Some minor annoyances, but be nice if they were fixed:-)

[sargebaldy]

I tried it a few times in ZDoom and found it very easy to do repeatedly, and it never took more than two shots. Trying it in Eternity it took me 3 shots half the time and otherwise 2. Something is definitely different.

[Grazza]

Yay! So I'm not mad after all.

Or at least, I'm no more insane than sargebaldy. :p

Seriously, thanks for testing it.

[Graf Zahl]

Grazza said:

Yay! So I'm not mad after all.

Or at least, I'm no more insane than sargebaldy. :p

Seriously, thanks for testing it.

I never need 3 shots for the spider (with any source port.) 2 should bring it down without problems. Just for fun I just killed the spider 20 times, only twice I needed only one shot (which according to the damage calculations and size of the spider is certainly possible - even with Doom2.exe) but there may be an explanation why this is easier in ZDoom: A few versions ago ZDoom fixed a bug in the trace code which might result in slightly more traces from the BFG hitting the spider if you are very close to it. Doom is losing some traces that cross the corner of blockmap blocks which might result in some misses that shouldn't really miss. Keep in mind that one BFG hit brings the spider down to very low health anyway (I often use one BFG hit and kill the spider with the shotgun or chaingun to save energy ammo) so a few more hits by BFG traces can easily make the difference here.

[sargebaldy]

Three because I'm not always the best aim. I can be very sloppy and still kill the spider in two in ZDoom.

[Grazza]

Graf Zahl: Yes, I thought it was most likely a blockmap-related issue (that's what I meant when I referred to "the way damage to enemies is calculated"). Any difference only seemed to become noticeable when Randy made that change (in 2.0.52 IIRC). The difference may not be so large, but it is enough to change the probabilities in this case quite a lot.

2/20 sounds like a run of very bad luck to me, assuming it was getting hit by a full portion of traces.

Yes, one-shot kills in Doom2.exe are possible - this happens in some demos (ddqr springs to mind). Indeed I got one a few hours ago when testing this, but it was only after a string of many unsuccessful attempts (I wasn't counting, but it was at least 20).

[Quasar]

The theoretical maximum damage of the BFG, minus damage from the projectile, is 5120. So the true holy grail of DOOM is a one-shot Cyberdemon kill. Someone on #zdoom ran a simulation of this which used code almost identical to that in DOOM to calculate BFG damage assuming all 40 tracers hit, and not once out of a few thousand tries did the damage exceed 4000. And since it's questionable whether you could actually hit a cyber with all 40 tracers, it may really not even be possible. I'd like to know for sure though ;)

If all 40 tracers hit, the average damage per tracer would need to be at least 100. Even if every tracer did the max 128 damage, you'd still need to hit with 32 of them.

[Anders]

Doom's random system makes this impossible, there isn't 15*40 consecutive numbers with all three lower bits set anywhere in the random table. Heck, the randomtable is only 256 numbers so for this to work ALL of them would have to have all three lower bits set. Someone should make a program that uses dooms random table to see what the real maxdamage is.... it's only 256 tests if you want to brute it.

But then, if you were talking about doing it in a sourceport with a different random system... never mind.

EDIT: i made a testprogram, and it turns out that the theoretical maxdamage from the bfgtracers (ie. not including the projectile) is 2728, which happens once in 256 shots. the theoretical mindamage is 2639 (also once in 256 shots)... the most probable damage is 2687, which happens 14 times out of 256. the medium damage dealt if all 40 tracers hit are close to 2686 (2685.9375).

EDIT2: Ah, i see the random function is called once every tracer for a non-damage thingy, updated the figures above to reflect the new behaviour.

[Hirogen2]

There is also some behavior already present in vanilla Doom that "deals" more
damage to enemies than by default. I think you this, just idclip into the
monster and hit the trigger.

[Quasar]

The *theoretical* maximum is 5120, because if you COULD roll the maximum every time, that's what you'd get. The actual empirical maximum is necessarily lower, as you have pointed out.

[Anders]

The random table is part of doom's damage theory. So 4800 (40*15*8) (i don't know where you got 5120 from) isn't even theoretically possible.

[Quasar]

Well ok if you want to consider the behavior of the RNG when talking about what is possible with the raw calculation of damage, then fine. And you're right. I miscounted the number of dice rolls in the summation for BFG damage, so the Eternity codepointer docs are incorrect. I seem to have thought there were 16 terms in the summation when there are only 15 (numbered 0 through 14).