Shooting targets with lasers
Sorry that I missed GURPS-Day yesterday. I was travelling in Germany for business. That was my first time in Europe since 1998 or so, when I spent three weeks near Eindhoven in the Netherlands. I need to go back.
Come to think of, there seems to be a litmus test for whether a given RoF system produces good results. I’m probably missing something, but here are some points that I noticed:
- For a Rcl 1 weapon (e.g. laser pointer), it should be possible to achieve 100% hit rates with reasonable skill, even when not engaging Close Stationary Targets (B408).
- For a Rcl 1 weapon with a very high RoF, doubling RoF should roughly double the number of hits, and shouldn’t increase the bonus to hit. A very high RoF in this case refers to such RoFs that in a single pass that intersects the target, there is no possibility of the pass occurring between the shots, given the weapon-training speed of the shooter (i.e. when a successful pass always indicates at least one hit).
- Firing 600 shots over a 1-minute turn should produce the same percentage of hits as firing 200 shots over a 20-second turn or as firing 10 shots over one second (unless there’s an Aim involved that only affects some of the shots), regardless of Rcl (unless Rcl is so bad that a penalty accumulates, which, in 4e, it doesn’t, ever).
- Shooting a target farther away (extra -5 for range) should produce the same results as shooting a smaller target (-5 SM), and the same as shooting a smaller target on a larger target (-5 SM smaller target on larger).
- A Homing weapon should produce roughly the same percentage of hits regardless of RoF, before things like warhead fratricide are taken into account.
I think a lot of these points are an outgrowth of the “people on infinite featureless field of battle” mentality.
Let’s take the first. With a laser weapon or something else with Rcl 1, you should get a lot more hits, right? In combat?
No. Not just no, but hell no. For my Dodge This article, +Peter V. Dell’Orto and I looked up the hit rate of – wait for it – people playing games like Photon or Lazer Tag. These are Rcl 1 weapons with RoF 1, usually. No autofire. But the hit rates tended to be about 6%. If you’re hitting that few times with semi-auto fire, you’re right in the center of the GURPS autofire principle: it’s for getting a few hits from many shots, not a lot of hits. Granted, this situation is chaotic, people aren’t aiming much, it’s usually dark. Hey, much like combat is supposed to be.
So point-the-first is dubious. Go ahead and try it. I bet that if you attach a laser sight to an airsoft rifle and could somehow calculate the percentage of time your beam was on-target in an actual chaotic situation, like an airsoft or laser tag game, it would be quite low. Your usage mode would also likely be “sweep the beam across the target,” not “hold the beam on target for a high percentage of the time.”
The second point – very high RoF where you can’t pass between the shots – is a misapplication of the rules. If you’re all or nothing, you handle this not with Rcl 1 and high RoF, but with RoF 1 and higher damage. This is a single-shot weapon for all intents and purposes. Rcl 1 is used for zero-recoil weapons like lasers, true, but it’s also used for firing three rounds from a shotgun loaded with shot rather than slugs. Rcl 1 is the lowest Rcl level, true – but it’s probably best read as “minimal spread,” but not single shots.
The third point totally neglects that a target can – and likely will – move. Hell, the shooter might move! Rules that achieve less, more, or the same hits as time increases can all be argued to make sense. Fewer hits because of much more opportunity to have relative chaos between shooter and target. More because of more time to compensate for that motion. The same due to a thought-to-be constant probability of hitting in aggregate taking all the shots as independent events. If you have an ability to put 76% of shots on target, that should be 76% over one second or one hundred. So this one assumes something that needn’t be true, but could be.
The fifth point is probably correct if you’re talking about cones of fire, and is probably a good design goal.
I’ll neglect the 6th because Homing is a game-mechanical term, and the precise thing its supposed to be modeling is left vague, probably intentionally.
Parting Shot
+Mark Langsdorf took a stab an an alternate way to look at Rapid Fire over on his own blog. As discussed in the forum thread above (and it’s worth reading, mostly) it’s not bad. Since Size Modifer and Range are both based on the Size and Speed/Range table, having the Rcl/RoF mechanic tied to that table makes good sense.
