Sawbones stat weight calculator

[RuQu]

With the assistance of some Operatives and Scoundrels on the healer forum, bobudo in particular, I have added a Sawbones module to my stat weight calculator project that previously was for Combat Medics, Bodyguards and tanks. It is available here.

 

If you press "Calculate HPS" the rotation it uses is spit out in one of the output windows, so you can examine it if you want. I recommend cutting it out and saving as a .csv file for easier analysis.

 

The rotation is decided on the fly based on specific rotational logic, so changing your stats will change the rotation (alacrity in particular, the others not so much).

 

There is a stat weight calculator, you can choose what stat to normalize to.

 

There is an upgrade calculator for gear comparison.

 

Rotations can be run as fixed time or fixed cast count.

 

There is also a "Compute Alacrity" button which will spit out lots of relevant data for analyzing the efficacy of your class at different levels of alacrity. To see the sort of analysis that can come of that button, see this post.

 

Your set bonuses aren't implemented because, frankly, they are terrible and don't have any use in a single target rotation (2pc), or at all (4pc).

Edited February 17, 2012 by RuQu

[Teirdome]

Cool stuff! Messed around with it a little while waiting on a build. I'm not much of a cpp guy, otherwise I'd try to contribute.

[RuQu]

Cool stuff! Messed around with it a little while waiting on a build. I'm not much of a cpp guy, otherwise I'd try to contribute.

 

Testing is always helpful, and currently still needs a lot of work just compiling the abilities and rotation used by other specs/classes.

[Azaranth]

Interesting. I applaud your efforts here. The presentation is superior to the tools that I use (mostly Excel), but there's a lot of magic happening behind the curtain here. A few questions about the tool:

 

1. Some of the numbers aren't lining up with the in-game tool tip. The estimated Crit Chance and Crit Size entries are close enough, but the Bonus Healing category is off by almost 3%.

 

2. Out of curiosity, why is the off-hand weapon damage relevant?

 

3. In what way is it weighing Accuracy? I see the word Accuracy greyed out, but it still gives a stat weight, which I don't understand. For future Scrapper/DirtyFighting calculations I'm guessing?

 

Finally, I want to caution people from staking too much on the 'perfect scenario' healing rotation information. While it's definitely useful to know, 99.99% of your healing is not done in these circumstances. Most of the time, you're dealing with damage spikes and damage droughts, fights where you have to cast on the run, and all sorts other environmental factors. I think this leads to people undervaluing Alacrity, which looks horrible when computing sustainable infinite heal rotations, but has much more practical value when trying to heal through Gharj's enrage.

Edited February 17, 2012 by Azaranth

[RuQu]

Interesting. I applaud your efforts here. The presentation is superior to the tools that I use (mostly Excel), but there's a lot of magic happening behind the curtain here. A few questions about the tool:

 

1. Some of the numbers aren't lining up with the in-game tool tip. The estimated Crit Chance and Crit Size entries are close enough, but the Bonus Healing category is off by almost 3%.

 

Can you list the numbers you are entering, and the values you see in game? When I enter 100 Cunning, I get a value of 14 Bonus Healing. When I enter 100 Power or Tech Power I get 17, when I enter 100 Cunning + 100 Power, I see 31. All of those are correct, and that field is fed by a very simple equation. If you list all your stats for me and what you see in game I'll see what I can do to make them line up.

 

Check if you have the consular buff checked. That will increase your Cunning by 5%, which in turn will increase your bonus healing by slightly less (since it contributes 0.14 vs 0.17 from Power).

 

2. Out of curiosity, why is the off-hand weapon damage relevant?

 

Heh. I added those fields in and forgot to turn them off. They are there for when Scrapper and dirty fighting eventually get implemented, and for copying the ui over for Gunslingers.

 

3. In what way is it weighing Accuracy? I see the word Accuracy greyed out, but it still gives a stat weight, which I don't understand. For future Scrapper/DirtyFighting calculations I'm guessing?

 

Good catch. Turns out the Alacrity and Accuracy boxes got inverted as I moved some UI elements around. That was the Alacrity weight.

 

Uploading a new version now with your bug fixes implemented. 1.0.7b3

 

Finally, I want to caution people from staking too much on the 'perfect scenario' healing rotation information. While it's definitely useful to know, 99.99% of your healing is not done in these circumstances. Most of the time, you're dealing with damage spikes and damage droughts, fights where you have to cast on the run, and all sorts other environmental factors. I think this leads to people undervaluing Alacrity, which looks horrible when computing sustainable infinite heal rotations, but has much more practical value when trying to heal through Gharj's enrage.

