reduced mana reservation vs. increased maximum mana (formula)

"

Zakaluka escribió:

While all this is well and good, the important thing to consider is how it all effects your uptime. "unreserved mana" - well that's nice to know, but regen has way more to do with your uptime than max mana.

Uptime means: what percentage of the time can I nuke? Downtime is: what percentage of the time am I forced to be inactive because I don't have mana? Some percent of downtime is okay (even efficient) in most builds because you have to spend time walking/herding.

This tool isn't perfect, doesn't account for every possibility. In fact, I don't actually completely understand how "reduced mana reserved by %" bonuses work, so they aren't supported right now. Neither is Inner Force at the moment. But this thing will show you your uptime considering all sources of regen, the cost of the skill you spam most, attack speed, size of mana pool, mana reserved.... And then how it changes your uptime to adjust your stats in various ways.

Uptime

If you feel like using it, make a copy first. It's read only. If you add in a section for reserved mana reduction, I'll merge it :). have a look.

the reduced mana reserved is exactly what I was looking for though...

not accustomed to working with spreadsheets, though I am positive they are great, and are very easy to use. I just forgot all the formulas and never can get myself to take a few hours to remind myself all of it. It could do powerful calculations.

but reduced mana reservation is easy. you could add it yourself, and it accounts to the passives as well as the reduced mana support gem. what you do is take the amount your auras reserve, and calculate the reduction as if those were your 100%. for example, if you have 40% reserved, then 1% reduction is 0.4%. a passive with 5% reduction is actually 0.4*5=2%. so taking one passive like that would make you reserve 38% instead of 40%

This is really good work Dao, and I certainly appreciate the analysis you've done here.

However, if you don't mind my saying, it suffers (and you may not really care) from something I see analysts do in my work place everyday, a failure to up-level the discussion outside of the circle of other analysts.

In other words, to gain broad acceptance (again, you may not care), translate your great hard work into concrete, actionable steps/conclusions for the dumb masses to understand the implications (i.e., the executives who have to make hard decisions).

Good work though!

"

Dao escribió:

I think it has two typos though.

dX/dMRR = (1+IMM)*MRR

should be

dX/dMRR = (1+IMM)*BMR

and I think that

IMM < 1.6/BMR - 2.6 + MRR

should be

IMM > 1.6/BMR - 2.6 + 1.6*MRR

correct me if I am wrong.

You are absolutely right. Thanks for checking!

here it is again, dao, with a couple new entries and a new row for reduced mana reservation.

Now because reserving less mana doesn't actually change your regen at all, %uptime stays the same. The only difference is that maybe you have enough mana for more attacks in a single uninterrupted salvo.

So %RMR (I am coining this abbreviation because typing out the phrase is awkward) gives you a longer time to OOM but provides no benefit for %uptime. Every other mana-related passive improves %uptime. I'd probably only take %RMR if I was running several auras and had well more than 70% of my manapool reserved.

In most of my builds I see 20%regen passives (but there aren't many of these) and 5%rmc winning out as having the largest impact on uptime. Followed next by 8%mana nodes because they indirectly allow you to also add +1 clarity. But if you weren't even running clarity in the first place, things would probably turn out differently.

"

Zakaluka escribió:

here it is again, dao, with a couple new entries and a new row for reduced mana reservation.

Now because reserving less mana doesn't actually change your regen at all, %uptime stays the same. The only difference is that maybe you have enough mana for more attacks in a single uninterrupted salvo.

So %RMR (I am coining this abbreviation because typing out the phrase is awkward) gives you a longer time to OOM but provides no benefit for %uptime. Every other mana-related passive improves %uptime. I'd probably only take %RMR if I was running several auras and had well more than 70% of my manapool reserved.

