AP Scaling and other DPS changes to the Warrior going up to PTS

[Ferboten]

If you're looking for proper weights for Str/Dex, it's a pretty simple formula

With both Resonance and Titan's Strength:

Str = AP*1.1*1.05
Dex = (AP*0.25 + Crit)*1.05

Plug your chosen weights for AP/Crit in and you're good to go

[Sedacleric]

The last problem what people (or pfalcon in particular) seem to neglect is that "statweights" never quite stay the same. The effectiveness of one stats is affected by the other and vice versa.

Thank you. Crit is multiplicitive benefit. AP is an additive benefit.

+1 AP will always give you the same marginal dps gain.
+1% crit will give you more dps as you go up the gear scale.

If you want to play around with it, go to Ferboten's calculator and see what 1% crit does. Now double your AP and see what it does.

[Havors]

The top and bottom of it is... Obviously there is an exact equation to work this out... hence people getting 2:1 5:3 etc etc.. BUT at the end of the day CRIT is still important but just not as important as it was because of the AP changes.

Just stacking AP/Str and not paying attention to CRIT at all will in the end lose you DPS if you **** the ratio up. i.e. someone hitting for 5K swings and criting once ever 20 secs for say 6K will do less DPS than someone hitting 4K but criting every 10 secs for 5K. Same as its always been. Its all about finding the exact sweet spot/ratio.

Obviously im just using totally random numbers but my point stands.

[Steelwind]

If you're looking for proper weights for Str/Dex, it's a pretty simple formula

With both Resonance and Titan's Strength:

Str = AP*1.1*1.05
Dex = (AP*0.25 + Crit)*1.05

Plug your chosen weights for AP/Crit in and you're good to go

Ferboten, did you take into account Proper Timing when calculating for Champ specs?

[Ferboten]

Ferboten, did you take into account Proper Timing when calculating for Champ specs?

Yes, it assumes proper timing is used on the highest damage finisher.

[PFalcon]

The top and bottom of it is... Obviously there is an exact equation to work this out... hence people getting 2:1 5:3 etc etc.. BUT at the end of the day CRIT is still important but just not as important as it was because of the AP changes.

Just stacking AP/Str and not paying attention to CRIT at all will in the end lose you DPS if you **** the ratio up. i.e. someone hitting for 5K swings and criting once ever 20 secs for say 6K will do less DPS than someone hitting 4K but criting every 10 secs for 5K. Same as its always been. Its all about finding the exact sweet spot/ratio.

Obviously im just using totally random numbers but my point stands.

Right, the random numbers can't exist within T3 with any reasonably geared warrior. You actually can't drop so much crit you start failing. On PTS I tried. Going pure STR/AP results in the same DPS as trying for crit cap, and trying to blend somewhere in the middle.

There is a ton of pseudochoice in 1.8 gear. Almost nothing you choose to do will actuall affect your performance. All you have to do is upgrade, and your DPS will upgrade. What you upgrade with almost doesn't even matter any more.

Sad in a way that there's no thought to what to equip any more, but it's better this way so you don't have to check BiS tables every time something drops.

[Rizaz]

If you're looking for proper weights for Str/Dex, it's a pretty simple formula

With both Resonance and Titan's Strength:

Str = AP*1.1*1.05
Dex = (AP*0.25 + Crit)*1.05

Plug your chosen weights for AP/Crit in and you're good to go

For STR ratios, Titan's Strength and Resonance are additive to a 1.15 modifier. Not multiplicative. Also, PA str, dex, and endurance is not modified by Titan's Strength or Resonance.

[wraithh]

It feels like what would be very useful is a "dps per item-value point" graph covering a range of AP and crit ratings from like 700-1300 in crit, and like 700-1500 in AP or so.

Basically that plot would tell you for a given AP/crit level and weapon damage (knowing what amount of that was PAs/etc) the best dps increase for a fixed "cost" of points on an item.

This seems doable if we know or can do the following:

1) Trion's valuation of stats relative to each other when they create items (which I think is 0.75 str = 0.75 dex = 1 AP = 2 crit)
2) Damage formulas for all abilities in a rotation (which I know Ferboten has and has published)
3) How Titan's Strength and Resonance affect the values of dex/str in the above (this is probably the most complicated bit in trying to make this a simple 2-D plot rather than 4-D, I think...)

But if this was doable...you'd be able to compute your dps for any combination of AP/crit ratings within the "reasonable" range, then divide it by the cost of items it'd take to reach those levels (obviously base stats + PAs are removed), at which point you could read off the contours of "equal item cost" and find the maximum dps at that item cost. Jumping from the max point on each contour of equal cost to the next one up should tell you the optimum stat distribution (at least in AP/crit in the end, if not the precise str/dex/ap/crit breakdown) for the quality ("cost") of items you have. Obviously you can't just arbitrarily pick your stats because Trion only gives us a limited set to pick from, but it should indicate the best path to try to follow at least regardless of where you are in gear.

Not sure if I want to attempt to do this or not, but it seems logically the proper approach because it takes into account the opportunity cost of giving up one stat for the other due to itemization.

[PFalcon]

It feels like what would be very useful is a "dps per item-value point" graph covering a range of AP and crit ratings from like 700-1300 in crit, and like 700-1500 in AP or so.

Basically that plot would tell you for a given AP/crit level and weapon damage (knowing what amount of that was PAs/etc) the best dps increase for a fixed "cost" of points on an item.

This seems doable if we know or can do the following:

1) Trion's valuation of stats relative to each other when they create items (which I think is 0.75 str = 0.75 dex = 1 AP = 2 crit)
2) Damage formulas for all abilities in a rotation (which I know Ferboten has and has published)
3) How Titan's Strength and Resonance affect the values of dex/str in the above (this is probably the most complicated bit in trying to make this a simple 2-D plot rather than 4-D, I think...)