 

First, yes, obviously these tools cannot model the complexity you see in game. They do, however, provide a good foundation for making decisions. Knowing that a change in Crit will give you, at a certain gear level, more benefit than a change in Power is true whether you are standing still or not, but some healers will always fear crit for its evil tendency to only happen when you don't need it and never when you do. Similarly, if people are dying before your spells can land, you should probably load up on some more Alacrity, even if something else is objectively a bigger HPS increase. A 10s cast that heals for 80 billion HP is worthless if the target dies after 5 seconds.

 

Acknowledging that there are valid amounts of Alacrity that are useful, Alacrity is absolutely terrible for any non-Sage/Sorc healer.

 

I will keep this post confined to Scoundrels/Operatives, but there is a lengthy analysis of just how terrible it is for Mercs, and even worse for Commandos located here. Yes, there is a difference between the two.

 

As for Scoundrels, the first thing about Alacrity that matters is that it does not decrease the GCD for instants. That is well known, and we can move past it to more important issues.

 

The second is that it makes cast-time spells less efficient. This chart shows the changes in Effective Cost of UWM and DS with Alacrity. Yes, fitting in extra DS casts will offset some of the loss in regen, but from 0-39% Alacrity, you go from a base regen of 18 to 11 during a cast. Yes, 39% is fully capped and massively overstacked, but the line is linear. You would need to fit in an extra DS for every 3 you cast at 0% just to maintain Energy neutrality from lost DS return, not even accounting for the fact that UWM goes from costing a net 13 to a net 17.7.

 

Next, let's examine the changes in rotational composition. This chart shows the change in cast counts for UWM, EMP, and DS, and the total change. As you can see, almost all of the change comes from increasing the number of times DS is used. Alacrity isn't letting you do more in the same time, it is letting you do what you would have done before in a shorter time, and then sit around and cast DS. This is why there are benefits to Alacrity in PvP, where fights are short and Energy is nearly irrelevant, but it does very little for your throughput in PvE content. How often do you say "Man, if only I'd gotten in an extra DS cast, that would have gone differently..."? Never, that's how often.

 

Count yourselves lucky that you actually see a minor but positive change in UWM casts with Alacrity. Mercs hold steady with their heals, but, due to differences in regen window sizes, Commandos actually see a 14% reduction in big heal casting from 0-39% Alacrity, and it is a steady drop along the way.

 

We now need to examine the other implications of this increase in DS while maintaining near steady UWM cast counts. This chart shows the percentage of your rotation comprised of either UWM or DS. I didn't bother to plot EMP, but it should roughly mirror UWM since SRMP UH procs are alacrity independent. We see that UWM drops from 40% down to 35%, and DS rises from 11% to 23.5%. Now realize that these two spells have different coefficients, they scale differently with Power. Adding 100 Power adds 17 Bonus Healing. With their coefficients, that means DS will heal for only 17 more, while UWM heals for an extra 46.24. So as the percentage of your casts made up of UWM (your highest coefficient) drops and DS (your lowest) rises, your average coefficient for your rotations decreases, which means you scale slower with gear upgrades. Alacrity actually makes all of your other stats worth less.

 

Reasons Alacrity is Terrible for Scoundrels

1) Doesn't affect instant abilities or HoTs.

2) Makes spells less energy efficient.

3) Even at extreme levels, only earns you extra casts of your weakest ability.

4) Decreases value of all other stats, makes all your upgrades weaker.

5) Opportunity Cost: you could have picked something else.

[Azaranth]

Can you list the numbers you are entering, and the values you see in game? When I enter 100 Cunning, I get a value of 14 Bonus Healing. When I enter 100 Power or Tech Power I get 17, when I enter 100 Cunning + 100 Power, I see 31. All of those are correct, and that field is fed by a very simple equation. If you list all your stats for me and what you see in game I'll see what I can do to make them line up.[/Quote]

 

Ah, you're right. The consular buff was checked. Sorry for the false alarm.

 

Good catch. Turns out the Alacrity and Accuracy boxes got inverted as I moved some UI elements around. That was the Alacrity weight.[/Quote]

 

What is Alacrity weighted against? Other healing values can be analyzed with regard to the average amount of healing boost given. Is Alacrity weighted with regard to it's impact on the 'perfect rotation' scenario?

 

Also, does the weight value of crit factor in energy returns for Diangostic Scan, and how it impacts the perfect rotation?

 

Excellent work so far. I'm simply making sure the methodology matches up with the results from my spreadsheet, before I abandon my spreadsheet entirely and move over to your calculator.

 

On to the Alacrity Discussion!

 

First, yes, obviously these tools cannot model the complexity you see in game. They do, however, provide a good foundation for making decisions. Knowing that a change in Crit will give you, at a certain gear level, more benefit than a change in Power is true whether you are standing still or not, but some healers will always fear crit for its evil tendency to only happen when you don't need it and never when you do. Similarly, if people are dying before your spells can land, you should probably load up on some more Alacrity, even if something else is objectively a bigger HPS increase. A 10s cast that heals for 80 billion HP is worthless if the target dies after 5 seconds.[/Quote]

 

I couldn't agree more. I'm merely wary of assigning weight values to Alacrity, because these models analyze a 'Perfect Scenario' healing rotation which is, for the most part, entirely hypothetical.