In most of my builds I see 20%regen passives (but there aren't many of these) and 5%rmc winning out as having the largest impact on uptime. Followed next by 8%mana nodes because they indirectly allow you to also add +1 clarity. But if you weren't even running clarity in the first place, things would probably turn out differently.

responding without examining the spreadsheet, sorry.

actually, you need to figure out what your regeneration is, and what your waste is.

if a skill uses 50 mana and cast time is 0.5 seconds then your waste is 100 mana per second.

if your regeneration is only 90 mana per second, then you "bleed" 10 mana per second. or, 90 mana per second is 45 per half a second (per skill activision) so you bleed 5 mana every time using the skill

if getting more regeneration is some 10 passives away or so, increasing your mana on the way is a viable option (you can later respec a few points), and that could include those reduced mana cost passives.

you need 50 mana for the skill, and every 5 mana on top of that would allow you to spam the skill one more time.

if you kill all monsters in say 10 spams, you need 100 mana to spam it non-stop 10 times. then 1 or 2 seconds of running and you are back at full mana and full 10-uses capacity of the skill.

your "up time" doesn't have to depend on just regenerating more than you waste. more mana could benefit it very much. i don't know what that %uptime you are talking about, but more mana is great no matter if it is coupled with regeneration or not. it all depends on your build.

as for mana regeneration, you can get 60(40+20) at witch, 60(40+20) at templar, 40(20+20) going down to middle between witch and shadow, 40(40) in the middle, 60(20+20+20) going down from middle between dualist and ranger, and 20 more going down the same path. total is 280% regeneration, or 3.8 times what you regenerate (and 3.8 times what clarity gives as well). just from passives no gear. not very viable. i am skipping templar so i can get myself up to 3.2 in the end, if it is really necessary and remember that gear can give a lot too (those last 20 would make me take 2 more +10 dexterity, and another 8% mana on the way, but it gives 20% regeneration and 20% mana, so 28% mana for 4 passives which is 4% less than you would get from 4 increased mana passives, but also 20% regeneration, so it is really appealing)

as for reduced mana reservation, the anlysis reply to my post shows exactly at which percentage it would be better to take those than increased mana. you can change the inequality to 10/8 instead of 5/8 to check if it is viable to take those (since if you're after one might as well take the other, so it really is better to check 10%) and to compensate how much passives are wasted to reach there, call it X passives in total, make it 10/(8*X) so if 3 passives are wasted, it is 10/24 thus you will see with that analysis inequality that the poster built, at which percentage it would be good to take those.

"

Tindraa escribió:

This is really good work Dao, and I certainly appreciate the analysis you've done here.

However, if you don't mind my saying, it suffers (and you may not really care) from something I see analysts do in my work place everyday, a failure to up-level the discussion outside of the circle of other analysts.

In other words, to gain broad acceptance (again, you may not care), translate your great hard work into concrete, actionable steps/conclusions for the dumb masses to understand the implications (i.e., the executives who have to make hard decisions).

Good work though!

i don't really care for any recognition or acceptance. if somebody understands what i wrote and finds it useful then it is not a wasted post. it is really low level though, some variables and an inequality. nothing fancy. didn't expect to get bombed with derivation, which shows the high level of the forum, and really a great post that i will use (both the post itself and the methods for my future calculations)

the process i am going through is easy - this game is not work, not a social race, not anything but fun. that passive tree alone would grant me hours of fun, and until i learn it i am actually theorizing over it more than i play (still fun though)

i produce intermediate results of some simple calculations. if i would ever think it to be a good idea to consolidate them i will, if i ever think that building a tool or spreadsheet is a good idea, i will. if i do, i'll share, if i don't, obviously there will be nothing to share.

i am helping myself, and will share anything i come up with if it is somewhat good or can benefit others. won't do anything more - like build a finished product for others unless i need it.

my process is first to search the forums, read the awesome mechanics thread, and see if anybody else did anything similar. if not, then do my own calculations. if somebody else has a similar process, they would find this post, and if it is somewhat useful to them, then they might speed up their own process. if that person is in the middle of building a tool, maybe my post would speed their work, and thus we will be benefiting each other.

haven't learned any mathematics short of in school (though my father is a math teacher so i picked up a few things here and there, and still can pick up more if i need), so really nothing fancy in what i am doing. just taking simple things and following through. that and i made the post twice as long to remind people of the simple subject of inequalities with, as you can see, really 1 variable in the end (just the % mana increased you have right now, when you want to decide if to increase it or not. the rest is numbers that you insert to replace all other variables).

"

i don't know what that %uptime you are talking about

Okay, this is actually rather discouraging.

I just made a worksheet that illustrates that rather long post you made, and lays out the benefit in hard cold numbers: actual effect of each mana/haste stat on uptime and percent uptime. You spent what, at least 20 minutes writing a rebuttal? If you had even opened the spreadsheet once you'd see that I already understand everything you're saying. It was a wasted 20 minutes.

