After SUPER shotgun, why go back (some do)?

Maes said:

Sure. We consider only cases the case where 8 calls to P_Random are made, as the last fatal pellet will carry over the damage of a sequence of "straight 8s", and the "inflictor" condition, whatever it's supposed to stand for, cannot apply to 7 consecutive shots on the same target:

There are 241/256 chances that the shot will be fatal (just add up all cases where damage is equal or greater than 60.

Yeah, this only means a 94.14% chance of getting a 1-shot kill on a healthy imp. With the "inflictor" condition on all shots, the chance would actually be lower: 226/256 = 88.28%. In either case it's quite not "a shell, an imp". More like "A shell an imp...most of the time". Sorry if I shattered dreams or wrinkled e-shafts ;-)

Vs a CG dude, odds change dramatically: getting over 70 HP of damage in one shot only has a 176/256 = 68.75% chance, which seems about right.

A one-shot kill vs a healthy lost soul only has a 4/256 =1.5% chance of happening. In other words, don't bet on it.

According to http://www.doom2.net/single/weaponfaq.html, the chance to kill an imp with one shell is 7/8, or 87.5%. That is roughly correspondent with the "inflictor" conditions of your calculations.

Hope I'm not going off-topic, but what I also find extremely interesting, however, is the ridiculously minute chance (1 in 5 million) to kill a revenant with one ssg blast, due to the pseudo-rng calculations of the doom engine. However, the 1 in 5 million statistic from the chart also assumes 0 misses of the pellets, meaning the conical blast radius of the ssg and the relative thin stature of the revenant itself are irrelevant. In this case, it means that purely due to the doom engine, even if all 21 pellets hit, a weapon that does a theoretical maximum damage of 315 would still only have a 1 in 5 million chance of killing a monster with 300 health. I find that mind-boggling, yet in my demo-watching and playing experience, I've never seen anyone one-shot a healthy revenant.

I love this page (DOOM Weapon Damage Information). A pity there are no more probabilities there which are relevant in casual play:
- odds of killing a cacodemon with 2 rockets
- odds of killing a baron with 5 rockets
- odds of killing an arch-vile with 3 rockets, 4 rockets

Too lazy to calculate it ATM, but I find killing a caco with 2 rockets less likely than a baron with 5 rockets, which is kinda strange.

And a revenant in one SSG shot, come on! in some situations killing it in 2 shots is quite a success. I also think that with Maes' extended calculations this chance is in fact not 1/5000000, but 0.

vdgg said:
Too lazy to calculate it ATM, but I find killing a caco with 2 rockets less likely than a baron with 5 rockets, which is kinda strange.

More hits means the results tend to get closer to average, thus barons usually get around 218 points of damage per rocket in total. With only two hits, the damage to cacos varies more, which means below-average damage is more common while above-average damage just causes the same effect as median damage. Likewise, it's easier to kill a hell knight with two rockets than a baron with four.

Qaatar said:

According to http://www.doom2.net/single/weaponfaq.html, the chance to kill an imp with one shell is 7/8, or 87.5%. That is roughly correspondent with the "inflictor" conditions of your calculations.

Yeah, but they used a much more simplified model for their calculations e.g. that all 7 pellet shots are likely to assume ANY values from the table, which is clearly not the case in an actual call ro the P_FireShotgun() since random values are used in a (well, several) specific sequence(s) which makes some values probable and others not as much. Of course, they fully acknowledge this deficiency themselves.

The good news is that a more accurate simulation is possible, as I proved with my simple program, once the gory bits of the actual attacking functions are figured out. The SG was one example, the SSG may soon follow ;-)

If anything, I'd like to know what's the purpose of having that complex "inflictor" condition (which generates one extra call to P_Random() if several conditions such as target's low health and opponent height difference apply). For some types of weapons and monsters, it's simply technically impossible to ever occur (seems more probable if you strike weak monsters with powerful weapons such as rockets and BFG blasts).