 

I'd posit to you that this perfect scenario almost never happens in game. This doesn't negate the weight values for Cunning, Power, Surge and barely impacts Crit. However, it really invalidates the weight value for Alacrity.

 

Don't mistake my meaning. It is important for us to model Alacrity. I just put forth that the modelling numbers and weights are essentially a 'worst case' scenario for the stat, and that in almost all practical applications, alacrity is more valuable than these numbers will represent.

 

As for Scoundrels, the first thing about Alacrity that matters is that it does not decrease the GCD for instants. That is well known, and we can move past it to more important issues.[/Quote]

 

Agreed

 

The second is that it makes cast-time spells less efficient. Yes, fitting in extra DS casts will offset some of the loss in regen, but from 0-39% Alacrity, you go from a base regen of 18 to 11 during a cast. Yes, 39% is fully capped and massively overstacked, but the line is linear. You would need to fit in an extra DS for every 3 you cast at 0% just to maintain Energy neutrality from lost DS return, not even accounting for the fact that UWM goes from costing a net 13 to a net 17.7.[/Quote]

 

This is what I mean: that's a fabricated scenario. For this perfect healing, full sustain scenario to be applicable, you need an extremely unlikely situation.

 

There's three possible scenarios:

 

1. Your healing output exceeds the rate of incoming damage. Perfect healing rotation is inapplicable, since you'd simply be wasting resources on overheal.
    
   
2. Your healing output is insufficient to match the rate of incoming damage. Other healing assistance is needed, else you wipe. If this additionally healing assistance now exceeds the incoming damage rate, you're back to wasting resources on overheal.
    
   
3. Your healing output EXACTLY matches the rate of incoming damage. Perfect healing equilibrium! However, maintaining this requires the absence of environmental effects which can interrupt it (running from AoEs, pushback from monsters, mind traps, etc), or inconsistent lag.
   

 

I think we can agree that this is an EXTREMELY unlikely scenario.

 

More often, you top a tank off, run from an AoE, blow a few GCD's to heal raid damage, back to channelling on the tank, etc. In practice, casting speed gives you a much wider margin of error for unlucky AoEs, interrupts and everything else. There's no way for our computer model to reflect this.

 

You can go boss by boss, and in none of them is the Perfect Rotation really applicable. In the Annihilation Droid fight, a 'Perfect Rotation' isn't going to happen. Tank is knocked out of range, healers need to dodge missile storms and ground targets. For Gharj, Again, this fight has frequent knockbacks and mandatory movement for the entire raid. The Perfect Rotation is inapplicable. Jarg and Sorno: Grappling Hooks and Missiles frequently interrupt you. Damage is not consistent, there's more raid healing to be done than tank healing, really. Etc, etc. The only fights where you're even able to stand in one place long enough to establish a healing rotation is the Fabricator, and maybe Bonethrasher.

 

Again, I do think it's important to model these healing simulations. I'm only bringing this up to emphasize that the Alacrity weights are going to be essentially a worst-case-scenario, and that 99% of the time Alacrity is actually more beneficial than the sim suggests.

 

(on the drawbacks of Alacrity): The Opportunity Cost: you could have picked (a different stat).

 

This is a the crux of the issue. It's my personal opinion that the first 300 points of Alacrity are extremely cost efficient. I think it's a good idea to aim for 300 points in each of the secondary stats, minimizing the effects of diminishing returns, before you consider pushing any single stat higher than that.

 

I worry that people will simply look at the stat weights in your sim, and conclude that because Alacrity under-performs in the hypothetical Perfect Rotation scenario, it's a better decision to push Crit from 500 -> 700 (a 3% jump) rather than Alacrity from 0 -> 200 (a 9% jump).

 

Again, great work on the sim. It's a great tool, and I'm not suggesting you change the way that you weight Alacrity. I'm suggesting that people need to understand the context in which these weights are given, and that Alacrity is a more useful stat than this sim suggests.

[RuQu]

First, on the weighting, they are all done by adding +1 to their respective stat, then essentially clicking the Calculate HPS button internally, logging the value and comparing it to the base HPS.

 

This can lead to spikiness in Alacrity's HPS since the change in a couple of points could mean getting in a final big heal or slipping in a final Hammer Shot or Diagnostic scan, which will push HPS up or down respectively. In comparison, the other stats all use an identical rotation, while Alacrity causes the rotation to change.

 

Yes, DS crits are modeled, as are SRMP UH procs, by calling a random number generator and comparing that to (your crit chance + 24) for DS or 30 for SRMP. This RNG might result in slight variation with different runs, but I honestly think the variation gets lost when compared to the differences you will face in game vs the static situation modeled, so I don't see the point of wasting the time by having "Calculate HPS" average 10000 runs.

