New efforts!

[Tubbius]

So many numbers.  They make this English teacher's head hurt.  :(

[Arcana]

Fair enough.  I was thinking that with a single SO enhancement against even-level enemies, a slight decrease in expected damage from accuracy over damage slotting might be worth it by virtue of the significantly reduced probability of a substantial number of misses over a few attacks, which could potentially place the character at real risk of defeat (particularly for squishies).

This is also true, but in I3 the forums just were not ready for me to bust out a markov analysis.  I was still trying to get people to buy into average damage mitigation algebra.  Many people were making the argument that the math might be technically correct but not properly represent the game, even though the game is run by a computer that essentially uses those same calculations to execute the game.

[brothermutant]

Ok, I like your example and totally understand what your goal is here, FASTEST burn time really with the BEST end cost. If this is correct, I get it now.

Thing I always fall back on is a couple of things: 1) my way allows me to cast the power MORE times than you (assuming a full end bar, and NO end recovery), I get 13 attacks to your 10. And 2) with guaranteed missing, you could only hit the same target 9 times in a row, where I COULD hit a total of 12 (yes I am aware that I could easily miss more than once, but so could you). While not as fast as your way, it allows me the freedom to change things up if needed. Also, 3) what would happen if we throw in resistances from the mobs? Would it not affect the outcome? Again, I am just curious. I think my issue is my playstyle vs what you guys seem to know is better. That's probably what is hanging me up.

[Arcana]

Level-shifting could be left in at the tier four still though yes? Or was it tier 3 and 4? Been too long since I played.

Level shift appeared in tier 3 (and of course also in tier 4)

You lost me with this point. How does buffing all my powers with just earning one alpha slot (assuming the buff is used in the power, e.g. auto powers don't need recharge) encourage more grinding than steadily earning a slot or two per incarnate slot (what I am saying would be Alpha tier 3 or 4 would give a single slot, Destiny tier 3 or 4 would give a single slot, etc)? Unless you are referring to those people who went for more than one alpha slot so they can decide which buffs to use? If that is the point, that would make sense I guess.

Basically.  Grinding to earn an new enhancement slot is functionally identical to grinding to unlock a new tier in the Alpha tree, but the Alpha tree is 2-dimensional: you can earn tiers, and *also* different enhancements.  What's more, even if you only want one, and you choose the one, and you earn the one, the devs could always add more Alpha options that would encourage you to earn a new Alpha - and that did happen with Alpha expansion.  Regular enhancement slots can't really do that.  And at the point where you're making special incarnate enhancement unlocks for special incarnate slots unlocked with an incarnate tree, you're really just replicating the Alpha design and shoe-horning its benefits into the enhancement system.

And then you'd *still* have to make the incarnate tree system anyway for the other incarnate powers.  They can't all be just more slots.

[MM3squints]

At this point someone would say, "this is all theory crafting lets test it out" and then at this point I would make myself feel sad for saying that ~O.o

[Arcana]

Ok, I like your example and totally understand what your goal is here, FASTEST burn time really with the BEST end cost. If this is correct, I get it now.

Thing I always fall back on is a couple of things: 1) my way allows me to cast the power MORE times than you (assuming a full end bar, and NO end recovery), I get 13 attacks to your 10. And 2) with guaranteed missing, you could only hit the same target 9 times in a row, where I COULD hit a total of 12 (yes I am aware that I could easily miss more than once, but so could you). While not as fast as your way, it allows me the freedom to change things up if needed. Also, 3) what would happen if we throw in resistances from the mobs? Would it not affect the outcome? Again, I am just curious. I think my issue is my playstyle vs what you guys seem to know is better. That's probably what is hanging me up.

1.  Swinging more often doesn't improve your overall ability to hit.  And there's no such thing as guaranteed missing.  The streakbreaker broke streaks of misses, not streaks of hits.  I can hit 12 times out of 13 just as often as you can with higher recharge.