OK, here's a lost soul killed in one SG shot:
http://www.doomworld.com/vb/showthread.php?postid=868957#post868957

doom2day said:
The normal shotgun's extremely thin/nonexistent vertical spread brings it second to the chaingun for 'sniping' usefulness.

Also, if there's a lot of thinly spread, weak enemies, the normal shotgun is more efficient ammunition-wise.

Exactly what I think.

doom2day said:
Also, if there's a lot of thinly spread, weak enemies, the normal shotgun is more efficient ammunition-wise.

Now that I think about it, in general, the level design in DOOM and DOOM II seems to fit the main shotgun the player will be using, respectively. DOOM II tends more toward larger areas where more monsters or larger monsters may gather, while DOOM is more about hallways and separate rooms, some of which you may snipe across through windows or the like.

vdgg said:

I love this page (DOOM Weapon Damage Information). A pity there are no more probabilities there which are relevant in casual play:
- odds of killing a cacodemon with 2 rockets
- odds of killing a baron with 5 rockets
- odds of killing an arch-vile with 3 rockets, 4 rockets

Too lazy to calculate it ATM, but I find killing a caco with 2 rockets less likely than a baron with 5 rockets, which is kinda strange.

And a revenant in one SSG shot, come on! in some situations killing it in 2 shots is quite a success. I also think that with Maes' extended calculations this chance is in fact not 1/5000000, but 0.

Also, the doom wiki puts the chance of killing the Revenant with exactly two super shotgun blasts where all pellets hit at 100%
What calculations are you two using? Are the ones at doom.wikia.com wrong?


I would like to put out there that assuming that all pellets hit always, that the shots required to kill from the super shotgun have less spread than the rocket launcher.

The shotgun and SSG have different uses - shotgun for small groups of former humans, or lone imps, and the tighter shot spread makes it more effective at medium range than the SSG. The SSG is for big groups of imps or former humans and the stronger enemies. Neither is actually better than the other all the time, just in different situations.

doom2day said:

Also, the doom wiki puts the chance of killing the Revenant with exactly two super shotgun blasts where all pellets hit at 100%
What calculations are you two using? Are the ones at doom.wikia.com wrong?

Not wrong, but incomplete. In any case, they acknowledge their deficiencies.


I applied the same anal-ysis I did for the Shotgun to the Super Shotgun and here are the results for avg/min/max. The number of actual calls to P_Random() depends on the conditions, with 5 being the minimum (HITTING NOTHING) and 12 the maximum (VERY IMPROBABLE). The typical conditions encountered when playing are 10 and 11 hits.

      5 call    9 call  10 call 11 call 12 call
AVG   207       207     207     207	207
MIN   155	155	170	175	170
MAX   250	250	245	250	240

It may appear shocking to some that there's no way to achieve the maximum damage of 15 per pellet = 300 HP max damage, yet it's true: the way the values are arranged in the P_Random table in vanilla Doom, no matter in which position you start on it, and no matter how many times you trasverse it with any of the possible usage patterns.

So where's the beef with the wikia calculations? Those dismiss computing the actual P_Random() usage patterns as "too complex", but in reality each time you shoot with the hitscan weapons, the calls to P_Random() have a very predictable and, most importantly, unbreakable pattern ("atomic" for those CS junkies out there).

Again, this is for UNHACKED VANILLA DOOM, WITH ALL PELLETS HITTING (computations for partial cases are possible). Source ports have totally different RNG mechanisms, and certain TAS speedruns use RNG manipulation.

So what does this mean in terms of combat?

1. That a single-blast pinky kill is ALWAYS possible assuming ALL PELLETS HIT, and that the minimum damage in normal conditions is 170 HP, not bad at all.
   
2. One-kill revenant kills are impossible, as there's no legal trasversal of the P_Random() table that would give you 300 HP of damage, not even close. Even the wikia acknowledges that, the 300 HP damage mentioned in the theoretical max do no even appear in the table. And yeah, two-shot revvie kills ARE always possible, ASSUMING ALL PELLETS HIT, since the minimum combat damage is 150 or 175 HP.
   