 

As for the scenarios you presented, if we rephrase them slightly those scenarios are:

1) You are overgeared for this content, standing around, and bored.

2) People are dying.

3) You are in level appropriate content.

 

In my experience, (3) happens fairly often. I know I'm healing more often than not.

 

And, let's remember that a faster heal is a weaker heal, so not only does it cost more, but you have to cast more of them to get the same job done.

 

Let's look again at my five conclusions:

 

1) Doesn't affect instant abilities or HoTs.

2) Makes spells less energy efficient.

3) Even at extreme levels, only earns you extra casts of your weakest ability.

4) Decreases value of all other stats, makes all your upgrades weaker.

5) Opportunity Cost: you could have picked something else.

 

How do these conclusions change in a high mobility situation where you are constantly in and out of fire or chasing down a tank who gets knocked around?

 

1) This one gets worse! You are relying more on instants when you run and that HoT may be the difference between life and death. Alacrity has zero benefit on these now key abilities.

 

2) If you can regen while running around, ie dodging fire means no damage so its a break, this may not matter anymore...unless of course someone steps in some of that fire and you need to do some fast burst healing to get them up while doing your standard heals you'd been maintaining all along. Now that burst was even more punishing because everything costs more. I'd say we combine the two and call this one even, Alacrity is just as bad as during a static fight on this account.

 

3) Us Commandos can cast our weakest filler ability on the move, you can't, so the energy regen you might gain from those extra DS crits are lost...so you still can't cast any more UWM/EMP cycles, UWM is still less efficient...yup this one still stands.

 

4) This one might actually go away in a high mobility fight, as any fight mechanic that has you moving so you cant cast allows you to regen during that time, instead of by using DS, so UWM/EMP might make up a larger percentage of your total casts just because you never have time to cast DS. Of course, that's true regardless of your alacrity level, so it is just no longer a complaint, but not an endorsement.

 

5) Opportunity Cost. Alacrity is the only stat with negatives associated with it. So long as you have enough that people don't die waiting on UWM to cast, you should absolutely take that high DR crit over any value of Alacrity. And since you should normally keep 1 UH in reserve anyway, why didn't you use EMP in that rare emergency? Saving someone now is worth more than 3% healing loss for 3.5 seconds. Or invest in Power, it has no DR.

 

As for the weight assigned, I put an * next to it in the hopes people would click "*Note 1" and read the details there explaining the oddities of Alacrity and directing them to the boards.

[Azaranth]

First, on the weighting, they are all done by adding +1 to their respective stat, then essentially clicking the Calculate HPS button internally, logging the value and comparing it to the base HPS.

 

This can lead to spikiness in Alacrity's HPS since the change in a couple of points could mean getting in a final big heal or slipping in a final Hammer Shot or Diagnostic scan, which will push HPS up or down respectively. In comparison, the other stats all use an identical rotation, while Alacrity causes the rotation to change.

 

Yes, DS crits are modeled, as are SRMP UH procs, by calling a random number generator and comparing that to (your crit chance + 24) for DS or 30 for SRMP. This RNG might result in slight variation with different runs, but I honestly think the variation gets lost when compared to the differences you will face in game vs the static situation modeled, so I don't see the point of wasting the time by having "Calculate HPS" average 10000 runs.[/Quote]

 

Great. Thanks for the clarification, and I agree with your conclusions.

 

As for the scenarios you presented, if we rephrase them slightly those scenarios are:

1) You are overgeared for this content, standing around, and bored.

2) People are dying.

3) You are in level appropriate content.

[/Quote]

 

That's an inaccurate representation of my statement. It has nothing to do with level appropriate content. This is about windows of time where your heals are neither overhealing nor underhealing, and yet your're wasting ZERO time not casting heals. If you have ANY space between heals, then the entire concept of benchmarking stats by their efficiency in a zero-downtime healing rotation is invalid.

 

Again, I would submit that this never happens. We all have frequent gaps in our healing, whether it's due to topping off the tank, a lucky string of shield procs on the tank, running from AoEs, switching targets for touch up heals, clicking pylon buttons, etc.

 

Here's a challenge:

 

Put your numbers into the modelling tool, and determine the perfect, sustainable rotation for your gear levels. Then, Fraps a video of you maintaining this perfect sustainable rotation for the duration of an Operation level boss fight, without stopping or overhealing.

 

I'd bet a large pile of credits that you couldn't do that, simply because these healing rotations are hypothetical, and impossible to really exact in a real world scenario. Contrast this with a DPS rotation, where infinite sustainability and a flawless 5-minute scenario is a very realistic situation which can and does happen on almost any tank-and-spank boss.

