Pitzerker Simulator for Optimization

[Luhkoh]

I'm reposting this simulator I posted on the old forum on Jan 14, 2020. I would like this to not move to the archive since I have some new discussion resulting from using it lately.

LINK TO SHEET (must download to use)​

Introduction
Hi all. There was some interesting discussion recently in the Hall of Records thread about damage ranges and breakpoints for the pitzerker. I was always frustrated with the fact that, at least to my knowledge, there was never a firm way to tell the best way to allocate stat points between strength and dex for the pitzerker. I also wanted to know how much time beefing up str/dex would save versus dumping all spare points into vitality. I decided to dig a little deeper and started trying to calculate average swings per enemy, but quickly realized this would be so difficult to calculate, that simulating it would be the much better route to go. So I made a simulator in excel (using VBA/macros). It takes all 36 relevant pit enemies (the 4 boss types with mana burn, stone skin, both, and neither, as well as the 5 champion types), and runs a simulation to get average number of swings per run based off that. This result allows us to calculate the decrease in time spent attacking from adding points into strength or dex for example.

Sheet Overview
Instructions is a title page with lengthy discussion of how to use the sheet as well as some discussion on how I used the sheet and what data I found useful for it. This thread will contain just a brief overview of what's available, but there is much more detail in this Instructions sheet if someone is stuck or just wants more info.

Input is where you put in all the stats and gear and such for your pitzerker. This can be used both for a gear planner that will calculate your hp and mana, and is used for the input for the simulator.

Stat Optimization is a method developed by @ResTTe , where you can optimize your str/dex ratio without having to run the simulator many times. I wasn't sure how valid this method of comparison was until I made this simulator, but in the handful of test cases I ran, the simulator and this stat optimization agreed exactly on what was best. So I recommend, if you do want to put extra points into str/dex, that you use this sheet to figure out how to distribute them.

Monsters is the sheet that the simulator uses to run the sim. No interaction is required or recommended from the user. Just there to see how the calc is done (or for checking) or to get more information.

Discussion
First off, from the planner I concluded that I needed to use dual wizzy on swap and be ok with 54% FI for the foreseeable future, as I can't get the amount of mana I would like from dual branches without more mana/mf charms and more mana on my amulet or somewhere else.

Then, more importantly, I analyzed my setup with max vita (~3k hp) and what I considered to be my minimum desired hp (2600). For max vita I got 25.452 swings per run, and for my max dmg with 2603 hp (161 dex, 160 str), I got 24.330 swings per run. I used 8 boss packs per run for these numbers. With 9 frames per attack, this results in 0.404 expected seconds saved per run. Assuming gameplay is the same. Is this worth it? It depends. I decided I would rather have the extra life until my runtimes get down to more elite levels, and then I may test the difference empirically. Because it's entirely possible that the extra hp could save more than 0.4 seconds through reduced potting and more aggressive play. However it is also possible that 2600 hp feels the exact same as 3000, and 0.4 seconds is automatically saved. That must be tested manually, but this sheets both helps optimizes that setup and gives an idea of the approximate maximum amount of time one may give up or save by using more or less vitality. One could similarly use this simulator to see what is lost by moving some skill points from shout to BO. I concluded that was worse than putting more points into vitality, but surprisingly not by much.

Lastly, I did remake this simulator in google sheets, which ended up being a waste of a rather lot of time. The google sheets version took approximately 100x as long to run as the excel one, so I just scrapped it. I neither have the know-how nor the motivation to make the simulator in another form for those of us who don't have access to excel. But if someone wanted to do such a thing, they are more than welcome. Hope this is helpful or at least interesting. Enjoy!

Thanks to @Gripphon @Fabian @ffs for the discussion and past literature. And big thanks to @ResTTe for helping, checking, beta testing, and coming up with the dps stat optimization method.

Edit: Monster Spawn rates edited on 11-Sept-2020 based on the number in this thread.

