YAGBU: Slaying missiles and critical hits
Malte Helmert
more experiments with WADOMF and g16pre2.
s
p
o
i
l
e
r
s
p
a
c
e
s
p
o
i
l
e
r
s
p
a
c
e
s
p
o
i
l
e
r
s
p
a
c
e
All experiments consisted of fights against balors.
The numbers listed are the points of damage scored,
taking a PV of 40 into account. PV and hit points
of the opponent were measured with a blessed stethoscope.
The PC had the Find Weakness skill at level 100.
Experiment 1: Regular melee weapon
==================================
blessed scalpel
damage reported by "W": 1d3+155
expected base damage: 156-158
expected damage against demons (+50%): 234-237
observed damage: 237, 237, 235, 237, 234, 235
This supports the "+50%" hypothesis.
Experiment 2: Slaying melee weapon
==================================
holy avenger "Justifier"
damage reported by "W": 3d5+85
expected base damage: 86-100
expected damage against demons (+50%),
*not* taking into account slaying: 129-150
Balor was hit with uncursed and blessed "Justifier".
Each hit resulted in a "with full force" hit message.
However, the damage always was in the 86-100 (uncursed)
or 129-150 (blessed) interval. I.e., there seemed to be
no extra slaying damage. (?)
This is weird. Maybe it is an oddity in g16pre2, or it has to do with
the fact that the Justifier was hacked. Great identify showed the demon
slaying ability though. Some followup research on slaying melee weapons
seems to make sense.
Experiment 3: Arrows
====================
A first exploratory experiment showed that blessed arrows and a
blessed bow lead to 50% more damage against demons than uncursed
arrows and a blessed bow. Occasional critical hits seemed to deal much
more damage, observing double, triple and quintuple damage in the few
critical hits counted.
All further experiments were conducted arrows where "M" showed a base
damage of 1d6+16. Thus, the expected damage range for uncursed arrows
was 17-22 and the expected damage range for blessed arrows was 25-33.
Assuming that blessed damage is calculated by rolling uncursed damage,
then multiplying by 1.5 and rounding down, it was expected that only
six of the values in the interval 25-33 actually occur, namely 25, 27,
28, 30, 31, and 33. This was found to be true.
Experiment 3A: Critical hits
============================
Blessed arrows without slaying capabilities were used. Because of the
PV of 40, only critical hits could deal any damage. All critical hits
*did* deal damage, and the observed values fit a nice pattern: They
all consisted of some of the possible values for blessed arrows
mentioned above, multiplied by 2, 3, 4 or 5.
Possible values in detail:
x2: 50, 54, 56, 60, 62, 66
x3: 75, 81, 84, 90, 93, 99
x4: 100, 108, 112, 120, 124, 132
x5: 125, 135, 140, 150, 155, 165
Observed values:
x2: 50, 56, 60, 60, 62, 62, 62, 62, 62, 66
x3: 75, 75, 81, 81, 84, 84, 84, 84, 84, 90, 90, 93, 93, 93, 99, 99
x4: 100, 100, 100, 108, 112, 112, 112, 112, 120, 132, 132
x5: 125, 125, 125, 135, 135, 140, 140, 140, 140, 150, 150, 155, 165
Experiment 3B: demon slaying
============================
Same as 3A, except the blessed arrows were arrows of demon slaying. In
this case, all hits were critical (resulted in the "exactly hit"
message). The only difference to the data observed above was that the
multipliers 3, 4, 5, 6 were observed rather than 2, 3, 4, 5.
This lead to one problem in reporting the results: The value 150
can either be interpreted as 30*5 or as 25*6. I have split the various
occurrences of 150 between those two groups in a random fashion.
All following experiments are minor variations of this.
Observed values:
x3: 75, 75, 81, 81, 81, 84, 90, 93, 93
x4: 100, 108, 108, 108, 108, 112, 112, 120, 120, 124, 132, 132, 132
x5: 125, 135, 140, 140, 140, 150, 150, 150, 155, 155, 155, 155, 165,
165, 165, 165
x6: 150, 150, 162, 162, 162, 168, 198, 186, 186, 186, 186, 162
Experiment 3C: uncursed bow
===========================
Same as 3B, but with an uncursed bow. The multiplier x6 was somewhat
rarer, but this is not statistically convincing. If blessing the bow
helps, it doesn't help a lot.
Observed values:
x3: 75, 75, 75, 75, 75, 75, 75, 81, 84, 84, 84, 90, 93, 93, 93, 99, 99
x4: 100, 100, 100, 100, 100, 108, 108, 108, 112, 120, 124, 132, 132
x5: 125, 125, 135, 135, 135, 140, 140, 145, 150, 150, 155, 165, 165,
165
x6: 150, 150, 180, 186, 186, 198
Experiment 3D: humanoid slaying
===============================
Same as 3B, but with humanoid slaying arrows instead of demon slaying
arrows. No qualitative differences in the observations.
