Miscellaneous Maths/Testing II

Eld

Sena said:

I've been playing with the numbers a bit, can't manage to make them add up nicely. Seeing the damage at 12 strength would likely help a lot.

Yeah, I was thinking the same thing. Just waiting till I can get a falcon to thwap and another blademaster to vitiate me.

Sena

I can provide the falcon, just send me a message when you're ready.

Sena

Testing with 12 strength and with a level 1 band helped a lot. Here's how it seems to work.

Without any strength modifier, drawslash damage is 427+23x, with x being, respectively, -3, -1, 0, 1, 2, 3 for Thyr, Mir, Unstanced, Sanya, Doya, Arash. A level 1 band adds 1, a level 3 band adds 2 (maybe a bug? I'd like to see a level 2 band). So, for Thyr with no band, that's 427+(-3*23)=358. Unstanced with a level 3 band is 427+(2*23)=473.

I don't know where the 23 comes from. If it's meant to be 5%, that could mean that Sanya is the base damage (450/20=22.5, rounded up to 23), and unstanced is reduced.

Rispok

Daeir said:

@Sena Really? I was always told staffcast was 1100 on denizens give or take 10% for resistances. So Blademasters are in effect, the strongest bashers in the game?

No. Monk is. Three chances for critical hits.

Sena

Staffcast isn't anywhere near 1100, even at 25 int with a level 3 collar. Denizens also don't have resistances.

16 int staffcast is 564 against denizens, 17 int is 589, and it has a long eq (4 seconds before quick-witted and diadem). Being better than that doesn't make them the strongest bashers in the game, it just means they have good offence. Mage is still far better defensively, serpent with a level 3 lash does similar damage to artied staffcast, and is tankier than blademaster, monk easily beats blademaster in both offence and defence, alchemists also do far more than staffcast and are tankier than serpents, plus they have a critical bonus.

Eld

Ok, did a little more testing. First off, I think that what @Sena mentions as a possible bug is in fact such. After testing just a couple numbers with Rom's level 3 band, I have for 12 strength:
                     no band   lvl 1   mine    Rom's
Thyr               358         381    404      427
Mir                 404         427    450      473
Unstanced     427         450   473
Sanya            450         473    496
Doya              473        496     519
Arash             496        519     542

I didn't put Rom through the full rigamarole, but based on those two, and the pattern Sena pointed out, I think that my band is indeed bugged and doing the damage of a level 2 instead of a level 3.

For unstanced and sanya drawslash at various strengths with no band, I have:
str:    unstanced  sanya
11:    399             420
12:    427             450
13:    454            
14:    481             507
15:    505             533
16:    529             557
17:    550             580
18:    570             600

I'm guessing the missing number for sanya at 13 str is 480. Haven't tried particularly hard yet to come up with a sensible formula for the diminishing returns yet.

Sena

Eld said:

Haven't tried particularly hard yet to come up with a sensible formula for the diminishing returns yet.

I haven't either, but it looks like a good estimation (far from perfect, but within a few points for 6-18 strength) is BaseDamage*(1+(12-Str)^0.9/13*0.85). Basically, the normal non-diminishing formula (BaseDamage*(1+(12-Str)/13*0.85)) but with the strength difference (from 12) reduced.

Sena

Also, level 3 bow damage seems to be 580+MaxHealth*19%.

Sylvance

Anyone have the ideal speed for a Bard rapier, to chase voice balance?

Sena

If you know how fast voice balance is, then I can tell you what speed would match it.

Sylvance

Sena said:

If you know how fast voice balance is, then I can tell you what speed would match it.

Thanks - out of interest how precise does it need to be (i.e. how many tests should I do to account for latency etc)?

Sena

With a decent connection, 10 or so samples should be enough.

Mizik

Voice 3s

Sena

Then it takes 200 speed to match it without nimble, 190-ish (185 might be enough) with nimble. Assuming trans swashbuckling.

Iocun

Sena said:

Eld said:

Haven't tried particularly hard yet to come up with a sensible formula for the diminishing returns yet.

I haven't either, but it looks like a good estimation (far from perfect, but within a few points for 6-18 strength) is BaseDamage*(1+(12-Str)^0.9/13*0.85). Basically, the normal non-diminishing formula (BaseDamage*(1+(12-Str)/13*0.85)) but with the strength difference (from 12) reduced.

Hrm, sure you didn't mix up a plus with a minus in those two formulas?

(12-Str)/13*0.85

is going to be a positive number for any strength lower than 12 and a negative one for strengths above 12.

Shouldn't it rather be:

(BaseDamage*(1-(12-Str)/13*0.85))

?

Or alternatively (Str-12) instead of (12-Str).

Tvistor

Whenever my heart says "Tvistor, you should go infernal, you'd rock it", I read these damage formulas and then feel better.

And confused.

At least doublestab has a static speed.