 

[Luhkoh]

My latest use of this simulator has led me to come to a perhaps controversial conclusion. I think that, for most pitzerkers, highlord's wrath is likely the best-in-slot amulet. Now hear me out on this one . There are so many inputs in such a problem that I absolutely cannot make sweeping statements for all pitzerkers. Depends on runtimes and gear and so many things. Nothing short of plugging all the stats for your own pitzerker in the simulator, with both setups will answer what is best conclusively. And even then it ignores defensive stats that may or may not affect runtimes. But the calc'ing I've done so far leads me to believe this is true for most. First I will explain how I came to the conclusion, then I want to talk about some of the positive implications I see if this is true.

So I have the stats and setup plugged into the sheet for my pitzerker as I have already discussed. I decided mana was too low for comfort with dual branch, so we're stuck with 54% hork. However mana total is acceptable to me with branch/wizzy (important for later). I also have an 18mf 2barb caster ammy, so the conventional setup is better for me than if you didn't have nice mf on your ammy or were running a 1barb like @ResTTe is. I wanted to sim the effects of using LoH and of using HLW, as well as both together and see the results. I have runtimes right around 85s, and higher dmg setups will be better the faster your runs are. Anyways I simmed the time saved by these items and plugged them into a hall of records style efficiency calc and came up with the following results.

Using 85s runtimes and 8 bosses
Setup          MF    Swings per Run   Time Saved   Adjeff
-----------    ---   --------------   ----------   ------
Conventional   759      25.069         0           2.422
LoH            727      23.131         0.698       2.419
Highlord's     727      22.701         0.852       2.415
LoH+HLW        683      20.895         1.503       2.425

Note that this sim was run without might aura active since we were calc'ing most dmg breakpoint information in the past with that condition. The sim should probably also be run with it active and compare the results, but I'll leave that for some work to be done later.

So to me this clearly shows that, for me, highlords is a better gear swap than laying of hands, and highlords is probably best choice overall. The disadvantage of highlords is lower mf (but that, I think, is shown to be worth it from the simulation), and lower mana. But the advantages are light res and whatever the heck stats you can pack on your fcr/mf ring over nagel. So looks to me like highlords wins in the sim, and in the non-simmed defensive stats as well. Plus I think there's the sort of hidden defensive benefit that the less time you're swinging, the less dmg you're likely taking.

Next, I was curious and plugged in the same stats, but used a more elite runtime of 55s, and for that, HLW still came out on top, but LoH+HLW was virtually the same (demonstrating how higher dmg is better for faster runners).

Now things might change if you have the mana charms/amulet and/or disregard for quality of life necessary to run dual branch. Because HLW means you cant reach 55% hork. But if you're @Gripphon , you also run freaking fast, so that might end up favoring dmg setups after all. Tough to say without running it through the sim specifically. May also change a bit if HLW drops you down more than 2 mf breakpoints like it does for me. But I'm convinced that HLW is best for most typical pitzerkers. And definitely convinced that it is a top-tier gear choice for a pitzerker regardless.

Why is it important if it gives an increase in efficiency so small that you cant even notice? I'll tell you . This is good news for many. If you're a beginner, you dont have to worry about crafting a good amulet, which is really tough to do. And it's also an increase in safety as I discussed a bit above. Another side effect, is since you have ias on the amulet, you don't need ias from goldwrap or gloves. So a perhaps non obvious result is that now chamming a shako only gives you 5 extra mf (75 shako + 0 trangs + 40 chancies ...as opposed to... 50 shako 30 goldwrap 40 chancies). In my case that doesn't even hit a new mf bp. Lastly, the ias roll on grief doesn't matter. So those of you unlucky enough to get a 390+/31ias grief are back in business . So that's three ways in which it is beginner friendly, and I think I will start recommending it as "default" modern patch setup. So seems to me like its cheaper, safer, and a slight theoretical efficiency increase. So win/win/win.