But if this was doable...you'd be able to compute your dps for any combination of AP/crit ratings within the "reasonable" range, then divide it by the cost of items it'd take to reach those levels (obviously base stats + PAs are removed), at which point you could read off the contours of "equal item cost" and find the maximum dps at that item cost. Jumping from the max point on each contour of equal cost to the next one up should tell you the optimum stat distribution (at least in AP/crit in the end, if not the precise str/dex/ap/crit breakdown) for the quality ("cost") of items you have. Obviously you can't just arbitrarily pick your stats because Trion only gives us a limited set to pick from, but it should indicate the best path to try to follow at least regardless of where you are in gear.

Not sure if I want to attempt to do this or not, but it seems logically the proper approach because it takes into account the opportunity cost of giving up one stat for the other due to itemization.

You won't need to do that. People mistakenly believe that to find out what's best, they need to understand how the results happen. You don't need to know why and how. You just need to know what the results are. The developers design retrograde. They start with the results they want, then build equations to create those results. You, therefore, don't need to understand any of the equations. You can simply test results.

The assumption that AP and crit result in constant proportions of gain along T3 is appropriate, because that appears to be a result that the devs already started with. Again, to build a battle engine you start at the end result, and work backwards. To come up with a battle engine with shifting gains, you have to either be a bad developer, or have meant to design it that way from the start.

The only AP-focussed warrior who may fall off the normal ID linearity curve is a barely T2 warrior with full str and AP, 4 low vendor gear pieces for synergy, and a bunch of poorly itemized HK and ROTP gear. IE someone blatantly breaking the 2:1 rule.

Any warrior following the 2:1 rule as far as what to equip, across any conceivable T3 progression level, will parse optimally.

[wraithh]

You won't need to do that. People mistakenly believe that to find out what's best, they need to understand how the results happen. You don't need to know why and how. You just need to know what the results are. The developers design retrograde. They start with the results they want, then build equations to create those results. You, therefore, don't need to understand any of the equations. You can simply test results.

The assumption that AP and crit result in constant proportions of gain along T3 is appropriate, because that appears to be a result that the devs already started with. Again, to build a battle engine you start at the end result, and work backwards. To come up with a battle engine with shifting gains, you have to either be a bad developer, or have meant to design it that way from the start.

The only AP-focussed warrior who may fall off the normal ID linearity curve is a barely T2 warrior with full str and AP, 4 low vendor gear pieces for synergy, and a bunch of poorly itemized HK and ROTP gear. IE someone blatantly breaking the 2:1 rule.

Any warrior following the 2:1 rule as far as what to equip, across any conceivable T3 progression level, will parse optimally.

Good point. There's a lot of wasted calculation in that based on the itemization. I've been following (and tested myself a bit) the AP/crit tradeoffs and noticed that it didn't seem to matter as long as you stayed below cap, but it's always nice to be precise. I suspect though that within typical RNG of crit chance it really won't matter though as you say.

[PFalcon]

Good point. There's a lot of wasted calculation in that based on the itemization. I've been following (and tested myself a bit) the AP/crit tradeoffs and noticed that it didn't seem to matter as long as you stayed below cap, but it's always nice to be precise. I suspect though that within typical RNG of crit chance it really won't matter though as you say.

I'm finding the same thing, to the point where I like Ferboten's conclusions but think he's being a little hard on dex runes. I can't make rune choice result in a DPS loss or gain. I can't make crit cap versus 2:1 (which is essentially focussing only on STR/AP gear and ignoring crit cap and dex-heavy items) parse as a gain or loss.

I do not forsee that you will be trying to roll on pendant of the eel types of equipment in this tier because it is a technical gain.

Again, Trion probably started with the basic principal that warriors having to retool themselves half-way through the content, leather vendor gear, and loot stuff rogues want, was a mistake in T2. And this is the successful product of starting with those principals.

The initial urge was to get that crit after seeing how low crit goes on ID gear. But it is not there so you have to change focus mid tier. It is so that you know how powerful AP really is and that Trion has finally fixed this calling.

So this is great. Don't go above crit cap. Don't try to unduly get there. Don't give up 10 crit for 2 AP, but don't give up 10 AP for 15 crit either.

[Ferboten]

The reason I say Str > Dex and Str runes > Dex runes is because of Titan's Strength giving you that 10% extra strength. If it wasn't for that, I would probably get Dex runes.

[Gargoyles]

If you're looking for proper weights for Str/Dex, it's a pretty simple formula

With both Resonance and Titan's Strength:

Str = AP*1.1*1.05
Dex = (AP*0.25 + Crit)*1.05

Plug your chosen weights for AP/Crit in and you're good to go

sounds simple to me, no actual correlation to each other because crit does not weigh in at all in between the 2 statements, so no weight.

[Thargrom]

If you're looking for proper weights for Str/Dex, it's a pretty simple formula

With both Resonance and Titan's Strength:

Str = AP*1.1*1.05
Dex = (AP*0.25 + Crit)*1.05

Plug your chosen weights for AP/Crit in and you're good to go

Don't you mean:

Str = AP*0.75*1.15
Dex = (AP*0.25 + Crit)*1.05

Isn't STR only 0.75 AP? ...and taking Titan's Str and Resonance as additive bonuses into account

[Ferboten]

Don't you mean:

Str = AP*0.75*1.15
Dex = (AP*0.25 + Crit)*1.05

Isn't STR only 0.75 AP? ...and taking Titan's Str and Resonance as additive bonuses into account

You're right. And I can't remember if I tested Titan's Strength and Resonance being additive or multiplicative, I'll have to run a quick test again just to make sure. And for PA's, Titan's Strength indeed doesn't affect Str PA's, but I was told that Resonance did.