I’m looking at your data in the sheet, and it seems from your testing (the top part) that:
AS from 1.7526782 to 1.9067828 all produced 22 frame novas
AS from 1.9099821 to 2.0987408 all produced 20 frame novas
and above that you get this weird mic of sometimes ~18 frames, sometimes ~27 frames. Your data average to 23.4 frames, but you say the “true” rate is 22.5.
So these AS breakpoints from your tests seem different from the formula that indicate that breakpoints should be at AS = 21/(N*1.15) with N being 9, 10 or 11, which works out to 1.66, 1.82 and 2.02. While your measures breakpoints are at 1.75, 1.91 and 2.10.
Am I misunderstanding things, or is that a discrepancy between the formula and your data? Is the AS in your sheet the same as the AS in the formula, which is the sheet attack speed?
Maxroll says. to get to the 1.82 breakpoint, but stay below the 2.02 breakpoint.
It seems similar for you, where you get the best nova proc rate above 1.91, but below 2.10, slightly higher than the maxroll numbers that come from the formula.
I will try this once I get better gloves with AD. I already have a primal funery pick which I can roll vit off to AD or elite dmg as was suggested by Abella (I haven’t done this yet).
Hey sorry, I’ll get back to you after I’m back. Just a quick note, that sheet that I linked is not mine, I had my own local version from my own testing and calculations.
You dont really factor in speed pylon when calc bp. Its better to exclude it and ignore speed than to lose damage from lower frames since its not always available.
Yeah, skip, or save for RG and use it to stack stricken first 2 minutes. Worst case you want to click it to spawn elites for a condi, but then you can click it in advance and use the move speed to kite mobs to the condi, zap for a minute, and then the speed is mostly over after that. I really don’t think it’s worth giving up 10% damage the rest of the rift for those rare cases where you’ll be dpsing with a speed pylon.
Yeah I’m not factoring it in, but in my particular example the 5% IAS compared to 0% IAS is a no-gain at best, and a loss with speed pylon. There is literally 0 benefit to it, whereas you could instead have a free 15% Life roll replace it because it hits the same breakpoints outside of pylon.
Hey so, just a reminder that this is not my sheet. Those values at the top align with non-swift harvest values.
1.82 * 1.15 = 2.093. Similar to how 1.82 with swift harvest is right at the edge of 10FPA with 1.826087 being the exact value of 10FPA and anything above that rounding down ot 9 FPA, WITHOUT swift harvest 2.0987408 is right on the edge of 10FPA, with 2.1 being the exact value of 10FPA and anything above that rounding down to 9 FPA. If you look at the bottom left table, you see that 1.83 APS corresponds to 2.1045 APS with swift harvest, which is the 9FPA breakpoint and results in the inconsistent frame differences you see at the top right. As to what the creator of that doc meant by True (Avg) FPA, I’m not sure. Maybe that’s what the calculated true FPA is, whereas the observed FPA was 23.36 due to a limited sample size (trust me it’s not fun counting frames).
If no swift harvest is being used (for example builds with Andariel IAS on top of krsy/rorg/witching hour), the formula would be 21/(N1) instead of 21/(N1.15).
Maxroll specifically lists 1.83 as the 10FPA breakpoint (with swift harvest), and warns to stay under 2.02. In reality, 1.83 is already just barely into the 9FPA breakpoint up until 2.02, and 2.03 is the 8FPA breakpoint.
Ahh yep that makes sense then, thanks. So your tests confirmed that you got more inconsistent and on average fewer Nova procs above sheet AS of 1.82, with swift harvest?
Makes me wonder what other passives you could replace swift harvest with…
I’ve tested and noticed that at exactly 40.6% IAS it becomes inconsistent (varies from 19-27 frames between novas). Anything up to 40.4% IAS it’s a consistent 20 ±1, but both show 1.83 APS on the sheet.
40.4% IAS = 1.8252, which is rounded to 1.83 on the sheet. FPA of 10.00486 rounded down to 10 FPA.
Exactly 10.00 FPA = 1.826087.
40.6% IAS = 1.8278, which is rounded to 1.83 on the sheet. FPA of 9.990628 rounded down to 9 FPA.
So in practice, yes. If you decided, for whatever reason, to put 0.4% paragon points instead of full percent you could technically hit 1.83 sheet APS but still be at 10FPA, but there’s no reason to ever do that so in practice yeah aim for 1.82.
