Power vs Critical Chance

[mastirkal]

I recently wrote on the wall in the healer subforums under the BiS thread and it got me thinking of exactly how useful critical chance is compared to power. The following is what I wrote. The second part is more accurate, but I need to check on swtor to see the exact math.

 

For healers the power for healing is 1:0.17 knowing this, if we have a crit rate of 76.28(300 surge) or 77.74 (360 surge) it would take more than 44.870 (45.72 for 360 surge) points of crit per one percent of crit chance to make power worth more than critical chance. Until you hit the point of critical chance costing more than 44.870 or rather 44 points, critical chance is always more valuable than power.

 

7.628/0.17= 44.870 power rating = 7.628 bonus healing power

 

The numbers for DPS is 1:0.23 meaning critical chance diminished once it takes more than 33.165 (33.8 for 360 surge) points in crit to equal a single percentage boost.

 

If my math on this are incorrect please tell me. I'm trying to find the min max just like everyone else.

 

If my math is in fact correct, you can go to 472 critical rating and that is the quote on quote maximum you can achieve with 63 gear without it becoming less potent than power.

 

http://swtor.askmrrobot.com/character/dd8a9f71-5acb-46a4-9d04-325e93b8b10c

 

The above is my build and what I believe to be the best high crit build achievable for sorc healers. Mr.Robot's math for the endurance and willpower is off by 10 points each way and I haven't the slightest idea why. It also lost 1% between the sage and the sorcerer for no apparent reason. You should be able to achieve 45.54% critical with this build.

 

 

AGAIN if I miscalculated something, please tell me. I'm simply trying to find the best PVE build I can. This is also only for sorcerer healers, I haven't checked out the other classes.

Edited February 19, 2013 by mastirkal

[CaptainApop]

Neat. I'll probably check out the maths when I have a more than 5 minutes to spare but if you could stick your workings and justifications into spoiler tags or something I think that'd help us double check your work.

Or if you used a spreadsheet of some kind I'd definitely appreciate a link.

 

It's worth noting that there is a secondary component to the crit vs power argument that isn't necessarily raw throughput. Power raises the guaranteed "floor" of your healing whereas crit raises the potential "ceiling" . You need both.

In addition to class based secondary bonuses (100% crits on sorcs, energy return on ops etc etc)

 

The are also other class based concerns. For example sorc bubble benefits from bonus healing but cannot crit.

Edited February 19, 2013 by CaptainApop

[mastirkal]

That is a valid concern, for sorc healers the two abilities that sufferfrom a high critical are a force bent dark heal and static bubble. The first is no big deal as you'll never really need it except emergencies. The second is a bit more of a concern, but overall it won't burn your house down and if the rumored changes to static bubble in the coming expansion are true, that concern might not be a concern at all if Static barrier begins healing people, for now however it is a valid downside.

 

The base vs the ceiling I did take into account. This is where my math comes in. If I did a single heal over a hundred or even a thousand times the law of big numbers would show the average heal would be higher. This is excluding outside influences that plague sorc healers the most obvious being overhealing. Sorcs have an average effective rate that can range anywhere from 60-85%. With high crit I have no idea where the number sits without a high healing rate area like TfB or EC NiM situations.

[RuQu]

One of the big things you are overlooking is that different abilities scale differently with healing power. Each ability has a base range, and then a multiplier that is applied to your bonus healing. If you have two abilities with a base average of 500 and one has a multiplier of 1 and the other of 2, the value of Power will vary drastically based on how often you use each of those abilities. In the case of Commandos, Hammer Shot exactly equals your Bonus Healing. Medical Probe has a multiplier of 2.72. So 1 Power = 0.17 extra HS healing, but 0.46 extra MP healing. If those were my only abilities, the actual return in Healing Done would be the weighted average based on typical cast counts.

 

If you want to do a proper comparison of the two, you need to first take a look at your standard rotation, then find the Expected Value for each ability factoring in your Bonus Healing and Crit/Surge values, and then recalculate that twice, once each for an increase in power and an increase in crit.

 

Assuming an ability does "H" healing on average with no crit, then we get:

 

Expected Value = [(1-crit%) + (crit%)(surge%)]*H

 

If you want to get a more accurate answer, try this:

 

1. Identify a typical rotation for your class. Or use a parser and look at your cast counts over a week's worth of FP/Ops runs.
   