I had discussed this some with @ResTTe on discord already, and I think he's currently running HLW and some fcr rare ring with MF and loads of res for max safety/convenience. I want to try a setup with a fcr/mf ring with AR, cold res, max dmg, str, and mana that I have. At some point I'll get back to doing a lot of runs with my zerker and get a practical feel for the difference this makes, if any. Right now too many other run projects going, but wanted to post the discussion anyways .

EDIT: I also just did some math to compare to a dual branch setup. So in my opinion you need a nice mf charms with mana, as well as mf and a lot of mana on your caster ammy (like gripphon) to make this worth playing. But if you have all that, or just dont find it a problem to play with low mana, then what? I get that, if you're slower than approximately 60s runs (for 8 bosses, so 7.5s per bosspack), then dual branch with caster ammy is better. And if you are faster than that, then hlw starts to break even then become better. Again, this is still doing the simulation with no might aura, and taking into account having might active for part of the run might devalue hlw's damage. But I hope I've at least been convincing that it is competitive for a top tier setup with some significant defensive and budgetary advantages as well.

So maybe something like this?
Tier 3 (my new default recommendation): highlords, ist shako, fcr/mf/anything ring, branch+wizzy setup, achievable by anyone, most comfortable, highly competitive with tiers 2 and 1
Tier 2: dual branch for more hork with 2barb caster ammy with mf. likely requires nice mana on charms and on caster ammy for comfortable play
Tier 1: ridiculous old patch barb helm stuff like fabian had

 

[Babyhell]

Hey @Luhkoh! Just to complete, for tier 3 recommendation you go with TO's belt for CBF right?

My amulet only has 8mf, whatever the other mods, so I may give a try at some point to HW+Chancies+FCR ring (only have 39MF but it will do still better than 30+8, and I'll get better AR from a rare - at least I can, I have to see what's in GoMule stashes). Will have to respec him for max Find Item also, I went the damage route but I only get 50% hork on switch.

Anyway I will keep him like he is until RFL, I will see to tune him better afterwards, i'll just run as is the Pits before the 20 hours of AT

 

[Luhkoh]

@Babyhell yep! Ptopaz shako, HLW, Enigma, Grief, 6ist pb (or gull dagger if you have less than 4 ists in the pb), chancies, trangs belt, nagelring, rare ring with fcr, war travs, and then on swap you want wizspike and suicide branch so you'll reach 54%FI after BC. Imo, desirable mods on that rare ring include mf, AR, light res, cold res, and mana. But anything you want will work. I know ResTTe is using one with fcr mf and tons of res, just to increase safety and reduce potting. The one I want to try out has fcr, AR, cold res, min dmg, and mana. Though I doubt that one will be better than one with mf.

As far as the skill points go, one could certainly move some points from shout to BO if desired, but points in shout do more for your dmg than points in str/dex, so I went with the standard 1pt BO and put all I could in vita. I would definitely max that find item over extra points into shout for more dmg though. This sim keeps drilling home the point to me that, while more dmg is absolutely nice, it will result in very minor improvements for zerker.

 

[Babyhell]

Hey

just a thought, as for CBF mod i can use (and do so!) Cham'ed Shako: alongside with HLW, isn't Tal Rasha's belt an alternative to Goldwrap, as it has up to 15MF and 30 increasable mana, in addition with 20 dex which is nice. No life however, but if we count 20 dex out from stats and put them in vitality, it's 80 increasable life for barb, so... I guess it could be worth a try, i would if i had an HLW (i thought i had one, but either it's hiding somewhere, lost it, or an old memory )

 

[Luhkoh]

@Babyhell Yep I'm sure that would work fine! Would probably rather have the mf from goldwrap... I don't personally like to give up mf unless it increases killing speed (or saves a cham for 5mf ). And you get a similar amount of life and mana with the ptopaz shako + trangs belt as well as a bit more mf than cham shako + tals belt. But i suppose the damage taken goes to mana could be a small plus too. If you use it let us know how it goes!

 

[ffs]

I want to post an update thread on Berserker soon which compiles all sorts of findings, insights, tools and resources that are helpful/relevant for Pit Berserker and have only been available since the likes of Gripphon and Fabian have last posted things on Zerker.