 

1) This one gets worse! You are relying more on instants when you run and that HoT may be the difference between life and death. Alacrity has zero benefit on these now key abilities.[/Quote]

 

Agreed

 

2) If you can regen while running around, ie dodging fire means no damage so its a break, this may not matter anymore...unless of course someone steps in some of that fire and you need to do some fast burst healing to get them up while doing your standard heals you'd been maintaining all along. Now that burst was even more punishing because everything costs more. I'd say we combine the two and call this one even, Alacrity is just as bad as during a static fight on this account.[/Quote]

 

Disagree with your premise. It only makes a spell more energy inefficient if you exercise the opportunity to cast more of them. Alacrity does not force you to cast spells more frequently, but it does enable you to reap the benefits of casting that spell sooner in the activation cycle.

 

It's entirely possible that after running from an AoE, you need to throw a clutch heal to save someones life. Alacrity increases your chances of landing this heal in time, which no other stat can do. Energy efficiency doesn't matter in this scenario. And let's be honest, healing someone who stood too long in the fire is a large part of our job in current Operations.

 

3) Us Commandos can cast our weakest filler ability on the move, you can't, so the energy regen you might gain from those extra DS crits are lost...so you still can't cast any more UWM/EMP cycles, UWM is still less efficient...yup this one still stands.[/Quote]

 

Again, this only applies to some sort of long-term, perfectly sustainable healing rotation with no downtime. This scenario does not actually occur with any regularity. Far less regular than the 'clutch heal to save someone' scenario does.

 

4) This one might actually go away in a high mobility fight, as any fight mechanic that has you moving so you cant cast allows you to regen during that time, instead of by using DS, so UWM/EMP might make up a larger percentage of your total casts just because you never have time to cast DS. Of course, that's true regardless of your alacrity level, so it is just no longer a complaint, but not an endorsement.

 

5) Opportunity Cost. Alacrity is the only stat with negatives associated with it. So long as you have enough that people don't die waiting on UWM to cast, you should absolutely take that high DR crit over any value of Alacrity. And since you should normally keep 1 UH in reserve anyway, why didn't you use EMP in that rare emergency? Saving someone now is worth more than 3% healing loss for 3.5 seconds. Or invest in Power, it has no DR.[/Quote]

 

Again, I disagree with the premise. There's no negatives associated with Alacrity, excluding opportunity cost... and once you're reasonably well equipped, then the other secondary stats should be far enough up in the diminishing returns range that the opportunity costs of picking up Alacrity are pretty low.

 

If you can provide a video showing a real world application of the mythical Perfect Healing rotation in practice, then I'd accept some of these arguments. Until then, I think it's a bad idea to make itemization decisions based on impossible scenarios.

Edited February 17, 2012 by Azaranth

[RuQu]

Great. Thanks for the clarification, and I agree with your conclusions.

 

That's an inaccurate representation of my statement. It has nothing to do with level appropriate content. This is about windows of time where your heals are neither overhealing nor underhealing, and yet your're wasting ZERO time not casting heals. If you have ANY space between heals, then the entire concept of benchmarking stats by their efficiency in a zero-downtime healing rotation is invalid.

 

Perhaps some of our disagreement comes from me being a Commando. Hammer Shot is roughly identical to not healing. It provides no regen beyond the passive amount of not casting for 1.5s, unlike DS with its crits. So, from where I am sitting, the filler spells, HS and DS, take place in this "everyone's topped off" space. If you burn 30 energy and stand around or burn 30 energy and cast DS, you still need that recovery time for your resource, whether you are casting or not. Your time when Alacrity is non-punishing is only when you are both not needing to heal anyone AND at full Ammo/Energy. If you are still regenning, that regen period will be longer with alacrity than without.

 

 

Again, I would submit that this never happens. We all have frequent gaps in our healing, whether it's due to topping off the tank, a lucky string of shield procs on the tank, running from AoEs, switching targets for touch up heals, clicking pylon buttons, etc.

 

Here's a challenge:

 

Put your numbers into the modelling tool, and determine the perfect, sustainable rotation for your gear levels. Then, Fraps a video of you maintaining this perfect sustainable rotation for the duration of an Operation level boss fight, without stopping or overhealing.

 

I'd bet a large pile of credits that you couldn't do that, simply because these healing rotations are hypothetical, and impossible to really exact in a real world scenario. Contrast this with a DPS rotation, where infinite sustainability and a flawless 5-minute scenario is a very realistic situation which can and does happen on almost any tank-and-spank boss.

 

The model has a field to change the fight duration. Pick a window that you do believe, click Compute Alacrity, save it as a .csv and make some graphs.

 

Disagree with your premise. It only makes a spell more energy inefficient if you exercise the opportunity to cast more of them. Alacrity does not force you to cast spells more frequently, but it does enable you to reap the benefits of casting that spell sooner in the activation cycle.

