Rate of fire testing

[Singular]

Ok, so I tested the RoF/minute of alchemical repeating heavy xbows (AL) against Needle (Ne), with Armor of Speed, Haste, and nothing. This is what I found.

No boost:

AL: 105, sometimes 108 (probably one or the other is my human error; they were in equal number)
Ne: 117

Armor of Speed:

AL: 117
Ne: 117

Haste:

AL: 114
Ne: 117

Interpretation:

Haste, as noted before, isn't as useful an effect on repeating xbows as AoS. It's good for run speed, though.

For some reason, AoS, which is stated to give a +15% alacrity bonus, equates to Needle's +20% alacrity bonus in my tests. Admittedly, I only did about 10 tests each, for one minute, and I am far from perfect, so there could be some error. If my test is correct, though, how ... disappointing - I mean, why bother putting this 20% alacrity bonus, when it's the highest such in the game, on the most sought after repeater in the game, if it adds nothing?

If you don't have Needle, or some alacrity boost, use Armor of Speed. I don't have a quiver of alacrity, but it would be nice to compare Needle's firing rate to that with a non-Needle xbow.

So, use Haste for the run speed, AoS for the boost to damage output. If you're good at twitch firing, you can ignore my advice since you'll reach 120 bolts/second regardless and it's likely the speed boosts will just interfere with your timing.

[methodology: using stoneskin's timer, I noted the number of quarrels in my quiver, and placed it on a hotbar, and pushed the attack button for one minute. I continued the test up until the last second of the minute and regardless of whether the xbow required reloading at the beginning. I discarded one outlier that was off by 12 bolts - it came in at 96 for reasons I don't understand - w/no boosts]

[Dandonk]

Isn't haste noted as a 22% bonus on ranged? Or something?

[Loriac]

Ok, so I tested the RoF/minute of alchemical repeating heavy xbows (AL) against Needle (Ne), with Armor of Speed, Haste, and nothing. This is what I found.

No boost:

AL: 105, sometimes 108 (probably one or the other is my human error; they were in equal number)
Ne: 117

Armor of Speed:

AL: 117
Ne: 117

Haste:

AL: 114
Ne: 117

Interpretation:

Haste, as noted before, isn't as useful an effect on repeating xbows as AoS. It's good for run speed, though.

For some reason, AoS, which is stated to give a +15% alacrity bonus, equates to Needle's +20% alacrity bonus in my tests. Admittedly, I only did about 10 tests each, for one minute, and I am far from perfect, so there could be some error. If my test is correct, though, how ... disappointing - I mean, why bother putting this 20% alacrity bonus, when it's the highest such in the game, on the most sought after repeater in the game, if it adds nothing?

If you don't have Needle, or some alacrity boost, use Armor of Speed. I don't have a quiver of alacrity, but it would be nice to compare Needle's firing rate to that with a non-Needle xbow.

So, use Haste for the run speed, AoS for the boost to damage output. If you're good at twitch firing, you can ignore my advice since you'll reach 120 bolts/second regardless and it's likely the speed boosts will just interfere with your timing.

[methodology: using stoneskin's timer, I noted the number of quarrels in my quiver, and placed it on a hotbar, and pushed the attack button for one minute. I continued the test up until the last second of the minute and regardless of whether the xbow required reloading at the beginning. I discarded one outlier that was off by 12 bolts - it came in at 96 for reasons I don't understand - w/no boosts]

I suspect the 114/117 for Haste/AoS is a measurement error tbh. You're getting a single volley's worth of difference over 60seconds, which is an easy error to make*. When I did similar testing a while back, I started using 2+mins as my test period because I suspected that I was getting errors up to +/-3 bolts per min, and tbh if you're testing the autoattack rate I'd be tempted to say that its best to run each test for 5 mins, as this will dramatically reduce errors (I was testing twitch firing, which is harder to maintain for long periods, as even momentary lapses of concentration can cost you 1-2volleys worth of shots).

Needle's alacrity doesn't stack with haste (afaik, the only bonuses that would stack with haste / AoS are the ranger capstone, and the two ranged sets from ToD).