Obviously this simulator is huge in really putting things to the test, and I've used it a lot – both personally and for that "Berserker update" post.

I hope I'll be able to finish it soon, but in the meantime just wanted to share this table here since this simulator was key in putting it together, and there are a few surprising takeaways:

EDIT: LOL used wrong table, will post the correct one in a bit.
EDIT 2: Here we go:

	Setup 1	Setup 2	Setup 3	Setup 4	Setup 5	Setup 6
Gear	HLW & Arach	HLW & Goldwrap	HLW & Trang belt	Godly Craft & Goldwrap	Good Craft & Goldwrap	Budget: Oath, HLW & Arach
Head	Shako Cham	Shako Cham	Shako Ist	Shako Cham	Shako Cham	Shako PTopaz
Body	Enigma	Enigma	Enigma	Enigma	Enigma	Enigma
Amulet	HLW	HLW	HLW	2 Barb, 15 FCR, 30 MF	2 WC, 5 FCR, 20 MF	HLW
Ring 1	40 MF	40 MF	40 MF	40 MF	40 MF	Nagel
Ring 2	Nagel 30	FCR/20 MF	FCR/20 MF	Nagel	Nagel	FCR/15 MF
Belt	Arach	Goldwrap	TO Belt	Goldwrap	Goldwrap	Arach
Kill W1	Grief PB	Grief PB	Grief PB	Grief PB	Grief PB	Oath 340 ED WS
Kill W2	180 MF PB	180 MF PB	180 MF PB	180 MF PB	180 MF PB	Ali Baba Ist Ist
Cast W1	Wizzy Ist	Wizzy Ist	Wizzy Ist	Wizzy Ist	Suicide Ist	Wizzy Ist
Cast W2	Hoto	Suicide B. Ist	Suicide B. Ist	Hoto	Suicide Ist	Hoto
Gloves	Chance G.	Chance G.	Chance G.	Chance G.	Chance G.	Chance G.
Boots	WT	WT	WT	WT	WT	WT
+ skills	6	6 with BC	6 with BC	6	5 with BC	6
Swings/run	22.06 (7.94s)	22.06 (7.94s)	22.06 (7.94s)	24.92 (8.97s)	25.35 (9.13s)	31.35 (11.29s)
MF kill	760	775	770	815	805	489
MF hork	620	655	660	665	675	393
FI	55	54	54	55	55	55
BC req?	No	Yes	Yes	No	Yes	No
Adjeff depending on run times
50s	1.546	1.557	1.560	1.557	1.571	1.786
60s	1.812	1.824	1.828	1.821	1.835	2.077
75s	2.213	2.225	2.230	2.217	2.232	2.515

Test parameters:

Spoiler

• Lvl 95 Barb
• “Adjeff (XXs)” columns refers to run time average other than from fighting and casting BC (time spent attacking as per relevant setup is added to this below, so is time from casting BC)
• 8.5 boss average
• No Might aura
• 400 dmg Grief
• 450 AR from charms
• +30 max dmg from charms
• Max vitality build
• 270 MF from inventory
• 20 Howl & 18 Shout hard points
• perfect MF rolls on S&Us

It really requires quite a special craft ammy and likely some time-travel gear to beat HLW – I doubt anyone has it tbh. I believe go-to standard in most cases should be the setup I recently switched to, that is HLW + Cham Shako + Arach. But TO Belt + no Cham in Shako is much "cheaper" and almost as good, plus life/mana from TO belt. Craft amulet needs to be at least very good in order to get even close to HLW setups.

Note that this sim was run without might aura active since we were calc'ing most dmg breakpoint information in the past with that condition. The sim should probably also be run with it active and compare the results, but I'll leave that for some work to be done later.

