Hmmm well if you have the command to shoot being spammed wouldn't the I'm Want to shoot work? Could see if there is any correlation to the phases existing like has long been theorized about ammo or whatever. Or are you already sorta doing?
Hmmm well if you have the command to shoot being spammed wouldn't the I'm Want to shoot work? Could see if there is any correlation to the phases existing like has long been theorized about ammo or whatever. Or are you already sorta doing?
I think most people are just generally interested in "what is better" (I am, anyway). Exact values would be great, but determining whether manibozho gloves beat BFG, for instance, would be very valuable. I imagine it's difficult with rapid shot procs, though =/.
Can test both on different non rapid shot job, nin with lamian kaman(+1) for Manibozho, rdm for BFG.
comparing snapshot to damage pieces is like comparing fast cast and magic attack bonus. Unless I misinterpreted you
You are. I basically was meaning some people want snapshot completely quantified so that not only can we compare one peice to another but get a reasonable way of comparing 1 ranged DD to a melee DD for example outside of a parse if you are lucky enough to convince people to let you try rng on something lol.
Or like certain weapons and builds between rngs accurately
Nahtirah Trousers seem to be 9% Snapshot.
With Trousers: 7.04 mean, 7.82 mode, 4.05 mean of the minimum cluster
Without pants!!!: 7.75 mean, 8.60 mode, 4.52 mean of the minimum cluster (from the previous page)
Mean: 7.01/7.75 = 90.4%
Mode: 7.82/8.60 = 90.9%
Min: 4.08/4.52 = 90.2%
The minimum cluster makes it most obvious whether snapshot is multiplicative or additive with the rapid shot. While the minimum decreased more than we would have expected from the %decrease of the mode, it definitely didn't drop by 0.8, so it's multiplicative.
If New delay = delay*(1-Snapshot%), then they're 9% Snapshot (or some /1024 fraction).
If New delay = delay/(1+Snapshot%), then they're 10%.
Safe to assume 10%? OH NO WERE TALKING ABOUT SE HERE.
If we assume its 10% how does that factor into other pieces, mainly manibozho hands vs. bld. Ffng. Gauntlets
And sylvan head +2
I'm not sure. Next step for me will be to re-merit RNG. I mostly rushed Nahtirah Trousers because they're adding STR to them next update and I wouldn't be able to use the D0 trick on them anymore.
While they don't live forever you could try some of the soa stuff that doesn't attack back maybe. Curiuos have you been looking into see if there is a cap on snapshot?
* Snapshot with 5/5 merits is 10% Snapshot (as described). The upper cluster is still around 7.82, but the mean is 6.96.
** 6.96/7.75 = 89.8%, or approximately -10%. (n=825)
* Repeated the test with Nahtirah and Snapshot 5/5. The upper limit is about 6.936, but the mean is 6.286.
** 6.286/7.75 = 81.1%, or approximately -19%. (n=475)
* Repeated the test with Velocity Shot and Snapshot 5/5. The upper limit is about 6.596, but the mean is 5.964.
** 5.964/7.75 = 76.9%, or approximately -23%. (n=1051)
** This confirms that Velocity Shot reduces ranged aiming delay by ~13% (not the reported 10%), which corresponds better to its +15% melee Slow.
Edit: it is also possible that velocity shot is 15% multiplicative with snapshot, which I can try to test later with the lower bound.
** lower: 3.298 vs. 3.638
** upper: 6.596 vs. 6.936
*** They shifted the same amount, while we would have expected the lower bound to shift down less if it was multiplicative. This indicates that they are in the same term.
* Tested Manibozho Gloves and Snapshot 5/5. The upper limit is 7.379, but the mean is 6.656.
** 6.656/7.75 = 85.8% or approximately -14%. (n=1044)
** Looks like Manibozho hands are about 4% Snapshot.
Also, I'll note that 1044 shots was one stack of recycled bullets (same as 1051 above). That indicates approximately a 90% cap on Recycle, not 95%. (1044-99) = 945 recycle procs. 945/1044 = 90.5% recycle rate with AF2 hat, 5/5 merits, and Mani hands. So that's likely the cap.
Edit: confirmed that the 90% recycle cap isn't due to missed shots. I've conserved ammo even on shots that I missed.
* Tested AF3+1 head and Snapshot 5/5. The upper limit is 7.377 and the mean is 6.56.
** 6.56/7.75 = 84.6%, or approximately -15% (n=660)
** Looks like AF3+1 head is 5% Snapshot.
Those 660 shots cost me 158 bullets, giving me a recycle rate of 76% with 5/5 merits and nothing else to enhance it, indicating that the RNG base rate is actually 50%.
* Combined all my compatible snapshot gear (AF3+1, Manibozho hands, and Nahtirah Trousers) with 5/5 snapshot merits. The upper limit was 6.19 and the mean was 5.52.
** 5.52/7.75 = 71.2%, or approximately -29%.
** Looks like there isn't a 25% total cap). 5.2% + 4% + 9.6% + 10.2% = 29%, so we have the full effect of all of the pieces used.
Well, I'm out of Spartan Bullets and too lazy to make more. If I find some Vision Stones on my mule, I'll consider cranking out a stack or two and testing AF3+2 head, but otherwise I'm putting the pause on this testing for now. I've saved all the distributions from the tests I've done on this page, so I can re-analyze them later for Rapid Shot information if people are interested.
How awful would it be to test at least a few of the values (say lucky and XI) on Courser's Roll?
I could do it, but I'd need to find bullets and a Corsair on my server.
*cough*
I wouldn't have suggested it if I weren't willing to COR XD
Took COR out with Bedlam +2 on the test server to get a baseline 600 delay value with 240 delay bullets (comparable to my 999 Relic Gun + 240 delay bullet). COR and RNG both have Rapid Shot, so I'll use the mean of the upper distribution for this.
* SMN averaged 2.76 seconds per shot with Mamoolbane (276 delay, no bullet) :: 106.6 delay per second
* COR average 5.29 seconds per shot with Bedlam +2 (600 delay, 240 bullet) :: 113.4 delay per second
** (n=565 in 99*4 bullets, 30% Recycle Rate for 99 COR confirmed)
* RNG averaged 8.60 seconds per shot with Relic Gun (999 delay, 240 bullet) :: 116.2 delay per second
There could be an offset, or it could simply be variance in delay per second, which has been shown before. At the moment, I'd assume it's just an average of about 110 delay per second that ignores the ammo delay.
* 4.46 seconds average delay with Acinaces (n=303) vs. 4.77 without (n=565).
** 4.46/4.77 = ~6.5% Snapshot
* 4.46 seconds average delay with Navarch's Mantle (n=321) vs. 4.77 without (n=565).
** 4.46/4.77 = 6.5% Snapshot (probably 70/1024 or something)
* 4.34 seconds average delay with Commodore Gants +2 (n=313) vs. 4.77 without (n=565).
** 4.34/4.77 = 9% Snapshot
* 4.46 seconds average delay with Navarch's Coulottes +2 (n=323) vs. 4.77 without (n=565).
** 4.46/4.77 = 6.5% Snapshot
(just going to update this post with more COR gear)
Wasn't expecting relic hands to have more snapshot than manibozho, looks like I grabbed delve gloves for nothing! Thanks a lot for these tests btw!