Softcaps discussion

[Skeleton Jack]

(02-03-2012 10:23 AM)Freehugs Wrote:  "Normal" is a bad description too, better to call them Weapon attacks.
I'm happy to change things in my formula post if it makes them clearer. I could probably split the damage formula's in Force/Tech, and Weapon attacks, I'm eternally looking for ways to make that section less confusing.

(02-03-2012 10:08 AM)Kor Wrote:  I'm sorry, but you're just flat wrong here. The community definition of "melee" and "ranged" may simply refer to the range of the attack, but the game actually defines attacks as one of the 4 types I listed. If you open your ability window, all abilities are listed as one 6 types: Melee, Ranged, Force, Tech, Active, Passive.

At the end of the day I don't care what kind of semantics anyone wants to argue about the names or types or the data mined names. People, real live people, who just play the game and don't data mine or create spread sheets come here to read this stuff. And they don't want to sort through technicalities of word choices in game files.

Normal attacks (or pick a better suited name) do WHITE damage. Greetings Defense!!

Tech/Force attacks do YELLOW damage. Greetings Resistance!!

That's it. End of story. That's all the readers care about. White damage attacks go against Defense and Yellow damage attacks go against Resistance. Ranged and Melee designators don't mean anything once you break it down. You're just adding words to confuse the matter. So that's three types. Which can even be simplified down to 2 types, yellow and white.

Explaining the data mined names in an OP or a guide for reference is ideal. But in the everyday discussion of the game, it's just an added complication for people who are newbies trying to come here to better understand their class.

[Kaedis]

Quote:Apologies if this is too much of a digression, but if you are correct, then does it not follow that most, if not all, the Accuracy on many classes' gear is virtually worthless (or at least significantly less valuable than alternative stats)?

You are in fact correct, but then, we already knew that Bioware's itemization is completely borked.

Actually, one really good piece of evidence that non-weapon attacks can't miss is the fact that the tier gear for the only AC that doesn't use weapon attacks at all (Sorcerers) has zero accuracy on it.

Unfortunately, Bioware seems to be in bed with Accuracy and Alacrity. Every single enhancement on every single piece of tier gear, PvP and PvE both, has either Accuracy or Alacrity on it (excluding a couple tanking ones with Defense). This, combined with the inability of crafters to generate higher than Mark 22 enhancements, means that you either have to drop a considerable amount of itemization points (by dropping back to Mark 22 enhancements) or be stuck with at least 255 Accuracy or 255 Alacrity on your gear (and that's excluding any from non-set slots). Yay....

I think Operatives are hit the hardest by this, as their gear is swamped in Accuracy, and the overwhelming majority of their damage (at least as Concealment) is from Tech attacks. Add to that the fact that they get no benefit from Alacrity either, and they are just sorta boned.

Quote:1. Do we have any evidence that bosses do not have any Resistance Chance?

Well, you can't really "prove" negatives with inductive reasoning, you can only generate evidence supporting them. That said, thus far we've not had any evidence showing they do have Resistance, despite a number of people watching for resists.

Quote:2. Do we have independent reasons for thinking that it is likely that bosses do or do not have any Resistance Chance? (For example, does it fall under a standard practice in designing MMO mechanics?)
3. If we do have such independent reasons, what are they, how much weight should we place on them, and why?

The major justification I can give here is that the PvE DPS tier set for the only class that uses no weapon attacks (Sorcerers) has no accuracy on it at all.

Now, it's possible that Bioware just screwed up and meant to put Accuracy on them. However, given that the other dps classes (and even the tanks!) are swamped in Accuracy, including classes like the Operative that deals almost entirely Tech damage, combined with the above lack of evidence for resists, I'd say it's justifiably conclusive given the current volume of evidence.

---

Quote:At the end of the day I don't care what kind of semantics anyone wants to argue about the names or types or the data mined names. People, real live people, who just play the game and don't data mine or create spread sheets come here to read this stuff. And they don't want to sort through technicalities of word choices in game files.

And that's why we define them. However, this is a theorycrafting forum, not a guidebook. Yes, we have guides here, but the primary point of the discussion is to theorycraft the game, and to that end, particularly in a thread dedicated to said theorycrafting (which is not a guide), we will continue to use the terms the game defines for these effects so as to be explicitly clear to those theorycrafting the effects. If you disagree with that, take it up with Masterkiller or Altratan.

[Analysis]

I appreciate your reply, Kor. I'd just like to point out that the last quotation in your post was not mine. The way your post is formatted might lead some to think so. Since I fundamentally disagree with Skeleton Jack's argument, I feel compelled to clarify this.

[Kaedis]

Clarified by moving horizontal rule. My apologies. I tend to use horizontal rules to mark when I update posts, rather than marking who I'm responding to. However, you make a rather good point, it's nice for split posts by who I'm responding to.