The effect of Might is very small, that was quite surprising. I didn't include it above but based on some simulations it really doesn't help Barb that much. It still helps of course, and in particular it's great for Merc himself quite a lot with his damage, in particular vs stone skin bosses or multiple champion fights. These things are of course hard to quantify and not accounted for by the simulator. But because of those scenarios only it probably remains best aura choice – however, for Barb himself it's not that relevant. Prayer or Blessed Aim could do more for you honestly.

----

BTW if anyone can think of any setups that should be included in such an overview, feel free to let me know

 

[ffs]

Disclaimer warranted though, the table is WIP and probably contains a mistake here or there but overall it gives a decent overview I think.

Edit: Point proven by posting completely wrong table first, so much for "decent overview" but now that should be the case.

Again, HLW is OP.

 

[Luhkoh]

What a post! So glad you're making use of this and compiling results. Do you have a link to a google sheet or anything? I find the tables on the forum kind of tough to read. So your runtimes/adjeff are adjusted based on BC and swings per run? Just not visibly? If so, just great work. Love it. Your (mf kill - mf hork)'s seem to be a bit wonky on a few.

I have to confess I did recently learn of a minor error in my simulator. Right now it assumes all 4 boss types have equal chances of spawning, and in reality archers have less of a chance to spawn. I can't imagine that this will appreciably affect anything at all, but I should get to changing it.

On your budget setup, I expect its worth dropping the offhand mf and running oath 2 handed. Would be interested to see the result of that. Ofc I guess it's of questionable worth since if you have enigma you can probably grind out a grief.

Very suprised at how little difference might aura makes. I guess it makes some sense, as your kill speed should be affected similarly to putting on LoH or something. Has anyone actually tried running a different merc on barb and seeing if they notice a runtime difference? Probably annoying to do since you have to lose your map if you want to re-hire might merc. But yeah prayer or even holy freeze (think of the champ devilkins) could be sweet in the pit if it didn't make much difference. But as you said, the merc's dmg himself surely makes a decent difference, which isn't captured in the sim. This seems worth testing further to me.

I'm really pleased that this HLW thing is taking off. It's awesome because it eliminates the need for a 32ias grief, to craft a decent amulet, and also makes the impact of chamming a shako very minimal. Top tier zerker is now much more accessible. I plan to try out your hoto/arach setup, but I don't think I can get a mana total I am comfortable with using it. On the hlw & goldwrap setup, I have to use an fcr ring with a lot of mana to get there. This is why I am wondering if I would still recommend HLW+goldwrap as default setup for most, but I'll see if I can get like 680+ mana on that setup and get back to you. It's kind of fun having the customizability in the rare ring, but that's a very ancillary point.

I expect the "conventional budget setup" with crafted amulet, trangs belt, and ist shako is top tier as well. Just if people like to use their rare gloves or something. But it's going to be strictly worse in performance according to the sim, even if barely so, so maybe no need to show it. Also did you run any through with LoH? Last time I was messing with it, it seems like that (plus HLW) could maybe be better for people with very fast runs.

I do think your AR and max dmg from charms are a little crazy, as well as the perf rolls on other stuff, and would love to see these numbers with more attainable gear. I agree it shouldn't make much, if any, difference on the setup comparisons, and maybe you are going for a "best" comparison, rather than seeing what is best for "most people". But if that's the case, why not use lvl 99? Also are you personally running a max vita build or was that just make it easy to compare across multiple setups?

So some suggestions on setups to present:
1. Drop the oath setup and add "conventional budget" setup. With ist shako, craft ammy and rare gloves. I know a lot of people run that setup.
2. Maybe cover the difference between "hlw + goldwrap" and "hlw + trangs" with a note rather than giving the latter its own column.
3. Investigate adding LoH into the mix (I expect it will outperform chancies on most of your 50s runtimes row).
4. And any interest in showing old patch stuff like fabian had or we just ignoring it? Certainly not realistic, but neither are the excellent charms and perf rolls on everyting imo. And might be fun to see how the "new" hlw tech stacks up against fabians setup.

Just some thoughts and recommendations to keep the ball rolling. Also really goes to show it doesn't matter much at all which of the setups you use unless going for a very optimized HoR score. Various budget setups make so little difference, which is comforting I suppose. Loving what you've done with it so far and looking forward to seeing it progress.

