Spiral Knights - Let's Experiment: The swing speed of swords.

Let's Experiment: The swing speed of swords.

30 replies [Last post]

Fri, 04/22/2011 - 21:46

Sirrocco

Legacy Username

This thread is spawn off of the "Attack Speed Hard Cap Limit?" thread - an attempt to get decent, experimental evidence on the effects of attack speed boosts on various swords, and the points where attack speed boosts are no longer applicable. Potential questions to be answered empirically...

- Are attack speed boosts on gear level-dependent?
- Which blades benefit more or less from attack speed boosts?
- Is there, in fact, a hard cap limit (or an effective hard cap limit) on sword attack speed other than that imposed by the generic limit on boosting anything (ie, +6)
- Are swing speed boosts subject to diminishing returns? Ie, is the jump from +0 to +3 more significant, less significant, or roughly the same as the jump from +3 to +6?

The basic technique is as follows: start a 60 second timer, while in a safe location that permits sword-swings. While the timer is running, spam the basic sword-swing combo of the sword in question as fast as you can, counting the swings. Record sword, number of swings, speed, and location. Worth noting that this is not a terribly precise method, for a variety of reasons - mostly observer error of one form or another. There's a fair amount of fuzz factor involved. Fortunately, we should be able to derive a decent amount of useful information even without a whole lot of precision. Unfortunately, I can't test all that much myself with the gear that I have, but let's see what I *can* come up with.

Calibur, no speed boosts, Moorcroft Manor: 94 or 95 in 60 seconds
Calibur, swiftstrike buckler only (+3, assuming full strength), Moorcroft Manor: 103 in 60 seconds
Calibur, swiftstrike buckler only (+3, assuming full strength), initial waiting area: 104 in 60 seconds

Further testing indicates that Moorcroft Manor and the initial waiting area do count as different tiers for the purposes of stat controls on gear. (Specifically, my level 10 wolver gear gives me 4 additional HP in Moorcroft, and only 2 in the initial waiting area). The one additional swing is almost certainly a result of the lack of precision I was discussing. The "94 or 95" was because the last swing was right at the end of the 60 seconds, and I couldn't quite tell which side of the line it was on. All tests were done with max heat gear, as it turns out, but I have no reason to believe that that had any effect on swing speed.

So - it's pretty clear that speed boosts on gear are not affected by floor (not surprising, but it certainly does simplify further testing) and that picking up a swiftstrike in tier 2 has a relatively small but certainly significant impact on swing speed for Calibur-users (should be no shocker to anyone).

Anyone else who wants to add some experimental results of their own is welcome to, and I'll be adding more myself as I acquire the gear to do it with. If nothing else, I do intend to get something in the cutter line some time in the not too distant future, and I expect that the results of that with and without swiftstrike will be somewhat illuminating.

Fri, 04/22/2011 - 23:22

RamuneDrink

Legacy Username

Bumping...though I can't add

Bumping...though I can't add any info.

Fri, 04/22/2011 - 23:53

Pupu

Legacy Username

I would help but I don't have the proper equip.
Looking forward to results~

Sat, 04/23/2011 - 00:22

Stardrinker

Legacy Username

Avenger data

Sword: Avenger (lvl10 but shouldn't matter)

No attack speed gear
58 in 60s trial 1
58 in 60s trial 2

1 piece vog cub (Medium)
62 in 60s trial 1
63 in 60s trial 2

Swiftstrike Buckler (High)
65 in 60s trial 1
65 in 60s trial 2

Full vog cub (Very High)
68 in 60s trial 1
68 in 60s trial 2

Swiftstrike Buckler + 1 piece vog cub (Ultra)
70 in 60s trial 1
70 in 60s trial 2

Swiftstrike Buckler + Full vog cub (Maximum)
72 in 60s trial 1
72 in 60s trial 2

Medium attack speed increase: ~7% - 8.6%
High attack speed increase: ~12%
Very High attack speed increase: ~17.25%
Ultra attack speed increase: ~20%
Maximum attack speed increase: ~24%

If we assume: Low=1, Med=2, High=3, Very High=4, Ultra=5, Maximum=6

then:
2=7 - 8.6 --> 1~=3.5 - 4.3
3=12 -------> 1~=4
4=17.25 ---> 1~=4.3
5=20 -------> 1~=4
6=24 -------> 1~=4

So it looks like 4% increase for each 'level' of attack speed.