2. Use TORHead to find the "StandardHealthPercentMin, StandardHealthPercentMax, HealingPowerCoefficient" values. These are typically hidden under "Effect Details." Click it to expand.
   
3. Take the average of the Min/Max values there and multiply by 7025. That's your base average.
   
4. Multiply Bonus Healing by HealingPowerCoefficient and add that to your base average. This is your non-crit heal for that ability, H.
   
5. Find the EV for the ability, using the equation above.
   
6. Multiply EV by the cast count from your parser.
   
7. Repeat for all abilities.
   
8. Divide by the total cast count.
   

This is your personal average heal.

 

Now, look at your projected stat budgets stacking Power and Crit, and recalculate the above personal average. Then, take the difference between the initial baseline calculation and each of these two and see which is higher.

 

As you can see, it is not anywhere near as simple as your initial comparison, but it's also not at all impossible.

 

For Commandos, the necessary ability values are here in spoiler tags, updated to patch 1.2 which should be fine since I hear there have been very few changes since then to healing.

 

AP = BH*1.94 + 629.05

MP = BH*2.72 + 915.96

TP = BH*0.481 + 170.04

BI = BH*2.28 + 807.69

KB = BH*1.08 + 382.59

HS = BH

 

Edited February 19, 2013 by RuQu

[RuQu]

Here, I made a spreadsheet you can copy and edit.

 

The parts in grey don't need changing. The parts in white are what you should enter based on your character. It's initial stats on the left, your new stats on the right (ie swapping crit for power), and the ability info from TORHead below.

 

I entered arbitrary values for the stats, and used the Commando ability values. Cast counts were arbitrary. You should use yours from a parser after running some FPs or Ops.

 

The final value at the bottom goes blue if the change is positive, red if negative, white if no change. If you are comparing two changes, just note the first value in the blue box and compare to the value after changing stats again.

 

Note that this does have a spot for "Crit% from skills," but it does not take into account any ability specific crit multipliers, such as Diagnostic Scan for Scoundrels. Hence the word "Basic" in the spreadsheet title.

 

[edit] The initial values for stats in there are from an old note I made about Tionese values shortly after launch. Swapping out 54 Power for crit was a loss of -5 healing per cast, replacing 54 crit with power was a gain of 5.

Edited February 19, 2013 by RuQu

[mastirkal]

Can you do one thing first, can you change that 57% crit rate to 76.26 and post the results again. My argument is completely base on having a surge rate of 300.

 

EDIT change the numbers to this for BiS

 

886 power, 340 crit

754 power, 472 crit

 

Both have 300 surge. Primary for a commando is 2434

Edited February 19, 2013 by mastirkal

[RuQu]

It's a google doc. Just select the menu option to create a copy and you can edit it to your heart's content. Swap out Sorc abilities, use the cast counts from your parse instead of the random numbers I chose, etc

[mastirkal]

It's a google doc. Just select the menu option to create a copy and you can edit it to your heart's content. Swap out Sorc abilities, use the cast counts from your parse instead of the random numbers I chose, etc

 

The numbers on the actual heals won't matter, the input and changed variables will. Either way, I'm simply trying to find the best build or at least a different, yet viable build. I'm goog to plug the numbers when I get to an actual computer.

Edited February 19, 2013 by mastirkal

[RuQu]

The numbers on the actual heals, and the cast counts per boss fight / Op / rotation absolutely matter.

 

Did you not read my longer post at all explaining why your numbers are not accurate and expanding o the other poster's comment about class variation?

 

Sages have no baseline heal like Hammer Shot that only heals for bonus healing. Commandos have no heals that can't crit.

 

If you don't either parse your cast counts or at least estimate a rotation, you won't get anything close to an accurate estimate on the increase per point of bonus healing.

 

You might as well just guess at that point. Go with what "feels right." You certainly can't say that math supports your conclusion.

[mastirkal]

The numbers on the actual heals, and the cast counts per boss fight / Op / rotation absolutely matter.

 

Did you not read my longer post at all explaining why your numbers are not accurate and expanding o the other poster's comment about class variation?

 

Sages have no baseline heal like Hammer Shot that only heals for bonus healing. Commandos have no heals that can't crit.

 

If you don't either parse your cast counts or at least estimate a rotation, you won't get anything close to an accurate estimate on the increase per point of bonus healing.