 

[saracen85]

@ffs
Wow! Nice compilation!
My setup is so close to number 5, except I'm on Ist Wizzy/Ist Suicides.
Does anybody ever use 2x suicides? Don't think you can max out resists without at least a wizzy...

Also, second LOH setups.

Just my 2 cents.

 

[Luhkoh]

@saracen85 gripphon uses 2x suicides. I don't think your res on tele swap really matters much. On weapon swap though, you can def take some dmg while zerking LE bosses. And the LE/CE invisible fhr novas. And kinda tough to get non-negative res on grief-swap.

 

[ffs]

Your (mf kill - mf hork)'s seem to be a bit wonky on a few.

Thanks for pointing it out, luckily seems like it's just typos mostly with actually calculated values being mostly correct – or so I think will get around to correcting that and some further inaccuracies before it's done. I'll also add Might numbers to show (they're not in this table, only in the wrong one I posted first...).

Do you have a link to a google sheet or anything? I find the tables on the forum kind of tough to read. So your runtimes/adjeff are adjusted based on BC and swings per run? Just not visibly? If so, just great work. Love it.

Yes. I ran the different setups through the sim and fed everything else into this quick and dirty MF adjeff calc that I made a while ago. Not many inputs required to get the adjeff based on simulator results. Originally it had only one table to quickly calculate adjeff for a set of runs, now I expanded it a bit to quickly compare 5-6 different setups by just changing the boss + run times inputs.

At this point I put the table together manually, but might combine it into spreadsheet which includes various setups tabs and adjeff calc, which automatically gets the relevant inputs from the tabs and spits out corresponding results. In an ideal world such a spreadhseet would include your simulator as well, but I understand it's not working with Google Sheets? Should be possible to incorporate the above things in the Excel file though. Will take some hours to do that but maybe I'll get around to it some time.

1. Drop the oath setup and add "conventional budget" setup. With ist shako, craft ammy and rare gloves. I know a lot of people run that setup.

Well, thanks to your simulator we know they shouldn't tbh the Oath setup is not really meant as a recommendation, just wanted to show a low-end setup and the gap to more optimized ones. I agree something like a low/mid roll Grief is something anyone can get their hands on. Might go for that in the final version.

2. Maybe cover the difference between "hlw + goldwrap" and "hlw + trangs" with a note rather than giving the latter its own column.

Is it really that way around? I would actually say that loss of 5 MF for a Cham with the upside of life/mana from belt seems better, with minimal (really negligible) MF loss. But yeah maybe including one of them is sufficient.

3. Investigate adding LoH into the mix (I expect it will outperform chancies on most of youre 50s runtimes row).

I did, and they were very close to Chancies in many instances, but impact of HLW is much bigger and I didn't get a result where they outperformed Chancies. But I can't claim I checked them for every scenario, will do so again in particular for fast run times.

Also this might be affected by Archers having lower probability to spawn – you're right, and I remember that in connection with Flayer spawning odds in Halls of Pain for the OBP tournament. Ranged mobs have different properties in that regard.

4. And any interest in showing old patch stuff like fabian had or we just ignoring it? Certainly not realistic, but neither are the excellent charms and perf rolls on everyting imo. And might be fun to see how the "new" hlw tech stacks up against fabians setup.

A rough number crunch puts Fabian's setup at exactly the same efficiency as HLW with the numbers of his HoR entry. Really goes to show it's nearly impossible to beat HLW, in particular for fast runners. It gets even harder with really fast runs.

I do think your AR and max dmg from charms are a little crazy, as well as the perf rolls on other stuff, and would love to see these numbers with more attainable gear. I agree it shouldn't make much, if any, difference on the setup comparisons, and maybe you are going for a "best" comparison, rather than seeing what is best for "most people". But if that's the case, why not use lvl 99? Also are you personally running a max vita build or was that just make it easy to compare across multiple setups?