3. There are 0/256 (10 calls) or 1/256 (11 calls) to kill romero's head in one blast, so there.
   

About situations that require two hits: instead of running a second sim I applied probability theory to multiple single-shot cases.

So, what does this mean? The, maximum damage from two shots is either 500 (with a 1/65536 chance) or 36/65536 ~ 0.055%, so two-shot arachnotron kills on undamaged arachnotrons should be exceedingly rare and/or impossible in vanilla Doom.

UNfortunately simulating all results from two shots would generate A LOT of data, too much to handle with Excel alone, so I'll just report what happens if I simulate 40 sequential pellets (keeping in mind that there are actually 65535 subcases now, and I'm merely reporting one).

Chance to kill a Caco/PE in two shots: 222/256 = 86.72% (223/256 with "inflictor")
Chance to kill a revenant in two shots: 256/256 (100%, no brainer ;-)
Chance to kill an arachnotron in two shots: 0% ( this probably contraddicts most people's experience, but don't mix source ports with vanilla).

I will try and perform a super-exhausting analysis when possible, but I'll probably have to rewrite the sim code to count for occurences where a certain amount of damage has been exceeded, rather than loggin g in ALL partial results.

Personally I use both shotguns. Shot gun has faster reload time and is great for zombies, shargents, imps, and lost souls. How ever the shotgun is not as great for other monsters. The super shotgun has the power to kill demons, hell knights, mancubus's, cacodemons, pain elementals, etc.
So I need both for different monsters.
PS. interesting post!

the normal SG is much more fun to use, it has the ability to go long-range, it takes less time to relaoas then the SSG, it has this over power vertical hit scan, and, in close combat you smash often more than 3 imps/zombies per hit.

in my E1 project, the normal shotgun will not be found before E1M5
just beacuse it is that good.

Maes said:
I went through GhostlyDeath's code analysis, and tried to count the actual number of calls to P_Random() per pellet on every possible code branch (assuming that they're all there):
[snipped by moderator - don't quote massive tracts of posts; only quote snippets that you're responding to, or don't quote at all if it is clear what you're responding to]
...

Well... In the branch for firing the shotgun only those calls are made and such. Also, the 7 seven pellets are done in succession. Due to Doom's single threaded nature, no other calls can be made. I only did the same case every exact time, not every possible combination of those variable methods.

Three words: Accuracy and Conservation.

Qaatar said:

I find that mind-boggling, yet in my demo-watching and playing experience, I've never seen anyone one-shot a healthy revenant.

I have with the SSG.

Yeah, I've done or seen it with the SSG.

The shotgun and SSG are very different weapons with different purposes, and I still use them both a lot.

Shotgun:
-Medium/long-range sniping
-Taking out dispersed groups of imps, sergeants, and other single/double-shot enemies
-Taking out imps in long, narrow lines (happens more than you think); you have a good chance of taking out multiple imps with one shell every so often.
-Ammo conservation in ammo-scarce maps

SSG:
-Taking out large clumps of imps, sergeants, etc. clustered closely together
-Close-quarters fighting against mid- to high-health enemies
-"Hit-and-run" tactics with stronger enemies: Rush up, quick SSG shot to the face, retreat/dodge retaliation. Rinse and repeat.
-Circle-strafing stronger enemies

Basically, the shotgun is most effective against lower-level enemies, from long range to short range, while the SSG is best against mid-level enemies at close quarters and crowd control of lower-level enemies. Usually the WAD or individual map will set the tone for which one I default to, but since most maps seem to be designed with shells as primary ammo it's almost always one of the two.

While the huge thud and power of the SSG is undeniably satisfying, it also feels great to zig-zag around with the regular shotgun and take down a (minor) monster with every shot. It's especially fun if you can just hold down the mouse button and aim well enough to hit something every time it fires. :)

Side note: Being able to take down a pinky demon with one SSG shot is nice, but if you're not chainsawing every one of those bastards as is practically possible you're doing it wrong. :)

Yeah the two shotguns are perfectly implemented. Alongside the additional monsters it's the key strength of doom 2.