I’ll return to this a few times in the upcoming days. There’s a scatter diagram I did a while ago that’s worth revisiting for fun, and I want to try and tie autofire rules to the area fire rules as well.
I'm less concerned about a model for rapid fire that perfectly reflects reality than one that:
1) Is easy to use in play, either because it has no table lookups or only requires 1 table that is commonly referenced; and that doesn't have more than a few special cases
2) doesn't make RoF weapons unreasonably powerful (they're useful, but they're not the only solution to the "I need to attack someone over there" problem)
3) doesn't overly penalize defenders against rapid fire attacks
4) produces a reasonable number of hits as a percentage of the number of shots fired across a reasonable range (RoF 2 to 2000 shots/second; if you're firing faster than that I don't care if things get wonky)
as a special bonus, a system that
5) allows the GM to aggregate large numbers of attacks in a single roll that produces appropriate numbers of hits and appropriate numbers of defenses
would be really nice, since I conceptually like being able to say "30 goblins shoot at you… (roll some dice) 15 of them potentially hit. Dodge at (some modifier)" and have that work out to about the same (+/- 20%) as rolling 30 attacks and 30 defenses (for attackers in the skill 8 to 18 range and dodgers in the 7 to 12 range).
I want what Mark wants.
And the Lazer Tag vs. the suppositions of that poster make me think that a good rule of thumb is, it's better to try to derive common sense results from statistics than statistics from common sense results.
I'm right there with you both. In fact, the RAW by and large doesn't bother me at all. I'd flirted in theory with a MoS = 0 is one bullet, MoS > 0 is 10% hits per Rcl increment, but hadn't worked implications. The entire point of the Rcl mechanic is to make rapid fire attacks EASY, under the general assumption of "trying to get only a few hits" rather than "trying to put every shot on target."
I'll expound on this more in specific posts, probably, since ideas and concepts are flying through my mind right now.
The question is whether massed simultaneous fire by multiple shooters against one target should have the same game mechanic as one person firing a weapon which fires the same number of shots at one target. Maybe, maybe not.
The massed shooters are all firing only once. If the gunner shoots 10 shots, in theory he's firing one shot every 0.1 seconds (RoF 10) which even with little recoil (Rcl 1) gives some time for body and target movements to vary off target, or 10 shots simultaneously (RoF 1×10) in shotgun style pattern (which does tend to disperse). The rules tend not treat the question of recoilless AND instant fire, but a 2d pistol firing an attack of that sort is, as Doug said, really 20d, or possibly 10d(0.1) if there is a mechanism preventing drilling.
A system for rapid resolution of massed fire against a single target by multiple shooters is an interesting question. Historically, that sort of things seems relatively rare outside of firing squads, since situations of 20-30 guys all deciding to target one opponent, all able to fire and deciding to do so at once, and all being close enough that they can see and shoot at once are likely pretty rare, with most such massed fire being poorly aimed volleys from the musket era and the like.
But it can sometimes happen, and a way of mechanically doing it to enable fast resolution of the attack seems good. It is probably best to be careful with anything that overly penalizes dodge; one the one hand dodging all at once is silly; on the other a system that might encourage "all the baddies carefully aim at character one on turn 1… on turn 2 they all carefully aim at character 2…" seems a tad unfair.
Sorry Doug, but the Lazer tag thing is bogus. The game is played largely by persons defaulting to DX 10 and from what I remember seeing it's done in adverse light conditions too. The few people who have played long enough to have taught themselves a single pts worth of skill could easily be what's driving the numbers higher than Gurps 2% minimum Crit Hit.
If it has relevance to Real world shooting with real guns that's because that also is done largely by defaulters w/DX10 under bad lighting conditions with only a few Skilled users to push the numbers upward.
There isn't a thing about the what the 4e Rapid Fire rules were supposed to fix that adhering to the Skill level rules in Tactical Shooting doesn't cover better.
