Commander
Join Date: Jan 2013
Posts: 416
# 21
02-26-2014, 12:15 AM
Quote:
Originally Posted by baudl View Post
what you ignored is, that the ratio is not linear 1:10...at higher crtH, the less amount crtD you need to reach the dmg increase of a single crtH %.

the optimum ratio moves from 1:10 at <10% crtH to 1:5 at around 20% crtH and so on....you can imagine that at 99% crtH the ratio looks more like 10:1 than 1:10.

i'd asume that at 50% crtH the optimum ratio is 1:1
The ratio does not change one bit. It stays 10 to 1 til you cap 100% crit hit. Example

example using 1000 base damage

51% crit hit 490% crit sev
1000*49%=490
1000*(1+490%)*51%=3009
490+3009=3499

50% crit hit 500% crit sev
1000*50%=500
1000*(1+500%)*50%=3000
500+3000=3500

49% crit hit 510% crit sev
1000*51%=510
1000*(1+510%)*49%=2989
510+2989=3499

Note the most optimal is at 50% and 500% crit severity. And it's always a 10 to 1 ratio. Just need to always remember to add 1+x% sev. So 100% is actually 200% or x2. and 200% is actually x3, etc.

Also while the OP's math is correct with his examples i can't agree with the if critH>20% use x console. By itself the tac consoles are 5 to 1. So you use the best consoles eg the crit Hit ones and then use your weapons mods as they are 10 to 1 to get you at that sweet spot. And that typically means using critDx2 or crtDx3 weapons.

Let's take my rom toon for example. Stats are as follows.

ship/attack/acc=25%
ship/attack/crit hit=22.7%
ship/attack/crt sev=113.2%
energy weapon spec 99 points
additional crit hit 2%
additional crit sev=25%
opponents def=-15%
To hit=128.57%
acc overflow
additional crit hit=28.57*.125=3.57%
additional crit sev=28.57*.5=14.28%
Total crit hit=28.28%
Total crit sev=152.48%

Using rom plasma beams with critDx2 for an extra crit sev=40%

Using 5x 1.6% crit hit consoles

36.28% crit hit 192.48% crit sev
1000*63.72%=637
1000*(1+192.48%)*36.28%=1061
637+1061=1698

Using 5x 8% crit sev consoles

28.28% crit hit 232.48% crit sev
1000*71.72%=717
1000*(1+232.48%)*28.28%=940
717+940=1657

As you can see the crit hit consoles won in this situation by 2.47%. So the answer is never quite that easy.

Last edited by dragonsbite; 02-26-2014 at 01:00 AM.
Career Officer
Join Date: Jun 2012
Posts: 270
# 22
02-26-2014, 05:13 AM
Quote:
Originally Posted by frtoaster View Post
You made a mistake. The expected damage is

n * crth * d * (1 + crtd) + n * (1 - crth) * d.
good spot Im glad somebody read my opus

I did however accounted for crtd value being 100% + x. However, a correction of

dmg = x * (1 + y) + (1-x)

will not change the shape of the function.
STF Elite Advanced Tactics
Captain
Join Date: Jun 2012
Posts: 2,559
# 23
02-26-2014, 07:00 AM
Quote:
Originally Posted by dragonsbite View Post
*snip*
yes i was wrong in my statement that the optimum ratio is not linear, i'm sorry for that. It was also not what i meant.

what i was trying to say was, that once you get over a certain point of crtH, your crtD falls off, because accumulating crtH is much easier.
This results in a deveation from the 1:10 ratio for optimal dmg.

so the question remains, or rather the statement i made to solve this question: How much more crtH do i need to compensate for the discrapency between crtH and crtD if i'm unable to increase crtD.

the further up i go with crtH, without the ability to increase crtD, the more crtH i need, to get the same results as if i was able to get my crtD up within a 1:10 ratio.

That's what i meant with "the ratio is not linear"...the best crtH/crtD ratio remains 1:10...BUT in the case of STO, where there is a cap to crtD stacking due to itemshortage, the question of how much crtH i'd need remains.

to get to amount of dmg X, you need y% crtH and z% crtD...the optimum way is y=1 and z=10...so much is clear.
but since z can't be raised indefinately due to the itemisation and shortage of crtD items, consoles, skills,...
To achieve X, how large does y need to be, when z is constant (or in our case can't be increased)

would this be linear? wouldn't it mean that every bit of crtD i get to fill up the gap between the crtH and crtD imbalance (1:10 ratio) be many times more valuable than crtH initialy was?


your calculations demonstrated, that as long as you remain inside the 1:10 ratio and add and substract crtH or crtD accordingly, the dmg output does not change.

i poorly formulated my statement on which you responded, proberbly because the subject confuses me myself a lot. i hope this new post is somewhat clearer.
Go pro or go home

Last edited by baudl; 02-26-2014 at 08:21 AM.
Captain
Join Date: Apr 2013
Posts: 3,246
# 24
02-26-2014, 09:00 AM
Quote:
Originally Posted by dragonsbite View Post

So you use the best consoles eg the crit Hit ones and then use your weapons mods as they are 10 to 1 to get you at that sweet spot. And that typically means using critDx2 or crtDx3 weapons.