Sat, 04/23/2011 - 00:33

Kryssa

Legacy Username

Vile Striker Heat lvl 10

Vile Striker, no speed boosts, Depth 23 Terminal - 119 swings in 20 secs
Vile Striker, no speed boosts, Depth 23 Terminal - 121 swings in 20 secs
Vile Striker, no speed boosts, Depth 23 Terminal - 120 swings in 20 secs

So on average, 120 swings per minute. It may be worth noting that the Cutter series gains a "bonus" swing for about 50% damage (of a normal swing) with each normal swing.

I don't have a Swiftstrike and don't think I'll get one anytime soon. My gear is full heat Ash Tail Set, Perpperbox and Wise Owlite Shield.

Sat, 04/23/2011 - 00:38

meowreka

Legacy Username

Let's compare this to Attack

Let's compare this to Attack Damage increases, so we can finally settle the argument.

Vog vs Skolver

Sat, 04/23/2011 - 02:04

Stardrinker

Legacy Username

Let's compare this to Attack

Let's compare this to Attack Damage increases, so we can finally settle the argument.
Vog vs Skolver

Can't really directly compare them because attack speed and increased damage both have pros and cons vs each other.

Attack speed:
Increases damage provided you keep swinging and stay on the target.
Reduces delay caused by animations so you are able to recover faster, meaning you can avoid potential damage.
Doesn't really increase damage of charge attacks.

In terms of just pure damage, Increased damage is better than Increased attack speed (IAS), but you don't get the other defensive benefits of IAS.

Also, I just had my data on no dmg bonus vs low dmg bonus vs med dmg bonus. on Depth 1, 9 and 19 eaten by the forum, guess that teaches me for not writing it down on a text editor to start with. Anyway, gist of it was:

Sword:Ascended Calibur
Depth 1
First atk, Last atk, Charge atk
~+5.5%-6%, ~+5.5%-6%, ~+4.5%-5% for Low, roughly double that for Med

Depth 9
First atk, Last atk, Charge atk
~+5.5%-6%, ~+5.5%-6%, ~+4.5%-5% for Low, roughly double that for Med

Depth 19
First atk, Last atk, Charge atk
~+6%-6.5%, ~+6%-6.5%, ~+5%-5.5% for Low, roughly double that for Med

This is just remembering data off the top of my head, that was eaten by the forum, and I can't be bothered to do more data gathering right now, but I'm not so sure if it's strictly % damage increase; there looks to be another modifier there. The % increases at depth 19 were noticeably higher than depth 9 or 1 (by noticeably I mean 1-2% higher total difference).

What this does mean, however, is that for pure damage increase, +damage > +attack speed. I'd still prefer my Vog Cub set because the damage increase is very slight but the survivability of IAS is hard to replace.

Sat, 04/23/2011 - 06:26

Sirrocco

Legacy Username

So Stardrinker's data seems

So Stardrinker's data seems to indicate pretty strongly that you don't have to worry about diminished returns with IAS - that the last point is just as useful as the first point. That's good to know. What would be really handy now is for someone with a full speed set (vog cup plus swiftstrike) to run a similar set of tests on something in the cutter line - both to see if the IAS has as much of an effect on the fast weapons as on the slow weapons (it didn't on the gun side, and the avenger and calibur tests *suggest* that it wouldn't here) and to get (semi)hard evidence one way or the other on the hard cap.

Sat, 04/23/2011 - 09:52

Shani

Legacy Username

Another for the pro side of

Another for the pro side of attack speed: for weapons like the Vile Striker line that inflict status ailments, if the ailment works on a straight-up percentage rate per hit, attack speed increase will also render that more effective by allowing you to get in more hits.

Mon, 04/25/2011 - 06:33

Pupu

Legacy Username

Bump for info on spur/cutter.