Eld

Iocun said:

Sena said:

Eld said:

Haven't tried particularly hard yet to come up with a sensible formula for the diminishing returns yet.

I haven't either, but it looks like a good estimation (far from perfect, but within a few points for 6-18 strength) is BaseDamage*(1+(12-Str)^0.9/13*0.85). Basically, the normal non-diminishing formula (BaseDamage*(1+(12-Str)/13*0.85)) but with the strength difference (from 12) reduced.

Hrm, sure you didn't mix up a plus with a minus in those two formulas?

(12-Str)/13*0.85

is going to be a positive number for any strength lower than 12 and a negative one for strengths above 12.

Shouldn't it rather be:

(BaseDamage*(1-(12-Str)/13*0.85))

?

Or alternatively (Str-12) instead of (12-Str).

There's also the problem that (12-str)^.9 is complex for str>12, in general.

Iocun

Tvistor said:

Whenever my heart says "Tvistor, you should go infernal, you'd rock it", I read these damage formulas and then feel better.

And confused.

At least doublestab has a static speed.

You hunt with doublestab?

No wonder you're still at Minia levels.

Zeon

songblessed rapiers have a bonus to all stats beyond what you see on weapon probe. So you're going to want something a little slower

Vaehl

But iirc they removed the entire speed bonus. Not 100% sure.

Zeon

wouldn't be surprised 

Sena

Iocun said:

Hrm, sure you didn't mix up a plus with a minus in those two formulas?

You're right, I meant Str-12, not 12-Str.

Eld said:

There's also the problem that (12-str)^.9 is complex for str>12, in general.

That's an even more embarrassing mistake.

So, an estimation of the strength multiplier is |(Str-12)|^0.9/13*0.85, added to the base damage for >12 strength and subtracted for <12. I think that should work.

Sena

As for songblessed rapier speed, from announce 3558:

Songblessed rapiers no longer receive a 10% speed boost.

Eld

Sena said:

Iocun said:

Hrm, sure you didn't mix up a plus with a minus in those two formulas?

You're right, I meant Str-12, not 12-Str.

Eld said:

There's also the problem that (12-str)^.9 is complex for str>12, in general.

That's an even more embarrassing mistake.

So, an estimation of the strength multiplier is |(Str-12)|^0.9/13*0.85, added to the base damage for >12 strength and subtracted for <12. I think that should work.

Base*(1+(str-12)*.85/13)-(str-13)^2, for str>=13, seems to work pretty well for most of the numbers I have, as well. Neither version is perfect.

Eld

To go along with the drawslash damage tests I've been doing, I decided to go ahead and check balance times, as well. Did a few hundred drawslashes in each stance with an unbanded sword, and a couple stances with my band, which is supposed to be level 3 but I'm pretty sure is effectively level 2. Lagging too much to get decent numbers for the other stances at the moment, but might continue later. This is all with nimble.
Looks to be pretty consistent with a base balance time (unstanced, unbanded, without nimble) of 2.5s, 10% reduction for Thyr, 5% for Arash, 20% longer balance for Mir and Doya, and Sanya same as unstanced. That would give, for example, for a level 2 band in Thyr with nimble, 2.5*.9*.9*.9 = 1.8225 s, pretty close to the ~1.86 I see.


band stance   trials  mean   std dev
0       neutral  400     2.228   0.091
0       Thyr      338     2.033   0.079
0       Mir        605     2.724   0.100
0       Sanya   600     2.241   0.080
0       Doya     600     2.752   0.084
0       Arash    604     2.129   0.086

2       neutral  624     2.036   0.080
2       Thyr      400     1.859   0.083
2       Mir        296     2.536   0.094

Aliath

Sena said:

Rangor said:

Could you check if the dmg stat/runes on bows have any impact on speed and damage?

Tested that before, runes and augmentations don't do anything.

Did you just look at the direct damage or on the bleeding as well, just a test I did a while back seemed to indicate a higher damage stat did higher bleeding

Sena

I only looked at speed and direct damage. I'll try testing bleeding later.

Achilles

@eld definitely have them check. My bow was not upgraded correctly a couple of times until I pointed it out.

Eld

Achilles said:

@eld definitely have them check. My bow was not upgraded correctly a couple of times until I pointed it out.

Yeah, I bugged it, just waiting for a reply.

Kafziel

Songblessed rapiers used weaponry for speed last time I tested it. Probably a bug.

edit: I did this a dozen or so times and they're all the same

Lightning-quick, you jab a black bear with an ornate steel rapier.
23:55:08:127
---------------RECOVERED BALANCE---------------
23:55:11:086

You viciously jab an ornate steel rapier into a black bear.
23:55:11:813
---------------RECOVERED BALANCE---------------
23:55:14:795


Gifted weaponry, trans swashbuckling, 215 speed songblessed rapier, nimble.