So how do you account for this...

BTW, just a little experiment

=====TEST FIFTEEN (yes i really did 11 more i didnt publish trying to prove myself wrong) =========

Romulan with 27 CritH 90 severity (reputations, SROs, universal consoles, etc, its very easy to get here with a romulan, this is actually one of mine)

1 + (.27*.9) = 1.243

Same romulan with critDx3 mods on his weapons

1 + (.27*1.5) = 1.405

Same romulan add 5x locator

1 + (.35 * 1.5) = 1.525

Same romulan with 5x exploiter instead

1 + (.27*1.9) = 1.513

Same romulan with 4xE 1xL (lets see if we can beat 1.525)

1+ (.286*1.82) = 1.5205

Same romulan with 3xE 2xL (This surpasses 5x locators)

1 + (.302*1.74) = 1.52548

Same romulan with 2xE 3xL (Optimum)

1 + (.318*1.66) = 1.52788

Same romulan with 1xE 4xL (going down again)

1 + (.334*1.58) = 1.52772



This romulan with 27 critH and 90 severity who then adds critDx3 weapons needs to use 2 exploiters and 3 locators to achieve optimal DPS. Even with critDx3 weapons, exploiters are still good (especially when mixed with locators to achieve the closest 5:1 ratio D/H). Locators alone actually hurt the DPS of the build.
Vice Admiral Rylana - U.S.S. DNT Omega X || Vice Admiral Lyzara - I.R.W. DNT Omicron X
Vice Admiral Kailiana - R.R.W. DNT Theta X || Vice Admiral Talina - I.R.W. DNT Tau X
Vice Admiral Victoria - U.S.S. DNT Upsilon X || Lt. General Dannika - I.K.S. DNT Sigma X
Vice Admiral Shondra - R.R.W. DNT Alpha X || Lt. General Rosanna - I.K.S. DNT Iota X
=== Vice Squad/House of Tlhap-Jen ===

Last edited by rylanadionysis; 02-26-2014 at 09:06 AM.
Captain
Join Date: Jun 2012
Posts: 1,214
# 25
02-26-2014, 09:50 AM
Quote:
Originally Posted by baudl View Post
yes i was wrong in my statement that the optimum ratio is not linear, i'm sorry for that. It was also not what i meant.

what i was trying to say was, that once you get over a certain point of crtH, your crtD falls off, because accumulating crtH is much easier.
This results in a deveation from the 1:10 ratio for optimal dmg.

so the question remains, or rather the statement i made to solve this question: How much more crtH do i need to compensate for the discrapency between crtH and crtD if i'm unable to increase crtD.

the further up i go with crtH, without the ability to increase crtD, the more crtH i need, to get the same results as if i was able to get my crtD up within a 1:10 ratio.

That's what i meant with "the ratio is not linear"...the best crtH/crtD ratio remains 1:10...BUT in the case of STO, where there is a cap to crtD stacking due to itemshortage, the question of how much crtH i'd need remains.

to get to amount of dmg X, you need y% crtH and z% crtD...the optimum way is y=1 and z=10...so much is clear.
but since z can't be raised indefinately due to the itemisation and shortage of crtD items, consoles, skills,...
To achieve X, how large does y need to be, when z is constant (or in our case can't be increased)

would this be linear? wouldn't it mean that every bit of crtD i get to fill up the gap between the crtH and crtD imbalance (1:10 ratio) be many times more valuable than crtH initialy was?


your calculations demonstrated, that as long as you remain inside the 1:10 ratio and add and substract crtH or crtD accordingly, the dmg output does not change.

i poorly formulated my statement on which you responded, proberbly because the subject confuses me myself a lot. i hope this new post is somewhat clearer.
This is why I'm against telling people about the 1:10 rule. It creates misunderstandings. Forget about whether the ratio is linear or not; in fact, forget about the ratio altogether. There isn't really an optimal ratio to achieve. If you want a rule to follow, then use this one:

If you are changing crit chance and crit severity only while keeping all other factors the same, then to maximize expected damage, maximize the product of your total crit chance and total crit severity.

Several players, including myself, have explained the mathematics behind that rule in other threads. The latest thread, by cerealplayer, is here:

http://sto-forum.perfectworld.com/sh....php?t=1034191
Commander
Join Date: Jan 2013
Posts: 416
# 26
02-26-2014, 10:10 AM
If you just look at the 2 types of tac consoles we all know they are at a 5 to 1 ratio as pointed out. This means that for the 2 consoles to be equal it must be x/(y/5)=1. Note x=crit hit and y=crit sev. So the formula to figure out which is best would be.

if x/(y/5)<1 use locator
if x/(y/5)=1 optimal
if x/(y/5)>1 use exploiter

However if you're just looking at your stats listed under ship/attack that wouldn't be enough data. You'd have to list your weapon mods. You'd have to figure out energy weapon spec which is a 12.5 to 1 ratio. And calculate acc overflow which is a 4 to 1 ratio. So the answer is never that simple.