Observed values:
x3: 81, 81, 81, 84, 90, 90, 93, 93, 93
x4: 100, 108, 108, 124, 124, 124, 132, 132, 132, 132
x5: 125, 125, 135, 135, 140, 140, 140, 150, 150, 150, 155, 160, 165,
165
x6: 150, 150, 150, 162, 162, 162, 168, 168, 180, 180, 180, 180, 186,
186, 198, 198, 198
Experiment 3E: uncursed arrows
==============================
Same as 3B, but the arrows were uncursed. The hypothesis was that the
same multipliers (x3...x6) would apply, but without the +50%
modification for fighting demons with blessed weapons. The
observations supported the hypothesis.
Observed values:
x3: 51, 51, 54, 54, 60, 60, 63, 63, 63, 66, 66
x4: 72, 72, 76, 76, 80, 84, 84, 84, 84, 88, 88, 88
x5: 85, 90, 95, 95, 95, 95, 100, 100, 100, 105, 105, 105, 105,
105, 110, 110, 110
x6: 102, 102, 102, 102, 114, 120, 126, 126, 132, 132
Experiment 3F: without Find Weakness
====================================
Same as 3B, but without the Find Weakness skill.
This was done to check whether critical hits from slaying weapons and
critical hits from the Find Weakness skill are cumulative in some way
or other (e.g., the x6 multiplier might only happen on successful Find
Weakness skill checks).
This was a smaller experiment because it was quickly seen that there
seems to be no influence of the Find Weakness skill for slaying
arrows.
Observed values:
x3: 90, 93, 99
x4: 108, 120
x5: 125, 125, 135, 135, 150, 155, 155, 155, 160, 165
x6: 180, 180, 198, 198, 198
Interpretation of the missile experiments:
Missile damage seems to be calculated as follows:
base_damage = roll dice reported by "M"
if fighting demons or undead and projectile is blessed:
Multiply base_damage by 1.5, rounding down.
if the projectile has an appropriate slaying property:
Multiply base_damage by 3, 4, 5, or 6 (choose randomly).
else if the PC manages to score a random critical hit:
Multiply base_damage by 2, 3, 4, or 5 (choose randomly).
if base_damage > PV:
Opponent loses (base_damage - PV) hit points
Some of this hasn't been explored very deeply. For example, dragon
slaying ammunition might behave differently to demon slaying ammunition.
Still, this is better than no information.
Malte
ToGu
"Malte Helmert" <hel...@informatik.uni-freiburg.de> wrote in message news:bl6rbp$97k$01$1...@news.t-online.com...
> holy avenger "Justifier"
> This is weird. Maybe it is an oddity in g16pre2, or it has to do with
> the fact that the Justifier was hacked.
Just a quick comment on this: Justifier is special and shouldn't be used
for those kinds of tests. And a side note about hacking: if ADOM actually
knows that something is wrong (run out of luck), then the internal
calculations are severely altered, making the results less useful. Other than
that, it depends on how the hack is done. If the game had simply been
hacked to generate the item "normally" then there wouldn't be any difference,
but if the hex data has been modified directly, some properties of the old
item might stay behind and mess up the test data.
- ToGu
Malte Helmert
Interesting. Special in what way?
Do other artifacts have similar properties?
Unfortunately, there seem to be no non-artifact slayers in g16pre2,
except rabbit knives (limit usefulness) and maces of disruption
(stethoscopes useless vs. undead).
> And a side note about hacking: if ADOM actually
> knows that something is wrong (run out of luck), then the internal
> calculations are severely altered, making the results less useful.
Running out of luck is something which is hard not to notice. :-)
> Other than that, it depends on how the hack is done.
In this case, simply generating with WADOMF (with modified to-hit and
damage bonus values).
> If the game had simply been
> hacked to generate the item "normally" then there wouldn't be any difference,
> but if the hex data has been modified directly, some properties of the old
> item might stay behind and mess up the test data.
I think WADOMF does a good job at rewriting the whole item structure.
I've never had problems with WADOMF-generated items, but then I've
rarely generated artifacts.
Malte
ToGu
"Malte Helmert" <hel...@informatik.uni-freiburg.de> wrote in message news:bl7ash$ks9$00$1...@news.t-online.com...
Special bonus for paladins, special penalty for everyone else IIRC.
Just one of those things I notice randomly, tell in #adom (ircnet) and
forget about :P
> Unfortunately, there seem to be no non-artifact slayers in g16pre2,
> except rabbit knives (limit usefulness) and maces of disruption
> (stethoscopes useless vs. undead).
Oh, I only meant Justifier, not artifacts in general. Haven't I mentioned
how much I enjoy ADOMs hardcoded approach to features? ;)
- ToGu