Is cut damage meant to be more dangerous than stun?

[pampl]

In init2, on line 1830, where the danger level of various attacks are calculated:

		switch (method)
		{
				/*stun definitely most dangerous*/
				case RBM_PUNCH:
				case RBM_KICK:
				case RBM_BUTT:
				case RBM_CRUSH:
				{
					atk_dam *= 4;
					atk_dam /= 3;
					break;
				}
				/*cut*/
				case RBM_CLAW:
				case RBM_BITE:
				{
					atk_dam *= 7;
					atk_dam /= 5;
					break;
				}
				default: 
				{
					break;
				}
			}

That means stun damage only counts for 133% normal damage but cut damage counts for 140% normal damage. Was cut supposed to be 6/5?

[fizzix]

Stun is more dangerous because of its side effects. Reduced to-hit, increase failure on activations and spells, and the possibility to get knocked out if you are stunned too much. Cuts only have a small leaking of HP, something that is unlikely to kill you except maybe in the very early game.

[Tiburon Silverflame]

Which means the OP's point is valid. Stun is given a lower rating, by those computations.

[Timo Pietilä]

Which means the OP's point is valid. Stun is given a lower rating, by those computations.

It might be that there have been in some ancient version of angband where cuts could have killed you. There is "cure mortal wounds" -spell in priest spell list. Mortal as in "you'll die without cure". I don't recall such situation though.

[Derakon]

"Mortal wounds" mean that the player's cuts timer is more than 1000. Without access to magical wound healing, this will deal at least 2700HP damage (3HP/turn for timer > 200, 2HP for > 100, 1HP otherwise). I think natural healing is also turned off when cut (certainly when poisoned). So yeah, those are "mortal" wounds. I certainly remember being very concerned for my frog-knows characters when they got mortally wounded, because I didn't carry healing items and the only healing spell I knew, Cure Light Wounds, didn't touch the cuts timer.

(I was not very old when I played frog-knows)

[camlost]

It might be that there have been in some ancient version of angband where cuts could have killed you. There is "cure mortal wounds" -spell in priest spell list. Mortal as in "you'll die without cure". I don't recall such situation though.

"Mortal wounds" in S don't heal themselves over time.

[Magnate]

Which means the OP's point is valid. Stun is given a lower rating, by those computations.

The original discussion of the relative dangers of different attack types was on rgra ten years ago. From what I can recall, stunning was considered more dangerous than cuts even then - so this is an error. I'll pick it up in the monster power rewrite ... ah, that reminds me, someone sent me patches to kick that off, back in the pre-git era. I'll dig them out and commit them in 3.3

[PowerDiver]

The original discussion of the relative dangers of different attack types was on rgra ten years ago. From what I can recall, stunning was considered more dangerous than cuts even then - so this is an error. I'll pick it up in the monster power rewrite ... ah, that reminds me, someone sent me patches to kick that off, back in the pre-git era. I'll dig them out and commit them in 3.3

If you plan a rewrite, shouldn't stun be mostly unrealated to damage? A 1d20 kick wouldn't matter, but a 10d1 kick is deadly. You really ought to compute stun probability from the dice, and also multiple stun attacks are possibly exponentially dangerous rather than additively dangerous.

[Magnate]

If you plan a rewrite, shouldn't stun be mostly unrealated to damage? A 1d20 kick wouldn't matter, but a 10d1 kick is deadly. You really ought to compute stun probability from the dice, and also multiple stun attacks are possibly exponentially dangerous rather than additively dangerous.

Absolutely. I think few if any involved in the original rgra discussion even knew how stunning was calculated.

[Derakon]

Since I'm curious how stuns work, I went and looked it up:

First, if an attack can both stun and cut, then it does a coinflip to see which one it tries to do (i.e. it won't try to both cut and stun on any single attack). Only hit, crush, butt, kick, and punch attacks can stun; of those, only hit can also cut.

Secondly, monster_critical() is called, which returns how strong of a critical hit was performed. Critical hits are 95% of max possible damage, so monsters with Nd1 dice attacks always critical but using larger dice gets you a much more even distribution. Critical hits also have to deal >20 damage (before AC damage reduction) or else make a (damage) in 100 roll. Then the quality of critical increases:

1) if the roll was maxed out (damage = dice * sides),
2) as the damage increases,
3) and there's an indefinite 2% supercharge (every time a 2 in 100 roll is made, the quality increases).