 

Faster spells are weaker spells, therefore you need more of them for the same benefit. By not putting those points into anything else, they will heal for less, and by not scaling resource costs to alacrity, you can't afford to cast more.

 

It's entirely possible that after running from an AoE, you need to throw a clutch heal to save someones life. Alacrity increases your chances of landing this heal in time, which no other stat can do. Energy efficiency doesn't matter in this scenario. And let's be honest, healing someone who stood too long in the fire is a large part of our job in current Operations.

 

No other stat, no, but smarter play can. Your AoE is instant, and the HoT sticks to the players, not the ground. Drop it as the group starts to run. You should have 2 stacks of UH based on SRMP ticking while you ran, so use those to save the person. If its a real crisis, EMP will refund the UH and save them. Why are you casting UWM in that situation?!

 

For me, as a Commando, I'm spamming KB and HS as I run, so the person should have the KB shield and heal buff on them, and then I can hit them with BI if its up, an instant MP if that cooldown is up, or hard cast AP which will be off of cooldown since I just finished running and will take 1.4 seconds to land (we want ~5% alacrity). If they died despite all of that, they were probably being stupid and an extra 0.2 seconds off of AP wouldn't fix that. You can't heal stupid.

 

Again, this only applies to some sort of long-term, perfectly sustainable healing rotation with no downtime. This scenario does not actually occur with any regularity. Far less regular than the 'clutch heal to save someone' scenario does.

Use your clutch heal tools for clutch healing, don't shoehorn in a slow heal with bad stats.

 

Again, I disagree with the premise. There's no negatives associated with Alacrity, excluding opportunity cost... and once you're reasonably well equipped, then the other secondary stats should be far enough up in the diminishing returns range that the opportunity costs of picking up Alacrity are pretty low.

 

I think we will just have to agree to disagree then. I think I've pretty well documented the changes in Effective Cost (real effect whether you think it matters or not) and the implications this has in any situation where your resource mechanic matters. Obviously if you are in situations with lots of pauses and regen opportunities like PvP these conclusions don't apply, but they do apply any time you are worried about dropping down into the next regen tier, and that happens all the time.

 

If you can provide a video showing a real world application of the mythical Perfect Healing rotation in practice, then I'd accept some of these arguments. Until then, I think it's a bad idea to make itemization decisions based on impossible scenarios.

 

I don't claim that the model is perfect, and I'm open to suggestion on improving it, but imperfect models are better than no models, and basing decisions on information, models, math, and reasoning is better than basing it on gut feelings.

 

If you don't models should guide gearing decisions, how would you suggest such decisions be made?

[Azaranth]

Faster spells are weaker spells, therefore you need more of them for the same benefit. By not putting those points into anything else, they will heal for less, and by not scaling resource costs to alacrity, you can't afford to cast more.

 

How so? Are you suggesting that reducing activation speed with alacrity reduces the coefficient on the spell gained from Bonus Healing? I've never seen anything else suggest this.

 

Edit: Upon re-reading this, I believe you're suggesting the case of stat opportunity cost. We've already covered this, and in most cases you're trading a fairly insignificant gain in other stats for a fairly substantial amount of alacrity (up to about 300 alacrity rating).

 

 

Obviously if you are in situations with lots of pauses and regen opportunities like PvP these conclusions don't apply.

 

Exactly, but the pauses and regen situations happen in every single encounter in the game. Sometimes, it's simply because the tank hit defensive cooldown, so you don't need to cast any heals for a second or two. Sometimes it's just some lucky shield procs. Either way, this happens with incredible frequency, which makes this healing rotation only a theoretical endeavor.

 

I don't claim that the model is perfect, and I'm open to suggestion on improving it, but imperfect models are better than no models, and basing decisions on information, models, math, and reasoning is better than basing it on gut feelings.

 

I agree 1000%. I hope you don't take me as arguing against statistical modelling, because I really do value it. (As previously mentioned, I'd done a similar project using a spreadsheet). Once again, thanks for putting together the tool, you've done a great job.

 

My entire point here, is just to caution folks against seeing incredibly low weight values for Alacrity, and concluding that it's a horrible stat. These models are simulating a theoretical situation, which is really a worst-case-scenario for Alacrity. The actual in-game impact of Alacrity is much larger than the numbers represent.

 

(Disclaimer: I'm by no means arguing that Alacrity is a more important stat that Power, Crit or Surge. I'm just trying to prove a point about the methodologies for establishing a weight for Alacrity)

Edited February 17, 2012 by Azaranth

[RuQu]

Unless the pauses you describe occur often and long enough for you to hit max resource, you are still suffering from the penalties of poor Alacrity implementation within this style of resource mechanic.