It's a mix of "good-to-ideal setup" and "random rolls" with those parameters. I figured it's not too unrealistic to include ~10 dmg/AR/MF charms and another handful of AR/MF charms. As for the rolls, they don't really make a difference, I went with perfect ones to keep things simple because as you said it doesn't have any effect on the comparison itself (at least I never observed one). But chosing some random MF or Grief dmg rolls seemed a bit silly/random as well..

 

[Luhkoh]

@ffs oh I see. So your 50s runtimes aren't really 50s. They're 50s plus time swinging plus BC time. Makes sense and seems a good way to do it. Yeah that's like right on the line of when I was seeing LoH starting to break even. And to be clear, I am talking about LoH+HLW. I agree HLW is better than LoH and should be made use of first when increasing dmg. If you get "true" runtimes around 55-56s, I think LoH starts to be more efficient than chancies. Though I think I was just using 8 bosses per run when I was running it, so may happen at higher runtimes for yours.

The simulator technically works with google sheets. The problem is that it runs like 100x slower. But I did code it in completely a long time ago when trying it. Pm me if you're interested in seeing the google sheets version. It was still a bit of a WIP when I gave up on it. If there's a way to do it, I'll be glad to assist. Or if you just want to make your final product in excel, let me know if you want any collaboration or clarification. Otherwise I'll look forward to your thread.

1. Fair enough on the oath setup. As far as the rare gloves conventional setup, I do think it has some advantages as it offers the most life, res, and mana of any of the setups for what I think will be an unnoticeable decrease in efficiency. But agreed it's the most minor.

2. Didn't even mean to imply a pecking order. Just meant they're so similar that that they might not warrant two columns. Especially since whether it even lowers you an mf breakpoint will vary player to player, where the mf on charms and uniques will differ.

3. Addressed in the first paragraph

4. Haha thats awesome. Might be fun to show that in your updated post, but just a suggestion, far from a necessity.

I agree it's tough to choose "how ideal" the gear in the examples should be. I def think the charms are not realistic. Seems impossible to acquire without trading, but agreed very minor point as it relates to the results. However you choose to go on that will work great I'm sure.

On the spawn frequency thing, it would be very easy to change. Archers have a rarity value of 1 and all other monster types have a rarity value of 2. I don't think the shamans come into play here since they can't be uniques, but someone correct me if I'm mistaken. So I think the chance of a given boss pack being archers is 1/7 and for other monsters it is 2/7. I will have that updated in the linked sheet in the next 10 min or so. If you want to change it on your own copy, all you should have to change are the entries under "chance monster" in the "Monsters" tab, as shown below, using the fractions i just mentioned.

Spoiler

EDIT: Sheet update in link at beginning of first post. If anyone wants to see if the sheet and the link seem to still be working as intended, I'd appreciate it.
Edit 2: nevermind. @art_vandelay says I'm wrong with the percentages. I'll get it figured in the next couple of days and update again. But in the meantime, dont bother changing those percents in your sheets.

 

[Albatross]

Outstanding work! Haven't gone through everything, but will definitely spend time in this thread and sheet if I get a pit zerker going.

 

[ffs]

Btw, if Fabian had been 0.5 seconds faster for his HoR entry, the HLW setup #1 would have broken the tie and he would have been better off with it.

@ffs oh I see. So your 50s runtimes aren't really 50s. They're 50s plus time swinging plus BC time. Makes sense and seems a good way to do it. Yeah that's like right on the line of when I was seeing LoH starting to break even.

Yes exactly, I think this is the easiest way to compare different setups once you have determined swings per run with your sim.

If you get "true" runtimes around 55-56s, I think LoH starts to be more efficient than chancies. Though I think I was just using 8 bosses per run when I was running it, so may happen at higher runtimes for yours.