So for example, a monster with a 40d1 melee attack would always get the "maxed roll" bonus, would get a tier-5 damage bonus, and then would get the supercharge bonus if any. Assuming supercharging fails, the monster would get a quality-6 critical.

Quality increases the stun timer by:
1: d5
2: d10+10
3: d20+20
4: d30+30
5: d40+40
6: 100
7 on up: 200; these latter two would require either very lucky supercharge rolls or a critical hit dealing >45 damage, disregarding the maxed-roll bonus.

Note that you become knocked out when your stun timer hits 100, so think carefully before making monsters with Nd1 melee attacks.

For reference, master mystics and grand master mystics both have 20d1 poison hits and 15d1 paralyze hits. If they land both of those in one round (and ignoring their kick attacks, which have d2 dice sizes and thus are much less likely to critical), and the game tries to stun you instead of cutting you, and they don't get any supercharge bonus, they'd get guaranteed quality-4 criticals on the first strike, and a 15% chance at quality-3 criticals on the second.

[Tiburon Silverflame]

So a 10d1 isn't *too* bad...it's still only got a 10% chance.

It seems a rather bizarre approach, tho, to say "X% of max possible == a crit" for monsters. It says that, say, 10d10 attacks basically *never* crit; rolling 95+ on 10d10 is exceptionally unlikely; the mean is 55, the stddev is 9, so we're out past the 4-sigma point.

[Magnate]

So a 10d1 isn't *too* bad...it's still only got a 10% chance.

It seems a rather bizarre approach, tho, to say "X% of max possible == a crit" for monsters. It says that, say, 10d10 attacks basically *never* crit; rolling 95+ on 10d10 is exceptionally unlikely; the mean is 55, the stddev is 9, so we're out past the 4-sigma point.

IMO the whole approach to critical hits needs re-thinking after 20+ years. Both for player crits and monster crits.

[Timo Pietilä]

IMO the whole approach to critical hits needs re-thinking after 20+ years. Both for player crits and monster crits.

For that I strongly emphasis on missile weapons criticals, if that is not yet fixed. Critical hit quality (not the possibility IIRC) is determined by weapon weight, which in case of bolts and arrows is always too low to get the biggest criticals, even that everyone knows that high velocity xbow bolt can go right thru human beings (and most medieval armors and bones).

That is just wrong. You should do a plain calculation based on archer skill+to_hit & shooter multiplier (not the missile multiplier IMO) vs monster AC and leave weight completely out of the picture. That makes heavy Xbows very very effective weapons (maybe Umbar needs tone down with that though. LXbow instead of HXBow?).

[d_m]

For that I strongly emphasis on missile weapons criticals, if that is not yet fixed. Critical hit quality (not the possibility IIRC) is determined by weapon weight, which in case of bolts and arrows is always too low to get the biggest criticals, even that everyone knows that high velocity xbow bolt can go right thru human beings (and most medieval armors and bones).

That is just wrong. You should do a plain calculation based on archer skill+to_hit & shooter multiplier (not the missile multiplier IMO) vs monster AC and leave weight completely out of the picture. That makes heavy Xbows very very effective weapons (maybe Umbar needs tone down with that though. LXbow instead of HXBow?).

I haven't run the numbers, but if this is an issue an obvious fix would be to use launcher weight instead.

That said, I haven't looked at the code doing criticals so it may be the case (as you and Magnate say) that we need a more drastic overhaul.

[Derakon]

So a 10d1 isn't *too* bad...it's still only got a 10% chance.

5% if it's a standard "hit" attack, and even then it'll only increase your stun counter by a marginal amount because the damage is so low.

By contrast, a 10d2 kick has only a 1 in 1024 + 1 in ~2694 chance of stunning (the first is the odds of getting a 20-damage max roll; the second is the odds of a 19-damage roll passing the percentile check). A 1d500 attack would only stun 5% of the time, though when it does stun it'd be an instant knockout.

The current system is quite odd, I'll grant.