so hold on, you're telling me that your 99 fcr build that gives up everything just to reach 99 fcr is... mostly teleporting with 48 fcr? a frame lower than a standard build?

this shit is getting psychedelic get me the foh

This post was edited by Dolphy on Sep 25 2025 10:45am

yes, i have tried that one. was definitely second best. i also tried 99 fcr on swap + 60 ias for some extra facets in monarch but it's not worth it IMO --> overkill. obviously we talk about a tiny margin but this one here with tele on swap was fastest for me: https://maxroll.gg/d2/d2planner/qudnq0cq
you obviously tele MOSTLY on swap, the more you practice and get better at swapping the more you tele at swap. as long as you tele once more on swap than main you saved frames which is always the case no matter where you farm.
a certain person here is unable to understand this though

anyway... it's fun to play around with these setups on hero editor and you could actually use tstrokes instead of the 6/40s or 4/40s and still reach 95 ias with 4 x 15 ias jeewels in a 4 os monarch which drastically lowers the cost.
but it doesn't rly matter much because the rest of the setup (especially the two rings and the amy) would be expensive enough

i appreciate the converstion with you and your behaviour

SP is nice for testing.
Hero editor cuz I’m far from being able to afford any of the expensive setups we discussed.
Static map seed layout and fully discovered map which negates any clear speed factors related to map reading or difference between the mobs strength/density, or luck.
So we’re left with only our game mechanics to care about.

I’d dare to say, humbly, that my game mechanics are way above average. So when I compare setups, if one of the build requires much more acuity and high level execution, I keep track of performances over many runs to include a vaste sample of minor human mistakes here and there.
I still try my best to avoid those mistakes and challenge myself at beating my current best performance.

But to be honest, here, about these 2 javazon setups…
I first recorded a single unperfect run using the 99/60 setup to clear TZ baal, CS, tombs and cows. No second try, just a single run as it was. I had in mind to switch back to this setup once I had beaten my score on the 68/95 setup.

Then I must have tried at least 30+ times, in each of the 4 TZs (so 120+ attempts) to beat that score on the 68/95 setup with 99fcr on swap. Trying harder and harder to avoid any potential frame loss. Felt pretty good about this near perfection I achieved in a few of those runs. But I couldn’t beat my timer achieved on the first try with the 99/60 setup. Some very close calls. Maybe doable if I keep trying harder. But then it feels like I would very easily beat that new record achieved on the 68/95 setup if I do more than 1 try with 99/60.

The only place I successfully recorded a slightly better time was when I compared both setups at doing a traditional baal run, clearing only the waves + baal instead of a full TZ clear. In which case, both having the same 99fcr breakpoint without risks of loosing time weapon swapping, while the difference between more dmg per hit vs 1fpa faster achieved the same results versus the waves, but the 95ias setup killed baal 3 frames faster each time.
3/25=0.12s
And this theoretical 0.12 second corresponds close enough to the 0.11s average advantage I’ve achieved over 10 regular baal runs with both setups.

I personally don’t tend to do regular baal runs.
I either run the good TZs or else I’m just messing around doing other stuff than farming.
So for me, the way I couldn’t beat a single time my timer achieved on 1st try with the 99/60 setup with 30+ tries in each of the 4 mentioned TZ… well it’s enough for me to conclude that, in my case at least, with my playstyle and my game mechanics, I have slightly more success with the 99/60 build, as well as much better QoL.

I compared both setups using 3 pairs of 4/40 javs shown in the maxroll planer I previously linked. It wouldn’t be fair to compare the efficiency of fcr/ias if we start using different javs on each setup.
I didn’t bother testing with 6/40, as I already know that the constant repair needed is more time consuming than the marginal time saved on average, as well as resulting in less mana sustain. I know 6/40 can feel a bit faster if you compare single quick runs, looking only at time from wp to game end, but it’s not as fast if you record multiple runs in a row including town travel.

I’m not gonna try to convince you further than that. Cuz I guess at this point that it depends on each player’s acuity, playstyle and mechanics, as well as our tolerance to QoL loss.

Again, only my personal opinion, based on my own performances, here’s my podium for each aspects:

Clear speed
99/60 -> 9.5/10
68/95 -> 9.4/10
68/55 -> 9.3/10

QoL
68/55 -> 8/10
99/60 -> 5/10
68/95 -> 4/10

Cost/Effectiveness ratio
68/55 -> 10/10
99/60 -> 2/10
68/95 -> 1/10

Global score (from the average of the 3 above)
68/55 -> 9.1 (3rd place in my personal ranking of all builds)
99/60 -> 5.5
68/95 -> 4.8

I exclude from the podium the setups using 99/95 without razor, and the (sadly popular) phoenix, both those options suck hard.