Not bogus, just of (admittedly, right in the post) limited utility. the point of bring it up is that zero recoil in reality is no guarantee of low shot spread or high hit rate in real life, which is what the game is trying to model.
One last point: the thing that the 4e Rapid Fire rules was trying to fix is stopping the game for minutes to roll many groups of burst of 3-4 shots in a large volley. One (or two, but one is better) roll to resolve is probably the primary design goal, and well worth pursuing.
Just as a technical note you were never supposed to roll more than 5 bursts in 3e. When ROF went over 20 you started breaking things down into 20 round groups. Then 100 round groups when ROF went over 100 and so on.
The real slow down was (and still is) rolling damage and hit location individually for each round that hits.
Printing the Rapid Fire Table on each and every character sheet would help too. Table look-ups are obnoxious.
Now that I think back, I do remember that part. I also created a Excel Macro or Visual Basic program called FireFight!, long since lost, that automated this fairly wonderfully, by burst, automatically calculating location, etc.
In addition, in Laser Tag, you are usually aiming for a small sensor, about the size of an eyeball.
So, the penalties for darkness (-3?), small target (-9), and moving and attacking (-2) all apply to that defaulted skill, along with the range penalties, which won't be too severe. Cumulatively, that is around -15 or a bit worse. It is a wonder that anyone ever scores a hit at all.
Add that to the fact that one way to know whether you have "come back to life" in Laser Tag is to shoot the wall and see if the gun is working again, and 6% sounds pretty reasonable, even with a high accuracy.
Rapid Fire in 3e was a nightmare, 20 shot groups or not – I was never sure in actual play how to deal with, say, the guy who fired a ROF 32 weapon with duplex rounds – 3 x 20 plus a 4 shot group? Okay, sure. Then resolved hitting, damage, and dodge for the piles of rounds he hit with. Took a long time. I don't miss that, for sure. That's why I am so on board with what Mark is asking for – simple to resolve, and if possible close to reality, too.
Yeah, what I feel is maybe (?) needed is a quick way of determining the size of the dispersed pattern. I think I know a way to do it. But basically, you look at the size modifier of your shots, and if it's less than that of your target, you just hit with everything, which takes care of that edge case. If the SM is bigger, you'd go into some other ruleset, like the RAW one, some of those proposed in the GURPS Forums thread, or just a simple percentage (compare SM of target relative to dispersion, convert to a 3d6 roll, resolve).
As a practical issue, hit rate tells you more about the value of ammunition than actual accuracy. The more precious your ammunition is, the less likely you are to waste it on low-probability shots. This in turn pushes out engagement ranges and pushes down engagement times, so everyone involved is less accurate.
I was with you until your last sentence, which doesn't follow.
On a practical level, GURPS rapid fire hit probabilities are deliberately low because otherwise cumulative HP damage and excessive die rolling both break the system too quickly and degrade playability.
On a less practical level, one problem with scaling up hit rates directly over time is that GURPS is not a simultaneous-move system but an alternate move turn sequence.
If for various reasons (large battle, space combat, whatever) you scale the turn to 60 seconds and scale the RoF accordingly, but still retain alternate movement, then you effectively saying that the other guy is letting you stay put and blaze away, uninterrupted, for 60 combat turns without any response before you turn around and (in the unlikely event you're still alive) do the same back to him.
Since this is not optimum, the reduced hit probabilities instead reflect that there's something else going – an exchange of fire with pauses for tactical assessment, evasion, cooling off of the weapons, and so on. (SPACESHIPS combines this with a general reduction of RoF by 1/20 from the maximum the weapons would be able to do so to allow for these pauses.)
However, the specific mechanics are red herring and debatable; there may be other ways to do it. The important point is scaling up the RoF over time simply doesn't work because it doesn't allow for the fact that people will react to constant standing still and firing well before the end of the extended period, and that reaction is likely to be some slackening in RoF.
Even if two sides were robots with no morale issues lined up on a flat plain, there would be some reaction, if only due to the fact that part way through that 60 seconds or whatever casualties would start to degrade both sides.
If