Now checked them again and the difference seems to be exact uMF breakpoints, i.e. whether 40 MF form Chancies provides 2-3 or 3-4 MF breakpoints across both switches. Chancies win for the 3-4 BP scenario and LoH win in the 2-3 BP scenario, provided runs are not longer than ~65s or so. Difference is negligible either way (0.002 - 0.003 adjeff, or ~0.15% adjeff difference). So pretty much interchangeable.

Not sure, maybe Chancies win in the 2-3 MF breakpoint scenario as well when accounting for Dark Archers spawning a little less often. Let me know when you update the sim to that effect.

 

[Gripphon]

One issue with making pitzerker faster is s/e times. I frequently finish my runs in ~45-50 seconds range even for 9 bosspack runs, but problem is too long s/e time after that ruining all effort. Hence I personally can't test the benefit of HLW unless I do the testing ignoring the s/e time in some way...

 

[Neck Romancer]

Would Arreats Face be viable on this build?

 

[ffs]

One issue with making pitzerker faster is s/e times. I frequently finish my runs in ~45-50 seconds range even for 9 bosspack runs, but problem is too long s/e time after that ruining all effort. Hence I personally can't test the benefit of HLW unless I do the testing ignoring the s/e time in some way...

Indeed there's is the effect of longer run times making HLW advantage smaller, and at some point MF becomes more beneficial. But no matter the exact rolls, charms and general MF gear: I think runs have to be really quite a bit slower for that effect to turn the tides away from HLW. (In case of the parameters above, even for 75s runs HLW is still slightly ahead of setups with very good craft amulets.)

Or do you mean in case of your run times you always have longer S&E times just by using HLW instead of a craft ammy? That would of course change things, but I didn't notice a difference in S&E times whether I used HLW or not.

Would Arreats Face be viable on this build?

Viable yes of course, but there's not really a reason to use it. You gain +2 Berserk and 20 str+dex. It doesn't make up for 50 MF from Shako, and its other stats aren't really helpful. If you don't like Shako (e.g. for aesthetic purposes ), Immortal King combo with headpiece and boots is another option. Very close to Shako + WT but in the end the latter combination is just better.

I think in terms of efficiency Shako can be outperformed only by a godly Barb helm with +Barb, + FI and sockets. Even then it's doubtful whether Shako's other stats aren't worth going for Shako anyway...

 

[Neck Romancer]

Viable yes of course, but there's not really a reason to use it. You gain +2 Berserk and 20 str+dex. It doesn't make up for 50 MF from Shako, and its other stats aren't really helpful. If you don't like Shako (e.g. for aesthetic purposes ), Immortal King combo with headpiece and boots is another option. Very close to Shako + WT but in the end the latter combination is just better.

I think in terms of efficiency Shako can be outperformed only by a godly Barb helm with +Barb, + FI and sockets. Even then it's doubtful whether Shako's other stats aren't worth going for Shako anyway...

Lol yes, i definitely want to make a cool looking barb for my next build. I can afford a pitzerker (Cham,Jah,Ber,Lo, 6 Ist etc... damn these guys are expensive!!) and definitely want one at some point though. Maybe i will go for a WW Barb.

The orther stats that are significant are the life/mana and Damage reduction i assume. These guys seem very hardy and the life doesnt seem like a deal breaker, though damage reduction and possibly mana are very valuable.

 

[Luhkoh]

@ffs yeah i think gripps point is hlw = faster runtimes = slower s/e. I'm still researching the spawn rate stuff a bit. It seems like it is really not well known/documented. @ResTTe manually counted pack types over 50 runs a while back, and the results were at least close to the 1/7 & 2/7 initial prediction. With plenty of room for error ofc. If you're still crunching numbers in the meantime (I don't expect to have this sorted out completely till sometime next week), then I would recommend going with the 1/7 & 2/7 thing that I initially thought was correct.

@Neck Romancer I recommend reading fabian's original guide and gripphons update. Hopefully that will give you an idea of what is needed for pitzerker. Our findings in this thread suggest you really dont need a cham. And like ffs said, if you want him to look cool, ik helm is the go-to. WWbarb is indeed the main build that might use arreats (though gface often is used on him).