That was enough javazon game time for me for now… I need a break from it now XD

This post was edited by SinsOfTheSun on Sep 26 2025 08:58am

i rly appreciate your testings. and i do believe you that the 60/99 setup worked best for you.
95/99 setup with tele on swap is definitely fastest on traditional baal run. there is no doubt. for me personally, it made not much difference after a while if i press w for teleporting or another button to switch from lf to teleport on right click. so i had the best results with the 95/99 tele on swap setup.
a setup that was 3rd best at my testings was actually a 95 ias/68 fcr setup with spirit on main, so no swap. basically the standard 55/68 setup but with a 4/40 or 6/40 instead of titans which let's you jump from 55 ias to 95. what i like about that one is the relatively good cost-efiiciency-ratio.
you "only" need an expensive jav and not amy and rings. LoL... still extremely expensive but hey

what i haven't tested so far is the 55/99 setup with magefists that you described in your first post and the idea of using one more jav on swap side and get rid of cta. as you mentioned and i fully agree with you, amazon is very tanky and cta might not be as important as one might think. at least on SC.
the only real downside i imagine would be a smaller manapool which could be anoying for long teleports. so possibly a QoL loss.

another thing i actually wanna try: half freeze duration on rare ring. could be good enough enough with cure on merc so you could replace cham with facet

This post was edited by pobo2 on Sep 26 2025 12:43pm

Lost me at phoenix

lol even got 2 more javs in cube for backup

The 99/55 setup using mages and a ias jewel in griff and a raven is just a cheaper/worse version TBH
I wouldn't say it's worth trying it

The 99/60 setup is more well-rounded, and that 5 extra ias allows you to combo throw+tele as fast as 99/95, cuz if you don't throw multiple times in a row then the shorter action frame reached at 60ias allows you to follow up with tele quite fast.
So just to picture how I play it :
Teleing with 99fcr
instantly throwing when I see a mob, but following up with tele to keep progressing forward
then throw again while I land closer or over the pack, again followed by tele to move on to next pack while this one is dying

To put it in perspective, let's say we look at a situation where you need 7 tele to reach first pack, and need 2 attacks to kill that pack, 2 tele to reach next pack, 2 attacks to kill this pack too. And everything is done pixel perfect.
99fcr = 12 frames
68fcr = 13 frames
weapon swapping = 7 frames, during which you can't perform cast/attacks.
With those absurdly high dmg builds, most packs require 2 throw to kill.
With 95 ias you need 18 frames to throw twice.
With 60 ias, you would need 20 frames, but since you have 99fcr on that side you can cut the animation with a tele follow up after 6 frames, resulting in 12 frames required for the 2 attacks instead of 20 on the 68/95 setup.
I mean, you could do it on the 68/95 setup too, but 2 frames slower if we look at the full 4 action sequence. So I'll list both ways...

99/60 :
12+12+12+12+12+12+12+6+12+6+12+6+12+6 = 144 frames

68/95 (who also starts the long tele sequence on cta side) :
12+12+12+12+12+12+12+7+9+6+13+13+9+9‎ = 150 frames (and 1 tele distance behind than the above who casted an extra tele past the 2nd pack between the 2 last attacks)
So 160 frames if we include that last tele missing (replacing the last attack by 6 action frames + 13 for tele)
or 167 frames if we assume another pixel perfect weapon swap + tele, in case the following thing is another long tele distance
or, if we want to be fair and use the same attack+tele combos on 68fcr side, we are still looking at a total of
12+12+12+12+12+12+12+7+6+13+6+13+6+13+6 = 154 frames instead of 160

I feel like that full sequence is a very fair representation of the average short sequence we could isolate to observe what happens during some TZ gameplay.
144 frames vs 154 frames if we compare what would happen if both setups killed both packs and tele'd the same distance, everything pixel perfect
And the 144 frame sequence is easier to execute than the 154 frame one, but also comes with 2 extra skills (cuz we're comparing both setups with same javs, either 4/40 or 6/40 it doesn't matter)

So in theory, it's impossible to beat 99/60 in any situations outside of baal runs.
In practice? The extra weapon swap required on 68/95 incorporates more room for human mistakes.
I'm looking back at my 25fps recordings, frame by frame, and even on my best runs where it's near perfect, there's still a frame or 2 here and there lost at least half of the time I need to weapon swap.

hahahah don't look at my cube, hidding my goodies in there
(I'm glad you caught that detail haha)

I feel like nothing can really be upgraded on that planner at this point, showing my wet dream zon that I enjoyed playing from hero editor.
Even the merc, using GPA at 14 fpa jabs
cuz even if you use GT and reach 13 fpa, the 1 frame faster is only saved after the 2nd hit has landed
and you kill so fast on that build that it never happens
so GPA = highest dps on short burst
not that it's really relevant, but higher dmg per hit at the same action frames simply means more sustain

Fun build to play in short period of time, but fuck I feel like playing javazon at maximum efficiency is sooo exhausting compared to other builds

i thought you didn't wanna convince me further, haha
but nice effort! i AM convinced now that 60/99 setup is faster overall

i don't like nitpicking and in the end it makes no difference but on my 95/99 swap setup i have a facet which is worth more than 2 skills and it's actually possible to put in another facet for 1 ias jewel --> 90 ias. a bit weird because 1 frame lower CS but LF only needs 89 for max speed. so due to facet(s) it's actually higher dmg than 60/99 even with 2 skills less. ofc in practice, that doesn't change anything
btw... do you have a source that weapon swap is a 7 frame animation/delay? in my recollection the delay for swapping is 0.2 secs which would result in 5 frames. but when YOU say it's 7 then it probably IS 7

anyway... nice job. was a pleasure reading your testings