Where it gets interesting is how you balance the different rates of fire, damage etc. with known things or things you can work out like monster pain chance. An SSG might allow you to yield more damage in a multiple monster fight, but the SG would give you more opportunity to trigger pain chance, possibly meaning that you don't get attacked once, even in a longer fight. And that's usually the trick in doom in tough maps.

From a gameplay perspective, ignoring realism of shotgun for a moment, I can't think of any game since that affords you two such different options within the same class of weapon, for all the reasons given in the posts above.

myk said:

Yeah, I've done or seen it with the SSG.

Not in vanilla you didn't.

I've switched to the regular shotgun over the SSG many times, particularly in DM maps that have giant courtyards. Usually against monsters the only reason I would use the regular shotgun would if the map is very generous in ammo and I want to give the monsters a bit of a handicap so that it's more difficult on my part.

Hi, lurked a while but never bothered to actually register until now. Hopefully I don't piss someone off or something.

To answer the OP, I tend to use the SSG over the regular shotgun, but the single-barreled shotty is great for taking out isolated, low-health enemies. Honestly, I always thought the DB was a bit overpowered for early in the game and Doom 2 gave it to the player too soon. That is, until I stumbled upon the red key trap in MAP2... those imps will rape you if you don't know what's coming.

SSG is king. Unless I have Plasma Gun. I just love shooting it!

TraderEddy said:

SSG is king. Unless I have Plasma Gun. I just love shooting it!

Especially with infinite ammo!

GhostlyDeath said:

Especially with infinite ammo!

Especially with infinite ammo. I have to say though once I found Skulltag I started loving the grenade launcher. What Doom 2 needed was that each weapon should of had a second weapon like SG and SSG. So Skulltag did well, still think that we need like a magnum or something to replace the pistol.

The chaingun is the pistol's second weapon already.

I usually use super shotgun for boss fights, and crowded rooms, and the shotgun for everything else. I use the chainsaw or fist to kill pinkies, unless there's other monsters in the room shooting at me already.

I think the shotgun is still very useful after getting the super shotgun, unlike in later id games..

The shotguns in Quake I and II.. In Quake I, the shotgun has very minimal use, since both shotguns shoot the pellets in same kind of spread and both shoot quite the same speed. In Quake II, the shotgun becomes pretty much useless after getting the super shotgun, shotguns functionality seems quite the same as in Quake I. In both Quakes even the weakest enemies need two shotgun shells and there aren't many situations where you would like to snipe far away monsters with normal shotgun (in Quake II, even less, because the rail gun does the sniping better).

In Quake that is justified because it's the starting weapon. You can say the same happens to the pistol in DOOM, which is also the starting weapon. Another thing about Quake is that its damage system is less random than DOOM's and weapon switching is very quick. This should make gauging that one needs a single shotgun blast to finish an opponent relatively easy, and should allow one to switch quickly to it and back to another weapon, in the process.

I recall killing an arch-vile with 3 ssg blasts once. What would be the probability of that happening?

xit-vono said:

I recall killing an arch-vile with 3 ssg blasts once. What would be the probability of that happening?

ON vanilla Doom, none, because there is NO possible trasversal of the random table that would give you 20 x 15 =300 HP damage, let alone thrice in a row (900 HP, the health of an undamaged ArchVile).

If you witnessed it in vanilla, then the Archvile was already damaged beforehand (don't let the fact that it can't become the target of infights confuse you, it can still be damaged by other monsters by mistake).

Some source ports have a different RNG, so it might possible although extremely broken if it happened thrice in a row. This sounds more like a RNG manipulation/TAS hack to give you always maximum damage.

I had even see a zombie apparently gibbing after a shotgun blast, but what had actually happened was that it was already low on health, and an imp fireball with maximal damage hit it, causing it to gib. Maximum damage imp fireballs themselves are not rare, damaging the zombie just enough to let that happen is hard but not outright impossible.