Reality testing firearm rules

[Varyon]

Originally Posted by Fred Brackin

Yes, that's how it works but it works out badly at larger scales.

If your target is at 100 yards and unmoving that's a-10. However 100 yards distant and 49 yards per second (98 mph) is also -10. Only at 1 yard farther or faster per second does it become -11. This does not pass the small test.

You probably need something like a penalty for distance with a seperate penalty for movement as a fraction of distance.

In the above case a better example would be something like 101 yards and unmoving being a -11 and 101 yards and moving at 49 yards per second also being -11, because in both cases the target is treated as being 150 yards away (anything that isn't exactly at one of the SSR breakpoints gets rounded up, so 101 is treated as 150... although most GM's would probably opt to round down instead).

And, yes, the nature of GURPS is that it has breakpoints, and these can lead to situations like the above. But breakpoints are necessary to keep the game playable, and typically you aren't going to run into issues with them unless someone is explicitly trying to break things in this fashion. Now, there's a deeper issue here, in that projectiles that actually take a bit of time to hit the target have more issues due to lateral movement than do hitscan weapons (where the issue is simply tracking the fast-moving target, not successfully aiming in front of it to account for the time the projectile takes to reach the target), but in general GURPS simply treats all ranged weapons as hitscan for simplicity.

[Fred Brackin]

Originally Posted by Varyon

And, yes, the nature of GURPS is that it has breakpoints, and these can lead to situations like the above. But breakpoints are necessary to keep the game playable,.

It's not so much that there "are" breakpoints it's the relative scale of what's on each side of the breakpoint.

100 yards would be huge for the Melee/Wizard arena combat that was Gurps birthplace (through the Fantasy Trip going on to Man-to-Man). 100 mph would also be huge but for a game that's trying to handle modern military combat 100 yards is the next thing to knifefighting range. Meanwhiel 100 mph is next to impossible for ground vehicles but near the minimum for air vehicles.

<shakes head>Arena scale just isn't going to cut it any more

[ericthered]

Originally Posted by kenclary

And in general, yes, plinking at a target is much easier than shooting in combat.

The gun I know how to shoot is a shotgun, which is usually fired against moving targets. There is a HUGE difference in difficulty between shooting at a machine thrown target (standard), a target thrown by hand (harder), and game (much harder). Its also easier to hit once you've "warmed up", another range condition. So yeah, the range is useful practice, but doesn't match actual conditions.

[mburr0003]

Originally Posted by Varyon

It does use both, but they stack before setting the penalty rather than after. That is, if a target is 10 yards away and moving at 5 yards per second, you are at -5 (for 15 yards) to hit rather than at -4 (for the worse, 10 yards) or -6 (-4 for 10 yards, -2 for 5 yards).

Right, sorry, by "use both" I'm to literally take both and add the penalties. Granted I think that should only apply for "perpendicular" movement, movement inline with the attacker shouldn't be as heavily penalized with a flat firing arc (or even firearms at long range). For arced fire (mostly bows or thrown weapons at longer ranges, which is more akin to indirect fire anyway) I'd probably want to add them as well.

Originally Posted by Fred Brackin

You probably need something like a penalty for distance with a seperate penalty for movement as a fraction of distance.

True, but there is a vast difference between a target moving at/away from the firer and one moving perpendicular. Especially at higher speeds.


So far I've just sighed and tried to ignore it, it's a fiddly bit that might just be "below granularity".

[Anthony]

Originally Posted by Fred Brackin

You probably need something like a penalty for distance with a seperate penalty for movement as a fraction of distance.

No, you really want a penalty for movement as a fraction of size, with an additional modifier for travel time. There are two main ways target movement is a problem:
Tracking
The basic way aiming works tends to follow Fitt's Law, which models target selection as starting at an distance from the target (as it was originally for designing control panels, that distance would be based on the size of the control panel), and then the user makes a series of corrections, each of which halves the error. However, when the target is moving, a new error is introduced every time you do a correction, meaning total error asymptotically approaches a limit of (target velocity) * (correction cycle time) (this can be thought of as an application of the Steering Law). There are some options for compensating when the target's velocity is highly predictable, but most combat interesting targets don't have highly predictable movement. High skill can probably be assumed to reduce the length of a correction cycle, so a general formula would be something like (Skill) + (Size Modifier) + (Speed Modifier) + (K).