Wed, 04/27/2011 - 07:04

#10

Sirrocco

Legacy Username

well, I have some information

well, I have some information on strikers, anyway. I have a swiftstrike, a standard striker, and a striker UV (IAS, low(+1)) to work with. Results of the one-minute test are as follows:

- striker+0: 115
- striker+1: 118
- striker+2:
- striker+3: 128
- striker+4: 130

I also noted during the process of acquiring data that getting data for the striker (and presumably the rest of the cutter line) is trickier than for other lines, partially because it goes so fast, partially because the second blade is distracting/confusing, and so forth. I believe that this data is somewhat less accurate than the Calibur data I posted (the jump from +1 to +3 being over three times as much as the jump from +0 to +1, for example). Given the apparent inaccuracy, I can't say that it provides any data one way or the other on speed capping - that *might* be speed cap starting to creep in at the +4 level, or it might be more data wonkiness - though it does appear that speed boosts at least do *something* up through +4.

It does look like speed boosts don't do quite as much for the cutter line as they do for the slower blades - this fits in with the behaviors we've seen from Calibur and Avenger. At the same time, that's arguably deceptive. I haven't yet really gotten into the swing of things with proper use of the striker, but as far as I can tell, optimal play involves a lot of dashing through groups, trusting your speed, and using the basic attack a few times (but not the entire combo) whereas optimal use of the Calibur line involves a lot of kiting to build up charge attacks, running in to unleash+shield, and then pulling back out to charge up again. It may well be that IAS does more for Calibur as far as the numbers are concerned, but rather less as far as practical impact on effectiveness of play.

It would still be helpful to have numbers for cutter speed +4 through +6, if someone else could field them, and possibly +1 also, as a baseline.

Wed, 04/27/2011 - 11:45

#11

Tive

Dont forget that +DMG makes

Dont forget that +DMG makes you deliver huger damage packets.
if you have to reposition after every 2 strikes you might as well want to finish it with 4 assaults instead of 5. (and dont forget charge attacks)

Wed, 04/27/2011 - 12:18

#12

FlashbackJon

Legacy Username

I approve of this thread.

Fascinating info! I don't have a wealth of equipment at my disposal but I'll give it a shot with the troika and sealed sword when I have time.

Wed, 04/27/2011 - 13:42

#13

Sirrocco

Legacy Username

Actually, i'm less than

Actually, i'm less than impressed with the charge attack of the cutter line. Of course, that may just be because I don't know how to use it yet. If there's anyone out there who uses cutters/strikers/etc regularly and has a valid tactical use for the charge attack, I'd be interested to hear it.

Wed, 04/27/2011 - 14:06

#14

Shani

Legacy Username

The only use I found for it

The only use I found for it was to take out Gunpuppies when soloing T2. A lesser use for it is to commit suicide by accidentally charging it instead of my bombs in T3.

Tue, 05/24/2011 - 23:25

#15

Evokie

pretend you have a sword with

pretend you have a sword with very high attack speed, and armor that provide very high attack speed as well. does that mean your weapon has plus 8 attack speed? so you have maximum (6) attack speed bonus + another (2)?

Tue, 05/24/2011 - 23:36

#16

Luden

Legacy Username

right now im using one piece

right now im using one piece of mad bomber, one piece of skolver + ctr leviathan blade

charge does around 3*2.5 times the normal hit damage per hit, so even with a 5% increase from skolver (not sure about the exact figures), a charge attack would multiply that by 7.5 times, whereas IAS would only make you finish your charge attack faster (not the charging, but the actual attack animation on charge attack)

the useful info for those who only cares about IAS is that IAS actually delivers your charge attack faster.

Tue, 05/24/2011 - 23:40

#17

Evokie

whats IAS and... btw i have a Cold Iron vanquisher. charge attack only does about 1.8 times the normal hit damage per hit.

Wed, 05/25/2011 - 06:22

#18

Droganis

Legacy Username

@Evokie

No, IAS (increased attack speed) is capped at 6. If you had that set up, you would just be wasting 2 points of IAS. It does not effect charge time or damage.