As Dandonk notes, haste is technically a 22% bonus, which ends up being an effective 15% (at least I think it is; I can't remember the exact numbers, but its around 2/3 of the nominal haste bonus). In other words, alacrity from Needle is very slightly worse than having AoS or Haste up.

The numbers you get for haste/AoS/twitch firing are about what I was seeing. Basically, I found that if you are twitch firing, you can ignore haste / AoS entirely (even feats like rapid shot may not be adding much to your effective rate).

However, and this is a huge caveat, twitch firing is hard to do when things get busy, so if you want to maintain constant dps, its best to keep AoS or haste up at all times (if not using Needle), and take rapid shot as a feat.


*One other source of error was the first volley. My methodology was to destroy all bolts, conjure up 1000, and click once to load them. What I was seeing was that sometimes, loading up the bolts with a left click actually fired 3 bolts, so if you just took the count as 1000-no of bolts you had left at the end of the time period, you'd be overestimating by 3 bolts. I started checking the exact numbers when the bolts were properly loaded before starting the 2 mins (or whatever) of firing.


Edit: for clarity, Haste and AoS should be giving identical fire rates, and I never had reason to suspect this wasn't the case in the fire rate tests I did.

Edit 2: I misremembered it seems, re: rapid shot; it is worthwhile taking this when twitch firing. For reference, my testing is recorded in the following thread:

http://forums.ddo.com/showthread.php?t=403841

[Singular]

I suspect the 114/117 for Haste/AoS is a measurement error tbh. You're getting a single volley's worth of difference over 60seconds, which is an easy error to make*. When I did similar testing a while back, I started using 2+mins as my test period because I suspected that I was getting errors up to +/-3 bolts per min, and tbh if you're testing the autoattack rate I'd be tempted to say that its best to run each test for 5 mins, as this will dramatically reduce errors (I was testing twitch firing, which is harder to maintain for long periods, as even momentary lapses of concentration can cost you 1-2volleys worth of shots).

Needle's alacrity doesn't stack with haste (afaik, the only bonuses that would stack with haste / AoS are the ranger capstone, and the two ranged sets from ToD).

As Dandonk notes, haste is technically a 22% bonus, which ends up being an effective 15% (at least I think it is; I can't remember the exact numbers, but its around 2/3 of the nominal haste bonus). In other words, alacrity from Needle is very slightly worse than having AoS or Haste up.

The numbers you get for haste/AoS/twitch firing are about what I was seeing. Basically, I found that if you are twitch firing, you can ignore haste / AoS entirely (even feats like rapid shot may not be adding much to your effective rate).

However, and this is a huge caveat, twitch firing is hard to do when things get busy, so if you want to maintain constant dps, its best to keep AoS or haste up at all times (if not using Needle), and take rapid shot as a feat.


*One other source of error was the first volley. My methodology was to destroy all bolts, conjure up 1000, and click once to load them. What I was seeing was that sometimes, loading up the bolts with a left click actually fired 3 bolts, so if you just took the count as 1000-no of bolts you had left at the end of the time period, you'd be overestimating by 3 bolts. I started checking the exact numbers when the bolts were properly loaded before starting the 2 mins (or whatever) of firing.


Edit: for clarity, Haste and AoS should be giving identical fire rates, and I never had reason to suspect this wasn't the case in the fire rate tests I did.

Edit 2: I misremembered it seems, re: rapid shot; it is worthwhile taking this when twitch firing. For reference, my testing is recorded in the following thread:

http://forums.ddo.com/showthread.php?t=403841

Excellent points.

I'm actually working from your previous posts - that's why I wrote in my initial statement that, if twitch firing, you could ignore my results.

An earlier thread noted that haste was less useful than AoS, so I tested that. I will very shortly test each, again.

Here are the threads:

http://forums.ddo.com/showthread.php...95#post4203095

http://forums.ddo.com/showthread.php?t=398713

edit: you could very well be correct about the human error. And, as you mention, there's no way to get rid of this except for extra time - I'll add a few minutes in my next few tests.