 

You might as well just guess at that point. Go with what "feels right." You certainly can't say that math supports your conclusion.

 

You make a very valid point ill have to look up the math to the all of the heals besides dark heal for a sorcerer healer.my parsing I have done is probably off kilter with some of your examples so I do thank you for helping me have a better understanding for what exactly in dealing with.

[RuQu]

Well, what you choose to parse is up to you. Pick whatever challenged you the most. Boss fights easy but trouble keeping people alive on trash? Then parsing only the bosses is probably a bad idea.

 

The important thing is to try and accurately capture your ability usage and the differences in how each ability will respond to stay changes, an then combine those two things to reflect your overall change in healing throughput.

[mastirkal]

I'm curious where you find the secondary stat. I haven't the foggiest idea and I know I need it because the critical chance percentage is off.

 

What I have so far I can see that I was very off when I saw what I did. The sorc healer is best at 340 critical chance rating rather than my estimate of 472.

 

https://docs.google.com/spreadsheet/ccc?key=0AkNtAK3jzVCgdEgxTF9IV21KR0liN1JXQXRHYVo5emc&usp=sharing

 

I also fixed your surge rating, it was slightly off.

[RuQu]

I'm curious where you find the secondary stat. I haven't the foggiest idea and I know I need it because the critical chance percentage is off.

 

What I have so far I can see that I was very off when I saw what I did. The sorc healer is best at 340 critical chance rating rather than my estimate of 472.

 

https://docs.google.com/spreadsheet/ccc?key=0AkNtAK3jzVCgdEgxTF9IV21KR0liN1JXQXRHYVo5emc&usp=sharing

 

I also fixed your surge rating, it was slightly off.

 

Your secondary stat varies by class. For a Sage/Sorc, Willpower is your primary. Strength is your secondary. Scoundrels are Cunning/Aim, Troopers are Aim/Cunning.

 

I used my old equations from 1.2. Did they adjust the DR curve of Surge again? What are the new exponent constants?

 

Your spreadsheet privacy settings are set so the link doesn't work.

[mastirkal]

Your secondary stat varies by class. For a Sage/Sorc, Willpower is your primary. Strength is your secondary. Scoundrels are Cunning/Aim, Troopers are Aim/Cunning.

 

I used my old equations from 1.2. Did they adjust the DR curve of Surge again? What are the new exponent constants?

 

Your spreadsheet privacy settings are set so the link doesn't work.

 

I did not notice that, anyway it has been set to public, but still the critical percentage is off, it should be sitting at 36.30 for the second rating. What your critical rate was sitting at was 85% at 300 surge. It is now 76.28 or rather 77.74 at 360 surge.

 

EDIT correction 36.31*

Edited February 20, 2013 by mastirkal

[RuQu]

I did not notice that, anyway it has been set to public, but still the critical percentage is off, it should be sitting at 36.30 for the second rating. What your critical rate was sitting at was 85% at 300 surge. It is now 76.28 or rather 77.74 at 360 surge.

 

EDIT correction 36.31*

 

I honestly have no idea what you are trying to say.

 

By default, with no gear or datacrons, your Secondary stat (ie strength) should be 100 at level 50. Looking at your spreadsheet, you have it set to zero.

 

I also see you trying to do some odd math where you are misusing the Expected Value field (B14). That field is the size of your average heal, taking into account the crit rate and surge bonus. It is not the size of an average non-crit, it is not the size of the average crit, it is the expected mean value over a very large number of casts.

[mastirkal]

I honestly have no idea what you are trying to say.

 

By default, with no gear or datacrons, your Secondary stat (ie strength) should be 100 at level 50. Looking at your spreadsheet, you have it set to zero.

 

I also see you trying to do some odd math where you are misusing the Expected Value field (B14). That field is the size of your average heal, taking into account the crit rate and surge bonus. It is not the size of an average non-crit, it is not the size of the average crit, it is the expected mean value over a very large number of casts.

 

Caught me at a bad time, I was adding the bonuses you get from the skill tree, that being a 8% buff to all heals, 10% for the AoE, 20% for the static bubble, 30% bonus to unnatural preservation (+8% because it's only used on yourself), and with that all the numbers seem to fall in line. As for the secondary value, isn't that only useful for hybrid classes where both are considered? aka assassin, marauder, jugg, and power tech? I also have not touched the expected value at all.

Edited February 20, 2013 by mastirkal