Arch-vile can't target certain monsters in vanilla, but friendly arch-viles can[edit]

The most recent series of edits excises all mention of friendly monsters from the main text and footnotes.  I have no opinion on that as a broad guideline (port-specific information has been added to articles without disruption in some cases), but here it is inconsistent because the new numbers in the "shots to kill" table presuppose friendly monsters.  That deterministic method of calculation can never be correct when a direct hit is forbidden (possibly excepting the barrel which has less than 70 hp).  So what do we do?  IMHO those lines should be blank, as before — a great many stochastic assumptions would be required to create a damage distribution, and AFAIK there is no preexisting research to refer to, so we just have to wait patiently for a more defensible analysis.    Ryan W 18:14, 3 October 2008 (UTC)

"Friendly monsters" kind of confused me at first glance because that's the name of a feature in some ports, that allow monsters that aid the player and such. But I see you mean those that can't or won't hurt others (harmless or non-targeting monsters).
I think the table should be consistent with what's possible in the game, and impossibilities should be addressed. In the case of the bosses, applying only the base damage, and in the cases of the monsters that aren't targeted, a value dependent on blast damage exclusively. With blast damage perhaps using about half of full damage may be convenient, as any hits that hurt the monster would range from the minimal to (almost?) the maximum, creating a mean somewhere between them. Perhaps it could be noted that a value for varying splash damage has been assumed somewhat arbitrarily just to give an idea because it depends on locational and random factors. Mainly because I don't think that is measurable in the same way as with targeted attacks. Who is like God? 20:02, 3 October 2008 (UTC)
Oh wait, doh, you are talking about friendly monsters (as per the title of the section). Normally, articles deal with "as is" data. Non-standard, usually impossible data could be included as notes to the table. The problem is that if you assume (use) arbitrary class monsters (also DeHackEd clones) in the table the information overlaps with the normal behavior. The normal info for the arch-vile against arch-viles is different than that of an archvile attacking an arch-vile with a different class applied to it. We could mention the info for the latter in the note, right after it explains that arch-viles don't target each other. Who is like God? 20:16, 3 October 2008 (UTC)

Detailed computation of fireball damage[edit]

Possible starting point for a technical discussion: [1].    Ryan W 20:35, 14 January 2009 (UTC)

Arch-vile weakness[edit]

Reading the entire arcticle I found that you forgot to mention the fact that the Arch-vile loses life or "hit-points" gradually when it revives or attacks a target. So the Arch-vile can die without being attacked by other monsters. This can be easily notice using cheats.


  • If you're hidden in a wall using "idclip" you can see how an Arch-vile becomes weaker while reviving other monsters; you'll notice there is a moment when the Arch-vile, in its attempt to revive a monster, suddenly dies.
  • In their way to attack a target Ach-viles also lose life and become weaker. You can note this for instance, in MAP32: Go 2 It in The Plutonia Experiment, when an Arch-vile is trying to kill a Cyberdemon (which is almost full), after some minutes, the Arch-vile dies being unable to kill the Cyberdemon. It can also occurs if the player is using "iddqd", there will be a moment when the Arch-vile dies being unsuccessful in its way to kill the player.

† JuanPa †

The first theory is almost certainly a myth. I did a test using Chocolate Doom and let the arch-vile resurrect over a 100 monsters and it still didn't die or even lose any hit points (it still took the regular amount of 4 SSG hits to put down).
The second EG is simply caused by the fact that arch-vile attacks do a small amount of splash damage to their attack target, which (in a similar fashion to the rocket launcher) also harms anything that is at close proximity of the actual target creature, the arch-vile itself included. -- Janizdreg 03:28, March 27, 2010 (UTC)

Maximum fireball range?[edit]

though they will not use their attack from more than 1024 map units away unless hurt   Maybe I'm misreading, but doesn't this imply that they won't even do it while hurt?    Ryan W 23:10, June 10, 2010 (UTC)

Yeah, misreading. As I said in the revenant talk page, the "just hit" flag will cause a retaliative missile attack regardless of range. It's only if that flag is not currently set that the function will look at the distance between the monster and its target. --Gez 00:04, June 11, 2010 (UTC)
Hunh.  Well, this demonstrates why I talk about code on talk pages instead of just editing the article immediately.  :>   Thanks.    Ryan W 00:23, June 11, 2010 (UTC)

On possible violation of demonic Geneva Convention (if they have one)[edit]

I remember reading somewhere on the Doomworld forums about the single case when the arch-vile can be retaliated against. As I recall it was something discovered by Maes in this Mocha Doom developments: that lost souls spawned by a pain elemental's death explosion somehow circumvent the exception and are launched with the arch-vile being their target. Can anybody corroborate with this? I hope I don't fable things here. Unmaker 16:14, 26 February 2013 (UTC)

Probability of pain state when hit with SSG[edit]

They are also the most resistant monster to pain, which makes canceling their attacks difficult; a point-blank shot with the super shotgun only disrupts their spell __% of the time.

The percentage was recently changed [2] from 55% to 49%, but according to this, both are incorrect:

  • Pain chance for each individual check: 8/256
  • Probability of no pain for one pellet hit: 248/256
  • Probability of no pain for 20 pellet hits: (248/256)20
  • Probability of stunning for 20 pellet hits: 1 - (248/256)20 ≈ 0.470051

Disclaimer: there were no replies to those talk posts, so either the data is obviously accurate or people couldn't care less.  :>    Ryan W (living fossil) 00:00, 28 April 2018 (CDT)

The DMINFO tool's '+paintab' option confirms that the RNG Results table is correct. Ditto for the (description of) tables for the other three games. --Xymph (talk) 00:33, 28 April 2018 (CDT)
LOL; why didn't I think of that?  Thank you.    Ryan W (living fossil) 18:05, 28 April 2018 (CDT)