Several months ago on Qo’nos First City I was witness to a
discussion as to which space weapons modifier was best: Acc or Dmg. The argument was vociferous, vehement, and
vitriolic, but the positions were supported only with bald assertions, basic
thought experiments, and a lot of name calling.
I think it’s time I got to the bottom of this.
Disclaimer: I don’t usually put disclaimers on these but I
thought that this time it was warranted as my results may be very much limited
to my build and other very similarly built ships. The main cause of concern is the bonus to
critical hit from weapons specialization, the Borg assimilated module, and the
passive bonus from the Romulan reputation system. Other confounding effects would be the
critical bonus from heavy cannons, the critical bonus from antiproton weapons,
whether your captain has the accuracy trait, and probably a few others that
slip my mind. As such, I’m not posting
this under Game Mechanics Revealed.
Premise
While accuracy is of paramount importance when fighting
smaller, faster moving ships, especially other human players with all their
various abilities and equipment that increase defense rating, what if you don’t
PvP? Is accuracy, with its bonus to
critical hit when there is no chance to miss still best? Or is it damage? I consider two experiments: one where I fire
at a stationary target, the other where, in a moving ship I fire at another
ship moving at the same velocity, such that my target’s defense rating is
somewhat higher than 0.
Experiment
The aggressor is armed with 4 rare tetryon dual cannons mk x. At 8 km and 125 weapons power including the Borg Assimilated Module universal console but no other equipment, the aggressor fires at the unshielded, unarmoured target and dps is recorded. The modifiers on the dual cannons are [acc]x2, [dmg]x2, [crth]x2, and [crtd]x2. The aggressor captain has full ranks of Starship Weapons Training, Starship Energy Weapons, Starship Targeting Systems, and Starship Energy Weapons Specialization, along with the bonus 3% critical hit chance from the Romulan Reputation system.
Results
Run

[acc]x2

[dmg]x2

[crth]x2

[crtd]x2

1

1606.4

1700.7

1615.2

1799.4

2

1711.1

1652.6

1730.9

1555.1

3

1743

1683.7

1588.1

1717.6

4

1676.3

1709.3

1736

1684.5

5

1625

1594

1635.6

1663.2

6

1681.3

1585.2

1634.4

1792.3

7

1673.6

1608

1710.4

1572.4

8

1595.6

1622.8

1688.9

1668.9

9

1663.5

1595.3

1609.1

1776.9

10

1581.3

1600.2

1614.6

1710.1

Average

1655.71

1635.18

1656.32

1694.04

Analysis
We see here that the average dps of each of the modifier types
is quite similar, but there are differences.
Surprisingly, at least to me, the [dmg]x2 came in last while the [crtd]x2
came in first. But are the differences
significant? Let’s look at the highest
and the lowest average dps modifiers, the [crtd]x2 and the [dmg]x2 and do some
fancypants statistics on them.
Null hypothesis, Ho, is that the means of dps are the same,
or u1 – u2 = 0. Therefore the alternate
hypothesis, Ha, is that the two means are not the same, or u1 – u2 ≠ 0. We will compare the outcome with Student’s t
distribution. There are 10 data points
per sample, implying 10 + 10 2 = 18 degrees of freedom and we are using a twotailed
test, Ta(18) = 2.101 at the 95% confidence interval. Therefore we will reject the null hypothesis
if t > 2.101.
n

10

10

average

1635.18

1694.04

sample standard deviation

52.10499

96.32923

sample variance

2714.93

9279.32

Pooled s^2

4145.316


s

64.38413


t

2.044214

Since t<2.101 the null hypothesis cannot be rejected at
the 95% confidence interval. But holy cow, is it close (and I might not even be
using the right percentage or number of tails for this application).
Conclusion
By the rigorous standards of statistical analysis, there is
no difference in hitting power between the four standard space energy weapon
modifier types while shooting at a stationary target. Since not all targets are stationary, [acc]
is best since it will help you in those occasional situations where you have to
shoot something that is moving quickly.
It is by far the best in PvP when you have slippery escorts and ships
with Aegis set. Consider me a proponent
of [acc] as the best modifier.
Since none of the other modifiers was noticeably better than
[acc] even when stationary, the second half of the experimental program is
cancelled.
Warning: Completely irresponsible conclusion jumping ahead
I find it interesting that the dps of [acc] and [crth] was
exactly the same. I guess that means
that with my accuracy rating, the bonus to critical hit percentage was
identical to the bonus from [crth].
The difference in average dps as calculated by my combat log
parser between the [dmg]x2 and [crtd]x2 was 3.6%. That’s rather a lot, really. Sure, it wasn’t quite statistically
significant at the level of sampling I performed, but it’s right there. I’ve made sweeping generalizations about
other weapons setups on less. But what
is going on here?
I’ve done some number crunching before about the modifiers,
always just from reading the tooltips and doing quick calculations, and each
time [dmg] came out far ahead of [crth] and [crtd]. Now once I actually do the experiment it
seems that [dmg] is lagging behind absolutely everything else. It makes me think that I wasted quite a bit
of ec on that modifier when cheaper ones existed. Of course, I did all that math back before I had
the Romulan +3% to critical hits, and it seems logical that as a higher
percentage of my hits are critical, the more important critical damage becomes. If both critical hit percentage and critical
damage bonus are very low, then of course [dmg] is more effective. So to follow this train of thought to its
conclusion, [dmg] would be most important at low levels as you level up, but
once you get things like the Borg console and whatever else that makes your
criticals more frequent, [crtd] takes over.
test
ReplyDeleteDamage and critical. Which one is the best depends on the ratio you are currently at. For basics here is the average Hit formula:
ReplyDeleteDamage * (1 + (Critical Multiplier * Critical Chance)) = Average Hit
So if we have a weapon with 1000 damage, 5% crit chance, and 100% multiplier we get
1000 * (1 + (1.00 * .05))
1000 * (1 + (.05)
1000 * 1.05
1050 average DPV
In this situation if we could choose between say 20% multiplier, 2% chance, or +10 damage which would be best?
20% multiplier = 1060
2% chance = 1070
20 damage = 1060
So it all depends on the particular stats you start with. But this much is always true
1) ACC will always win if you would miss even a single shot
2) ACC will still win if your multiplier and chance are at an even ratio
3) Otherwise keep multiplier and chance at a 10:1 ratio, if that ratio is not close that use whatever will get it closer
Typically Dmg doesn't add much damage. I think even on DHCs it only adds like 10.