Attack Speed & Weapon Speed

[oceanclub]

Can anyone explain what measures slow/very slow/etc weapon speeds are, and how they relate to attack speed?

 

For example, I have a slow weapon, and a very slow weapon which offers +20% attack speed. How do I determine which offers the most hits per second?

 

Thanks,

 

P.

[KDubya]

You need to dive into the frame rates.

 

http://steamcommunity.com/sharedfiles/filedetails/?id=416939844

 

The steam guide above has a nice description of how it all works.

 

Also look for Andrea's Lady of Pain  fighter build, it has a lot of analysis into how to improve attack speed.

[Sensuki]

There is some incorrect information in that guide, and all other guides I've seen that speak about attack speed. They forget that the game has a global recovery multiplier that modifies the recovery frames so that it is not directly as follows

 

Now, for the dual-wielded weapons, attack time always equals recovery time. So it's 20/20 frames for the small ones and 30/30 frames for the big ones (1 second is 30 frames, btw).

I'm fairly sure from memory that the global recovery modified is 1.4, so the recovery for a fast weapon is actually 28 frames, not 20 frames. The recovery for a normal weapon is 42 frames as a base, instead of 30 and after this calculation, you then apply modifiers.

 

I'm 99% positive I covered this in my youtube guide about UI and Controls, and there are definitely posts on the forum about it. For some reason this info seems to have slipped through the cracks.

[AndreaColombo]

Sen,

 

just found your video in which you explain the global recovery modifier is 1.2.

 

However, I can't seem to reconcile that with my observation in my frame by frame analysis. Have a look at the numbers in this post.

 

On a naked, 10-DEX character with a two-hander I get a recovery time of 50 frames. That's including a 4-frame delay that occurs from the moment the recovery bar becomes empty to the moment the attack animation starts again.

 

Using the math from your video, it should be 54 frames, so 4 frames off compared to the observed value.

 

Revising my math with the Blade of the Endless Paths (where I also mistakenly counted Armored Grace as -16% instead of -15%) we get:

 

Penalties:

Plate Armor (50%)

Vulnerable Attack (20%)

 

Bonuses:

Armored Grace (-15% penalty)

Durgan Steel armor (-15% penalty)

Durgan Steel weapon (+15% attack speed)

Speed weapon (+20% attack speed)

Deleterious Alacrity of Motion (+50% attack speed)

25 DEX (+45% action speed; applies to both attack and recovery)

 

Expected:

Attack: 30 * (1-0.45) = 16.5

Recovery: (16.5 * 1.2 * 1.5) * (1+0.5+0.2-0.15-0.15-0.15-0.2-0.5) / (1+0.45) = 29.7 * (1+0.5+0.2-0.15-0.15-0.15-0.2-0.5) / (1+0.45) = 29.7 * (1-0.45) / (1+0.45) = 16.335 / (1.45) = 11.26

 

 

Observed:

Attack: 24 frames (seems to be hard-capped.)

Recovery: 15 frames (11 + 4-frame delay)

 

In this case, we get really close if we do not consider the 4-frame delay as part of recovery but animation speed is not what I calculated. Using 24-frames:

 

Recovery: (24 * 1.2 * 1.5) * (1+0.5+0.2-0.15-0.15-0.15-0.2-0.5) / (1+0.45) = 43.2 * (1+0.5+0.2-0.15-0.15-0.15-0.2-0.5) / (1+0.45) = 43.2 * (1-0.45) / (1+0.45) = 23.76 / (1.45) = 16.38

 

This gets us very close to the observed 15 frames (really, I don't think 1.4 frames are visible to the naked eye.)

 

 

But still, the math for a naked character doesn't return true for some reason. Same with my observations on wearing plate mail, where I got 75+4 frames versus your calculated 72 frames. Stuff is starting to get confusing...

[Zenbane]

Sen,

 

just found your video in which you explain the global recovery modifier is 1.2.

 

However, I can't seem to reconcile that with my observation in my frame by frame analysis. Have a look at the numbers in this post.

 

Well Sens recently stated that he hasn't played PoE since April, so I imagine that most of his stats/analysis will be off by now.

 

As for the original question,

 

"Can anyone explain what measures slow/very slow/etc weapon speeds are, and how they relate to attack speed?"

 

Personally, I would just keep different weapons in my inventory and have a fun testing session. Pen, paper, stopwatch... just attack and record the time for each swing, 5-10 swings per weapon.

 

I'm not being sarcastic either with the pen/paper reference. I loved the pen/paper days of old (Dungeon Master, Eye of the Beholder, Shadow of Yserbius, Stonekeep, etc)

 

At the end of the day, the stats you record yourself hold true moreso than someone elses post, video, or patch notes

[AndreaColombo]

^ A frame by frame analysis is going to be quicker and more precise than recording seconds on a piece of paper (seconds vary depending on frame rate too)

[Sensuki]

You cannot have a half or partial frame, an animation frame is either rendered or it isn't. 11.26 frames would either just be 11 or 12 frames, I'm not sure if stuff like that is rounded up or down, but in this particular example it looks like down?

 

You would be better off asking Josh Sawyer on the Something Awful forums - I got my information from him during the beta. It is possible that they have not told us some things, or slightly changed some things.

 

I found that many times during the beta the way he wanted things to work and the way they actually worked did not match up, either because the person who implemented his mechanics did so differently than asked (misunderstanding?), or because of a bug.

 

Back when I originally looked at frame by frame recordings of attacks, my calculations matched what I saw on the screen, but I was not dealing with complex additive maths from many bonuses, just simple ones from raw character creation, armor (or lack of) and global recovery mult.

Edited November 3, 2015 by Sensuki