If you're afraid to follow links, examine the url. It's on google. Maybe that would add some confidence.

Percent uptime is the main point I'm driving at. You'd rather write a rebuttal than even look at my tool for 2 minutes to understand my main point x.x. Have your %RMR, it certainly does help. Just trying to illustrate what it really does, and that there are usually better ways to accomplish what you're after.

i understand your point, but for ranged characters %uptime may prove to be less valuable than actual uptime and unreserved mana. for example, what i do with vaal is run in circles and have short bursts on him. best burst spell i have is firestorm, so i link it to as many supports as i have (increased duration, cast speed are most important) and then i burst him and run away. my mana most of the fight is full, and i don't care if on a burst i empty it because in 2-3 seconds it'll be full again, and just avoiding the lasers (or seeing the cold floor which signals lasers) is more than that. since he can burst lasers at any time, and firestorm takes a while to start up even with cast speed (which means i can be stuck and get hit by 2) i usually give a burst right after lasers and wait for another laser before i give another burst. mana is wasted on temporal chains and phase run (100% uptime, only linked firestorm empties my pool)

easy fights are easy, but if plenty of white mobs come sure 100%uptime is a little more important, though running away a few steps every few seconds is still necessary so %uptime is better than is on paper. but if you shoot and run shoot and run, bursting ability could be much more important than 100%uptime.

as i said, i haven't used spreadsheets in a long time, and i don't really want/need one right now because the passive tree is so complex that 10 spreadsheets can't cover it... but i did examine yours a little bit. looks very nice for %uptime calculation. maybe i'll capitalize on this overcoming of my spreadsheet laziness and take some time to learn it again. it could be really powerful from what i recall, calculating very complex things. maybe even include a page with both my calculations and the derivation analysis reply that was posted in this thread, so i could have a very good tool for it.

a few remarks:

the reduced reservation passive, at least from what i gather from the forums, works only on skills that reserve a percentage. so you need at least 2 separate slots for it - %reduced reservation passive, and %reduced reservation from support gem (provided clarity and all other auras are linked to that gem).

if you plan on publicizing it on the forums, the attack speed part is usable with spells (ignoring %ias on weapons and just thinking it to be cast speed) but since you have 4 columns of attack speed, 1 column of %3 cast speed would be nice, as well as a true/false for an "is weapon/cast" somewhere

duration shouldn't be in percent? (question not typo) because it shows exactly how long 1 attack/cast takes.

what i am after in this post is just calculating the effectiveness of reduced mana reservation, nothing more. as you said, it won't do too much good unless you reserve high, or if you picked up plenty of increased mana nodes, in which case it will be much more effective to detour to it. we can use the derivation calculation that was replied here to see it. the effectiveness of a mana increase is constant, meaning it would increase your base mana by the percent amount divided by the reserved amount. if you reserve 50% then an 8% mana increase node would be effectively be 4% increased mana. also, if you reduce the reserved amount, it would actually make those mana increase nodes more effective. the effectiveness of mana reservation reduction depends on reservation as well as increased mana percent. so if you have say 200% increased mana, and same 50% reserved, the effectiveness would be 1.5. meaning each percent would increase your mana by 1.5%, and a 5% node would increase it by 7.5%. compare that to only 4% from an 8% increased mana node. if you have only 100% mana increased, then it would be 1 to 1, meaning 5% reduced cost would give you 5% more mana, still more than 4%. if you are going for those reduced skill cost nodes, near the middle, between templar and marauder, or between marauder and dualist, those 10% that you can pick up, if you reserve some mana, would be very very valuable, at minimum passive cost for making your way to it. sure going there just for them is not good, especially since you would be losing regeneration that those increased mana passives would give. but what i am after in this post is not anything but their effectiveness.

by the way, i've made some calculations, and without giving up too much passives, you could run 5 auras (110% reserved from purity, haste, hatred, and add to that maximum level clarity and discipline) with a support gem and 10% reduced mana reservation (which with 110% reservation would make absolute wonders for you, reduce the requirement of mana base of this build by plenty). you can still take some offensive passives and have a well balanced character. would require about 650-750 mana base to give you some 150-200 unreserved mana to play with (calculated with 100% increased mana, less base needed with more percentage) and a lot would come from levels and intelligence. no need to use the battery (and sacrifice your energy shield) to do it.

50 people just read this and started a blood magic build.