And i have to mostly agree with the last poster. Still though it is best to use a calculator specific to this game for the most optimal results. There will be quite a few cases where exploiter consoles will be the better choice. Although you should try to avoid this with other choices.

Last edited by dragonsbite; 02-26-2014 at 10:20 AM.
Captain
Join Date: Jul 2013
Posts: 833
# 27
02-26-2014, 10:29 AM
Oh wow, seriously? That's what I call lying by math. The calculations are right, but since the starting values are wrong, the result is wrong.

What's with that low 80% severity? With an absolute base of 50%? (The real base in game is higher than that, btw) How often do you hear people having such low severity? News for you, weapon specialization skills add bigger portion severity than chance, and acc overflow adds 1% severity per point. You'll far overshoot that based on skills alone in a ship with no consoles using white non-ap non-dhc weapons that has no modifiers.

What's with that 20% chance? Good luck getting that high chance if you're not a Romulan with a whole bunch of superior operatives. Sure, you hear some non-romulans reaching it these days by filling their boat with every universal console under the sun, but that's after a full set of Locators...

Oh, and the 1:10 rule of thumb doesn't come from CrtH modifier being 2% and CrtD being 20%. It comes from the typical tac newbie asking questions in the academy: He would have crit chance in the high single digits or low teens, and severity already well above 100 based on specialization skill, acc, dhc, ap, and a mix of odd modifiers. It assumes he has no resources, but the right skill set for tac. While making no sense mathematically, this rule of thumb shows wisdom and intention to help by the person who came up with it, and not egoism and intention to deceive, like OP of this thread.

It's on a per weapon modifier basis. If you think the 1:10 comes from CrtD having an extra zero, that's 1:100 for you.

And it doesn't make a 1:5 rule for loc/exp. Exploiters having half the bonus that it's supposed to have merely makes it a bad console. The chumps who end up using [crtH] plus exploiters instead of [crtD] plus locators deserve the nerf they get.

Here:
http://comatoes.github.io/sto-crit-calc/
This is a solid calculator, and illustrates the issue nicely. Look through the options on left hand side. It applies whether you're in the masses or the extreme minority who has a full set of superior operatives. Unlike this thread, the calculator does not deceive, whoever you may be.

/thread
Commander
Join Date: Jan 2013
Posts: 416
# 28
02-26-2014, 10:43 AM
Quote:
Originally Posted by noblet View Post
Here:
http://comatoes.github.io/sto-crit-calc/
This is a solid calculator, and illustrates the issue nicely. Look through the options on left hand side. It applies whether you're in the masses or the extreme minority who has a full set of superior operatives. Unlike this thread, the calculator does not deceive, whoever you may be.

/thread
Actually that calculator does deceive. It has no option for being a Romulan which adds crit hit and crit severity at a 1 to 5 ratio. It's also using the wrong formula for acc overflow.
Captain
Join Date: Jul 2013
Posts: 833
# 29
02-26-2014, 10:56 AM
Quote:
Originally Posted by dragonsbite View Post
Actually that calculator does deceive. It has no option for being a Romulan which adds crit hit and crit severity at a 1 to 5 ratio. It's also using the wrong formula for acc overflow.
If your char is an Romulan Operative, did it not occur to you to add an "Opertive" trait? It's on the left hand corner. And the character trait is not 1:5 ratio, btw, nor any boff traits.

And there's no "formula" for acc overflow to call "wrong" or "right" for. It assumes stf conditions, and your acc skill is maxed out. If it's not, that's your problem. Acc over flow is subject to too many conditions, it won't be pinned down exactly by a calculator, or any person in this thread or elsewhere. ("Accruate" character adds a pecentage based on certain criteria, but not others. Overflow is broken for faw atm. And how targets fly, when you hit them, are completely fluid. The calculator does the best job that can be expected of anyone on acc)

Last edited by noblet; 02-26-2014 at 11:05 AM.
Commander
Join Date: Jan 2013
Posts: 416
# 30
02-26-2014, 11:01 AM
Quote:
Originally Posted by noblet View Post
If your char is an Romulan Operative, did it not occur to you to add an "Opertive" trait? It's on the left hand corner. And the character trait is not 1:5 ratio, btw, nor any boff traits.

And there's no "formula" for acc overflow to call "wrong" or "right" for. It assumes stf conditions, and your acc skill is maxed out. If it's not, that's your problem.
Of course conditions will effect acc overflow. But there is a right and a wrong and the calculations are wrong for acc overflow. If you want to point people to that calculator you should maybe point out the inaccuracy of acc overflow by telling them, to ensure they don't use that portion of it. But as acc overflow calculations are incorrect the weapons with acc are listed higher then they should be. So even choosing not to enter your accuracy will still show incorrect results for weapons with acc. And that you can't ignore. So you'd be better off using something else or point out to not use acc overflow and devalue or ignore the weapons with acc? Geez this just gets more complicated. Better off if the calculator was redone.

And you're correct the operative trait is not 1:5 it's 1:2.5. ty for the 2 corrections.

Last edited by dragonsbite; 02-26-2014 at 11:30 AM.
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off


All times are GMT -7. The time now is 04:30 AM.