be rapid adaptation to differences in light, as it can take hours for people to achieve complete night vision. With the plethora of artifical lights though, most people have not been forced to operate in darkness during their childhood long enough to develop the proper visual pathways, so I would not be surprised if most people raised in urban environments suffer from a quirk level of night blindness.

I agree that it should be possible to reduce penalties for some or even many tasks with training at performing them in low-light conditions, but that's not what Night Vision represents. Night Vision 2 represents being able to resolve tiny details at twice the distance in low-light conditions and is a fairly optimistic treatment of the maximum variation in human night vision.

Techniques for working in low light might be more expensive than Night Vision, but they are the canonical way to represent training in low-light operations. Granted, I think that buying such a technique for all combat skills quickly gets too expensive and I would totally support a leveled trait that reduced darkness penalties to multiple related skills, like for example combat skills.

But Night Vision is really too cheap for the purpose and, importantly, also allows the character to see as well in the dark as a cat. Which is a cool superpower, but not actually realistic.

[Þorkell]

What's the TL of the scenario?
Are all the hundred soldiers asleep when the attack starts?
Is the ground perfectly level and flat with no obstructions?

[tbone]

Originally Posted by Icelander

A lot of people sleeping in the rough will be huddled close for body warmth and even if we assume somehow that there was no system for keeping watch, there might be people having trouble sleeping and/or having to get up to attend to bodily functions.

For the sake of fun, it's possible to assume all 100 soldiers are asleep; but in any realistic situation, I think your assumptions are sensible. Even ignoring sentries, with 100 people in the camp, somebody is going to be awake.

And to really get realistic (at least where game realism is concerned), it's only fair to ask whether some of those 100 will have Light Sleeper, Danger Sense, etc.

(Still. If the soldiers aren't packed too tightly together, a good assassin could really get some work done. There's the proverb I half-recall, something about a running man with a knife... [hits the Googles] Oh, it's "Four thousand throats may be cut in one night by a running man", apparently, and its source is... Klingon?? Huh. Well, apparently, Klingons would be willing to give the assassin good odds in this thread.)