Wed, 05/25/2011 - 06:30

#19

Bigindian

Another interesting measure

Another interesting measure would how many seconds it takes to do a full sequence of moves. e.g. Striker's 10 swings, or avenger's 2 wings?

Wed, 05/25/2011 - 18:49

#20

Evokie

if my normal attack damge is

if my normal attack damge is already maxed out on my sword and i have a damage bonus on my armor will i do extra damage dispite my sword having max damage already?

Wed, 05/25/2011 - 20:07

#21

culture

Legacy Username

Nah, the armor's damage bonus

Nah, the armor's damage bonus is applied to the sword. So if the sword is already maxed out then that's it.

Thu, 05/26/2011 - 19:51

#22

Evokie

today i found a friend with a

today i found a friend with a CIV with no UV and i compared my CIV with very high attack speed. Everything is exactly the same. swinging speed, dodging speed everything. its ridiculous. attack speed UV is absolutely useless for generally faster swords.

Wed, 07/27/2011 - 10:50

#23

Seiei

It may seem as there's no

It may seem as there's no difference, but to someone with fast eyes their is significant. My style of play consist in swinging like crazy till i see any of d monster about to attack me then blocking. The cutters ASI is very useful. Of cours with caliber sword its a little more noticeable.

Wed, 07/27/2011 - 13:11

#24

Thrull

Legacy Username

Is anyone testing swing speed

Is anyone testing swing speed by factoring in "Shield Canceling" tricks? Spamming swings might not be the most representative of tests. Most players will shield cancel to avoid a chunk of the swing animation, especially with bigger swords.

I did some frame-by-frame animation analysis and tentatively concluded that swing speed modifiers have an amplified effect when shield canceling (i.e. more than would be expected with a flat percentage). But I don't have the numbers now. Might be interesting to test further, although it's much harder to reproduce consistent shield-cancellations with lag and what-not...

Wed, 07/27/2011 - 13:17

#25

Bigfootm

For vig set, V high plus V

For vig set, V high plus V high equals more than MAX does itnot?

Wed, 07/27/2011 - 13:34

#26

Perp

Legacy Username

Yeah, a UV above medium is a waste. Though if you have one, you could use one piece vog and one piece skolver.

Wed, 07/27/2011 - 13:44

#27

Dirt

Legacy Username

This thread is interesting.

This thread is interesting. Maybe it should be moved to the Arsenal or Wiki forums?

Wed, 08/03/2011 - 18:19

#28

Darbish

Using Gran Faust, the time it

Using Gran Faust, the time it took to traverse from the bottom of the Training Area Help Desk until the time the screen starts to fade to black when you exit the area:

without AIS: 28.000 sec
AIS (low): 26.822 sec

It's about a 4% difference in speed which is consistent with Stardrinker's results. If I decide to waste energy to craft some other ais gear I'll update this.

Sun, 08/28/2011 - 20:13

#29

Nievem

With an autogun I did some

With an autogun I did some quick video analysis, it seemed to fit 16% ASI from Swiftstrike reduced the firing animation, not the reloading, and also reduced the animation when shield cancelling, but not the shielding animation. Just the firing animation until the point the shield can cancel. Of course it didn't have to be exactly 16% but there was a reduction. I suppose its cool because the main gripe with the gun is that you are stationary while firing, and this mod would allow you to reduce that time, but in the end it is just very few frames you save, so lets say 20ms tops.

Wed, 10/12/2011 - 02:07

#30

Xiaoa

Tested in Advanced Training Hall

Place: Advanced Training Hall
ASI: Very High
Weapons Tested:
Divine Avenger: 49 Combo/100sec -> 56 Combo/100sec = 3.6% increment per asi lvl
Tempered Calibur: 52 Combo/100sec -> 61.3 Combo/100sec = 4.5% increment per asi lvl
Dark Thorn Blade: 53.3 Combo/100sec -> 60 Combo/100sec = 3.1% increment per asi lvl

Note: No UV for all the tested weapons. Only ASI very high is tested.

Search

Let's Experiment: The swing speed of swords.