 

Alacrity will remain terrible, and I would say worse than the calculator suggests not better, until they make Ammo/Energy/Heat costs scale like this:

 

Actual_Ammo_Cost = Base_Ammo - regen*(delta_cast_time)

 

That will make a neutral rotation stay neutral no matter the alacrity value, while not passively increasing regen which would, with enough alacrity, let people ignore the mechanic.

[ragamer]

Alacrity is an ill designed stat, specially for Scoundrels... Because it's was born when action bars were linear.

 

IMHO, the only healer on the game that may think on gearing for alacrity is the Sage/Sorc... Because are the only ones with a big enough effective action bar pool to "eat" whatever burst is needed to save some1 and still get their energy regen intact.

 

Alacrity doesn't make your heals more efficient and that's what matters for healers with non-linear action bars (On a burst scenario... 1 or 2 "forced" casts are enough to leave the sweet energy regen zone for a scoundrel).

 

You may argue one of our extra energy sources, DS, benefits from more frequent ticks (which is absolutely true) but it also benefits from higher crit chance.

 

The bottom line is that once the diminishing returns hits hard our crit chance and surge...

 

...Alacrity is still not a good stat at all... Power is, because it really allows you to have to cast less heals overtime.

 

 

Regarding "infinite heal" rotations... Scoundrel Healers usually have a base of 80% critical effect on the main heals meaning that with some decent surge, you are easily getting double heal crits...

 

...Well, you can't heal on a perfect rotation when the difference between a crit and a non crit on UWM can be more than 3k health, sadly you have to pay attention to your crits. That's why theoretical HPS are not a good meassure for "burst healers"... Chance has to be "tamed" and not doing so has a more dramatic impact on effective healing output than any slight variation in gear allocation.

 

If your priority target is high enough, you have to let your HoTs do the rest and there is always some1 around that can benefit from an extra hand (Hell, even do some DPS if things are under control). That's why I think Alacrity is a waste... Heal for more and you will have to heal less often (I suppose I don't need to enter into the details on what Alacrity do not modify for us even).

 

...I'm not familiar with Commandos, but as they also have a non-linear action bar, I would be surprised if all the above didn't apply to them.

 

 

EDIT: Yes, I know that for some reasson, Devs decided to flood us with sets with Alacrity left and right, forcing us to grind double the gear for the same purpose... I hope that at some point the sets are corrected or Alacrity modify far more things for us, scoundrels.

Edited February 17, 2012 by ragamer

[RuQu]

It's actually worse for us. About the same for Mercs and Scoundrels/Operatives, worse for Commandos because they designed BHs first then swapped some numbers around and never really polished up the Trooper class. We drop into both medium and low regen sooner than a Merc does, but have the same cost to our abilities, and that smaller window results in us losing casts with alacrity while Mercs/Scoundrels/Operatives stay even.

[Solmors]

I am inputting my stats and was curious about some of the numbers.

 

Cunning: 1182 - Obvious.

Power: 229 - Hover mouse over Bonus Healing

Tech Power: 1112 - Hover mouse over Bonus Healing

Crit Rating: 32% - This is the main one I am curious about. Is this the Tech Crit % you have? So I have 32.02% crit chance, is that what I put in?

Surge Rating: 14 - Hover mouse over Critical Multiplier.

Alacrity Rating: 296 - Hover Mouse over Activation Speed.

 

The reason I ask is because some of the calculated values are correct (bonus healing, Crit Size, Bonus Damage), but two are off (Crit Chance and Cast Speed Reduction).

 

I put my values in red. The calculated values are 13.17 Cast Speed Reduction and 25.64 Crit chance. My actual values are 9.17 and 32 respectively. Did I do something wrong or is it off because you are still working on the algorithm? Thanks!

[schwiz]

Nice work! Thanks for making open source, but it looks like you haven't checked your sawbones rotation code into your repo yet.

[RuQu]

I am inputting my stats and was curious about some of the numbers.

 

Cunning: 1182 - Obvious.

Power: 229 - Hover mouse over Bonus Healing

Tech Power: 1112 - Hover mouse over Bonus Healing

Crit Rating: 32% - This is the main one I am curious about. Is this the Tech Crit % you have? So I have 32.02% crit chance, is that what I put in?

Surge Rating: 14 - Hover mouse over Critical Multiplier.

Alacrity Rating: 296 - Hover Mouse over Activation Speed.

 

The reason I ask is because some of the calculated values are correct (bonus healing, Crit Size, Bonus Damage), but two are off (Crit Chance and Cast Speed Reduction).

 

I put my values in red. The calculated values are 13.17 Cast Speed Reduction and 25.64 Crit chance. My actual values are 9.17 and 32 respectively. Did I do something wrong or is it off because you are still working on the algorithm? Thanks!

 

The crit rating, is the rating, not the percentage. It looks like you have about 200 Crit Rating to get 32.03%.

 

When you mouse over Tech Crit Chance, it should say something like:

Crit Rating (200): 7.8%

 

You would enter what is in parenthesis.