[CaseyTheRetard]

(02-03-2012 11:41 AM)Skeleton Jack Wrote:  At the end of the day I don't care what kind of semantics anyone wants to argue about the names or types or the data mined names. People, real live people, who just play the game and don't data mine or create spread sheets come here to read this stuff. And they don't want to sort through technicalities of word choices in game files.

This argument expresses the most ridiculous sense of entitlement I have ever seen on a theorycrafting site. If you are not happy with the semantics and language we have chosen to use in discussing the game, feel free to ask for a refund. I for one will be happy to return to you every cent you've paid me for the work that I personally have not performed to your satisfaction.

[Ezmode]

We would appreciate it if everyone could keep the discussion civil before it really gets out of hand. I will give everyone a stern talking to if needed, and no one wants that to happen.

Back on topic, has anyone else attempted to find the right Accuracy requirement for bosses yet? If so, please share your data!

[Eridian]

Hi everyone, I am not a math guru, so I won't talk about the numbers, but I would like to suggest if it is reasonable to do so, is that anyone posting up data take on a journalism approach to theory crafting, that is to say, include references to your sources that enabled you to come to your conclusions.

Putting up numbers is all fine and dandy but can't be taken seriously when no sources are available for referencing. Don't assume that the readers and even fellow theorycrafters know where you got your information from...

This is a great start to a great forum, let's feed it with facts and a positive vibe and avoid the bickering please. Everyone wants to be an expert, but if your information/data can't be validated by using source information as reference, it's nothing but conjecture. So please, theorycraft away, but please include reference and source of information/data.

Thanks guys and keep this up I am really interested to see more of your numbers on this subject.

[Kaedis]

Quote:Hi everyone, I am not a math guru, so I won't talk about the numbers, but I would like to suggest if it is reasonable to do so, is that anyone posting up data take on a journalism approach to theory crafting, that is to say, include references to your sources that enabled you to come to your conclusions.

Putting up numbers is all fine and dandy but can't be taken seriously when no sources are available for referencing. Don't assume that the readers and even fellow theorycrafters know where you got your information from...

Most of the information we get is either explicitly cited from mathematical formulae (which don't need to be re-derived every time they are used), or are from the game files, or both. The game files cannot easily be cited, though if something is in a location that is no commonly known the poster usually references the table or node it's from. Ability data is all from the ability nodes themselves, and easily verified.

Beyond that, per our "do your research" rule, anything previously discussed and solidified on these forums is open for inclusion as basis of discussion without need for citation (requiring it would quickly become prohibitively cumbersome even in the best cast scenario). These aren't news articles, the idea is that those that frequent the site and contribute to the knowledge and calculations here have a solid working knowledge of the game's mechanics already (which are discussed elsewhere on this site) and of the theorycrafting already done and currently being done on the site. I agree that certain things could use better citation at time, but a fair amount of the mechanics thus far are still based on peer-critiqued observation by some of our veteran members, with no citable "report" to link to. Some others tend to boil down to simply recognition of the poster as an expert in that particular area, which the information coming from a developed depth of knowledge rather than any single post or link.

[Kaedis]

It has occurred to me lately that even Stat Weights are a poor thing to focus calculations on, as they are wildly dynamic and dependent on gear. The mathematical construct that is most useful in defining a stat's relationship to the other stats is a set of constraint equation that define an isocline. For those not well versed in calculus, an Isocline is a curve the describes the points in which a parent function has the same slope. In other words, if you had a family of related curves (very common in differential equations), an isocline would be a curve that intersected each of the curves in the parent family at the points that have identical slope on each curve.

If that still confuses you, well, I'm not sure I can make it more clear. In any case, "isocline" is a bit of a misnomer for what I'm getting at anyway. Basically, finding constraint equations that define the points at which all of the relevant stats are the same stat weight, regardless of what that weight may be. Now, according to optimization theory, this point also represents the maximization of the overall equation (being HPS or DPS).

For example, given the following assumptions (somewhat oversimplified, but you get the idea), I've developed the below isocline equations for Corruption Sorcerers:

• The only healing spells used are Resurgence, Innervate, Static Barrier, and Dark Infusion.
• The character possesses the PvE 2-set bonus.
• Innervate is cast every 7.5 seconds.
• Resurgence is cast every 7.5 seconds to line up with Innervate.
• Static Barrier is cast every 20 seconds.
• Dark Infusion is used as filler during the remaining time.

Crit = 155 + 13% (Power + Force Power - 1200)
Surge = 260 + 6.5% (Power + Force Power - 1200)
Alacrity = 155 + 21% (Power + Force Power - 1200)

If all 4 stats (Power, Crit, Surge, Alacrity) are at the values provided by these equations, their stat weights will all be within 1% of the average of their stat weights, given the above assumptions. This gives a solid linear system that describes the stats, and an easy way to calculate relations between them. This, I think, should be our ultimate goal for stat metrics for each class.

[LagunaD]

Worth thinking about, but I'm not sure how one would go about computing such curves in a general way.

Also, you must have assumed some value for the primary stat, which doesn't appear in those equations but surely affects the weights.