Note: these numbers are WITH swift harvest. Without swift harvest, I use 1.92 with Eternal Torment replacing SH, for example with Andariel headpiece.
oh wow! I guess with the shadow clones, they don’t even make it worth while clicking on the speed pylon in s33 for higher GR pushes (in the 140s)?
Thanks for taking the time to test and add to the thread - very much appreciated.
Given that I had IAS on weapon/gloves and krysbin’s last season (plus SW passive), I was almost certainly gimping my damage. Strange though, it felt like I was doing more damage when using the higher IAS…mobs just felt easier to kill.
I don’t really have the skill or brain to test this on console. I can’t accurately see the nova explosions and trust my brain to count them accurately. I guess I could just record a snippet of 60 seconds video of me casting siphon blood/nova on an empty map with no mobs around me so I can concentrate on recording the data.
BTW, atm, my sheet says my IAS is 1.52 (with 10% from paragon, SH passive and 7% on krysbin’s). NO pylon of frenzy shine etc. Does that sound correct?
edit: where am I going wrong with my math - I tried to work backwards using AS = 21/(N*1.15) to find the break point “N” for AS of 1.52…I’ve obviously done something wrong as the maths doesn’t sound right at all. This was my working out:
I didn’t play last season but from what I understand, you had no choice over what attack speed breakpoint you hit because the ethereal item came with attack speed. However, the item itself was just so overpowered anyways that it didn’t really matter, and is still stronger than this season. If you want to find AS breakpoints for a given FPA, then it’s AS = 21 / FPA / 1.15 for swift harvest, or just AS = 21 / FPA for without. And remember, if you do this the AS value that you see is the MAXIMUM value for that FPA since it rounds down, so you want to stay under it.
I decided to make a google sheet for those calculating their breakpoints, you’ll have to make a copy to edit it though.
Blue cells are for you to edit, with 3 sets for you to setup for the calculations below it. Use dropdowns for shrine/pylon effect and swift harvest toggle.
Dang, I missed that division symbol! I look like I’m bad at math, I used to be good in my youth, but my cognitive abilities are declining with age sadly. Many thanks for the assist.
So for gearing, we want the 10 frame breakpoint with frenzy. Seems there are two main options for that.
Keep swift harvest and roll the now useless ias off krysbin.
Swap out swift harvest, keep ias on krysbin, and then we need another 2 ias rolls.
Did I get that right?
I don’t have a good feel for how much damage we can get from another passive, and how that compares to the area damage we’d be giving up for ias rolls. krysbin always comes with main stat (and ias), so we wont be able to get a crit-crit-AD krysbin, best we can hope for is int-crit-crit, so there is only some main stat to gain from rolling off the ias.
I guess we could also swap out swift harvest, no extra ias and roll ias off krysbin, and just stick to the 11 frame breakpoint. That’s dropping 9% proc speed to get both the new passive and the AD rolls.
So I guess we could rephrase it…
We can pick two out these three:
There are no extra passive interms of pure dps. There are lazier passives such as eternal torment which is obsolete with proper gameplays, as well as other QoL passives.
Most players on this forum doesnt obtain insane paragons so IAS > INT is beneficial. Even vitality is an option mentioned above since it boosts sim survival in bigger setups. Not at home, i ll update Sim HP formula later.
Worthless if you value quality setups due to knockbacks. Its only valuable elite spawned can be killed efficiently. Speed knockback is usually negatively affect sim builds since more often than not you have to adjust sim positions, speed knockbacks breaks setups.
Best value for speed is on RG but bear in mind to benefit most from speed aka stricken, you must manual DN since siphon is a channel skill and stricken is only stacked at a static 54f on channel skills. Iron rose ICD doesnt affect manual DN at all.
Right, so then the best setup then would be to roll all ias off gear and run swift harvest. No important passives left out, 10 frame siphon, extra int roll, and allows 120+ AD.
I made a slight change to my sheet, to account for weapon IAS rolls working differently than IAS rolls on other pieces of gear. If you’ve downloaded a copy for your own use, make sure to re-download it.
If your Krysbin’s is 5%, you don’t need to do anything. If it’s 6% or 7% you could just simply take off 1% or 2% respectively from paragon.
Otherwise if you can’t help it, you could swap Swift Harvest for Eternal Torment for situational DPS buff, for example when you have a very large pack and are running Aura of Frailty, you will inevitably leave some monsters out of your aura when you stand still and DPS, and Eternal Torment would help include those Spreading Malediction 1% buffs if you’ve previously walked over them and have never walked outside of ~1.5 screens distance from them. Otherwise you could run Rigor Mortis for a bit more survivability, but you could make it harder to pull monsters like that.