 

For cast speed reduction do you have points in Black Market Mods? The default is yes, and that adds 4% which is the difference you are seeing.

[RuQu]

Nice work! Thanks for making open source, but it looks like you haven't checked your sawbones rotation code into your repo yet.

 

Just getting a handle on the whole SVN thing.

 

All of that was outdated from the original terminal version.

 

I just pushed the source for the current 1.0.7 to the repository if you want to look at it.

 

I have a plan to revamp how Energy/Ammo/Heat regen is modeled as the current model has some shortcomings (ie it if you pass from <60 to >60 midcast, it doesn't catch that, it treats regen as static from the start of the cast, and Pugnacity is modeled as 1/s instead of 3 every third second).

 

So, keeping mind mind that the getEnergyRegened function will be changing and there are a lot of //comments blocking out lines used for troubleshooting and QA/QC that should probably be removed, I'm open to any advice on cleaning it up, making it more efficient and general good coding practices. This project started as me teaching myself to code, so constructive criticism and advice on all fronts is welcome.

[Solmors]

The crit rating, is the rating, not the percentage. It looks like you have about 200 Crit Rating to get 32.03%.

 

When you mouse over Tech Crit Chance, it should say something like:

Crit Rating (200): 7.8%

 

You would enter what is in parenthesis.

 

For cast speed reduction do you have points in Black Market Mods? The default is yes, and that adds 4% which is the difference you are seeing.

 

Ahh, thats exactly it. Thanks a bunch!

[rob_y]

I absolutely love this thread, even though I'm coming to it two months later.

 

I've always been down on Alacrity (my main is a Scoundrel), but I didn't like how everyone said that alacrity makes your heals less energy efficient. Ignoring opportunity cost (because that's discussed as a separate line item and we don't want to double count that - let's pretend you have a choice of 100 alacrity or 0 of any stat), that's only true if you use your extra time to spam more energy costing heals. If you instead imagine the excess time you have as just standing still. Your heals are exactly the same as before except any with a cast time (Underworld Medicine) land quicker. Now, let's pretend those gaps are filled by ticks of diagnostic scan (don't forget - you don't have to let DS run its whole course - especially now in 1.2 since they sped it up a bit). Your heals are now MORE energy efficient as a package because you're not only getting extra healing from DS, but also refunded energy.

[RuQu]

I absolutely love this thread, even though I'm coming to it two months later.

 

I've always been down on Alacrity (my main is a Scoundrel), but I didn't like how everyone said that alacrity makes your heals less energy efficient. Ignoring opportunity cost (because that's discussed as a separate line item and we don't want to double count that - let's pretend you have a choice of 100 alacrity or 0 of any stat), that's only true if you use your extra time to spam more energy costing heals. If you instead imagine the excess time you have as just standing still. Your heals are exactly the same as before except any with a cast time (Underworld Medicine) land quicker. Now, let's pretend those gaps are filled by ticks of diagnostic scan (don't forget - you don't have to let DS run its whole course - especially now in 1.2 since they sped it up a bit). Your heals are now MORE energy efficient as a package because you're not only getting extra healing from DS, but also refunded energy.

 

This thread was a bit dead and outdated, and my SWTOR interest is considerably waned from before, so I'll try and give the updates as succinctly as possible.

 

There are two ways of modeling healing, one is like on a healing dummy with "infinite damage." The other is to model incoming damage.

 

The first method is how I went at first, and is how I still calculate max burst and sustained HPS. It doesn't matter if you can do 1000HPS if the tank is only taking 800DPS, so if you want to find your limit, the tank needs to take as much DPS = max HPS or more. In this situation where the tank is just constantly taking damage and you are struggling to keep up, you really are chain casting.

 

With simulated damage, you start to get a more "natural" healing. If Underworld Medicine heals for 3.2k, you can look at a tank with 1k of damage and say "I'll just use Diagnostic Scan and wait." Using UWM would be wasteful over-healing, and the energy cost could make it so you can't afford to do two UWM/EMP combos in rapid fire when the next spike lands, depending on your starting energy level, of course.

 

When Boss DPS -> max HPS, a simulated damage model (and real gameplay) converges towards the first method. When you are forced to chain cast, Alacrity makes each cast more expensive due to decreased regen. You are correct, though, that when you have time to not chain cast and can afford to stand around between casts, this effect goes away. Each heal is a more costly burst, but you are building in recovery time to compensate it.

 

What this means in practice is that Alacrity is a Quality of Life stat that is largely only good once you are already able to heal the content. If you can't keep the tank alive, swapping out alacrity for something else is probably the answer. If you can keep the tank alive already, swapping surge out for alacrity will make your life easier.

 

I have not modeled the new DS yet, but Sc/Op healers are lucky that they have a regen tool that scales with both crit and alacrity.

Edited April 25, 2012 by RuQu