In addition, most braces are not designed for easy tracking, so you would expect to lose any bonus for a brace. Something like a weapon stabilization system would still function, however.
Leading the Target
Against a moving target, you have to lead the target by (target speed) * (range to target) / (projectile speed). This leads to two problems:

1. We aren't particularly well wired to do that, though mechanical systems (such as marks on sights) can assist.
2. We might not know the range in the first place. Knowing range by parallax is only particularly effective at short ranges and drops to zero at a couple hundred yards. Knowing range by apparent size vs expected size requires knowing the target's actual size, and isn't something we're super good at. Absent rangefinding sensors, the most reliable option we have at longer range is comparing target position to landmarks, which is part of why sighting guns is so powerful.

An approximate formula would be something like (Skill) + (Size Modifier) + (Speed Modifier) + (Range Modifier) - (Projectile Speed Modifier) + (K).

Note that projectile drop has the same issues as leading a moving target, except the effective target velocity is G * travel time / 2.

[Fred Brackin]

Originally Posted by Anthony

No, you really want a penalty for movement as a fraction of size, with an additional modifier for travel time. [/list]An approximate formula would be something like (Skill) + (Size Modifier) + (Speed Modifier) + (Range Modifier) - (Projectile Speed Modifier) + (K).

.

As usual I am confused by your math. I even missed wherever you defined "K".

[Ulzgoroth]

Originally Posted by Fred Brackin

Yes, that's how it works but it works out badly at larger scales.

If your target is at 100 yards and unmoving that's a-10. However 100 yards distant and 49 yards per second (98 mph) is also -10. Only at 1 yard farther or faster per second does it become -11. This does not pass the small test.

You probably need something like a penalty for distance with a seperate penalty for movement as a fraction of distance.

What part of this is supposed to be a scale problem?

It's equally true that something at 10.1 yards has the same penalty stationary or moving at 4.9 yards per second. There are problems, but the scaling is the part of that that doesn't have a problem.

[Fred Brackin]

Originally Posted by Ulzgoroth

What part of this is supposed to be a scale problem?

The part where I said that 100 yards is not a long distance for rifle fire but even 50 yards per second is a very high speed for ground combat. Then 50 yards per second at a distance of 100 yards would make accurate rifle fire far more difficult than if the target was unmoving ...... except that it wouldn't be by RAW.

If the system breaks that easily maybe it should be fixed.

[Rupert]

Originally Posted by Anthony

[*]We might not know the range in the first place. Knowing range by parallax is only particularly effective at short ranges and drops to zero at a couple hundred yards. Knowing range by apparent size vs expected size requires knowing the target's actual size, and isn't something we're super good at. Absent rangefinding sensors, the most reliable option we have at longer range is comparing target position to landmarks, which is part of why sighting guns is so powerful.

At the range small arms are used there's another ranging method, taught to soldiers. That is to use the level of detail you can resolve to tell how far away the target is. For example, at a certain distance humans will seem to lose their necks, and then their heads will apparently 'sink'; into their torsos and also lose their eye sockets. Further away even more detail is lost. This is learnable and is combined with apparent size to get ranges. It's good enough to allow hits with small arms at practical ranges, but nothing but lots and lots of practice will make leading running people at various ranges possible.

[Ulzgoroth]

Originally Posted by Fred Brackin

The part where I said that 100 yards is not a long distance for rifle fire but even 50 yards per second is a very high speed for ground combat. Then 50 yards per second at a distance of 100 yards would make accurate rifle fire far more difficult than if the target was unmoving ...... except that it wouldn't be by RAW.

If the system breaks that easily maybe it should be fixed.

I don't think absolute speeds matter much as a factor...

But certainly I agree that a crossing target and even more so a fast-crossing target (and 22.5 degrees per second is pretty fast) should be a more serious problem for a human-like shooter than it is.