2.  Higher recharge does mean you can attack more often than I can with that one power.  But the overall benefit is not as good as that seems because we don't have only one attack.  In reality, we tend to have four or five attacks, and when some are recharging others are ready.  If you slot recharge each individual attack will be able to fire more often, and if I'm idle a lot and you're not that's a significant benefit.  But that benefit drops if that recharge means you have two attacks ready when I have only one.  You might be able to use a better attack than me that has better DPA, but better attack vs less good attack is not as much benefit as attacking vs not attacking.  This gets complicated to try to summarize, and I even tried to create a metric to quantify it, but it didn't catch on.

3.  Playstyle preferences have nothing to do with quantitative efficiency.  Playstyle is how you want to play.  My +180% recharge build for my Energy/Energy Blaster is not as quantitatively strong as someone that, say, soft-caps their blaster.  But when I'm sending about a gajillion mag of knockback and energy projectiles down range almost faster than my fingers can move, I don't really care.  I'm what you call a preference maxer.  I have preferences, and then I try to generate the maximum quantitative build that delivers that.  I don't try to make the best build possible, I try to make the best possible realization of the build I want to play.  Everyone else should do likewise, to the extent they want to work on it.

Even back in the day, there were people building so-called "marathon builds."  These were builds that had a lot of endred, because they wanted a build that would allow them to play at maximum activity level without running out of endurance.  Were these builds generating the best DPS, or even the best DPE?  Nope.  But they did have great EPS, in fact by explicitly *not* increasing DPS, they lowered EPS to the point where they could remain continuously active without worrying about running out of gas.  Their own limited recharge prevented them from using attacks faster than they could power them.  That's perfectly viable, and perfectly reasonable, as long as you are doing that *deliberately*.

4.  Resistance has no effect on anything, except for the issue of overkill you mention.  But that's a complex thing to analyze without resorting to some very sophisticated analysis tools (stochastic or monte carlo markov-type stuff).

[Arcana]

So many numbers.  They make this English teacher's head hurt.  :(

Think about all the math majors that have to read all the words in this thread.

[Pyromantic]

Think about all the math majors that have to read all the words in this thread.

I'm a math major, but I did a BA and most of my electives were English.  So I suppose I'll tolerate all these words that allow us to talk about math.   :)

[brothermutant]

I always thought you could only hit 9 times in a row too. Really, I could hit every time if luck was on my side? That is awesome. Ok, so if that works with one dmg vs one endred or one rc, what happens when you have two slots to choose from. Should I still go dmg for both? a mix of dmg/endred or dmg/rc or endred/rc? At that point, my calculations get nuts.

And while I am sure "playstyle" didn't really affect anything, you have to admit that if fighting high level cons where missing was a pretty decent possibility (obviously not the 95% chance to hit cap), having the ability to cast a power more times than your high damage slotting COULD allow me to surpass you in total damage out, yes? I mean, it sounds like you are suggesting ideal conditions where we both hit our target an equal number of times. We cannot guarantee that (as far as I know) so how could you account for that? I am not trying to nitpick here, just trying to understand what I would shoot for when the game comes back. Has anyone done the math on say shooting +2 to +4 con bad guys in a statistical way that shows a general "sweet spot" for slotting using nothing but SOs (or IOs I suppose is ok too)?

And I won't point out that the more I attack with my -defense rad power, the better my chances to hit the next time. ;)

[Azrael]

I found if I put two slots in a given power I could slot for acc, dam and end.  It made a slight difference with TOs and more noticeable with DOs.  Some attacks I would double that ratio to six slot the key attacks in my attack chain.  Ergo x2 acc, x2 dam, x2 end red.

Which translated into an SO of each before SOs became available as standard.

It worked for me.  I would hit most things, did a dollop of damage and it shaved an SO of end off each attack.  In early game terms that was x4 TOs worth of acc, x4 for dam and x4 of end red.  12 slots!  Double the early games 6! :). Psychology.  As opposed to going dam only and burning end pre stamina inherent.   It took a while to come to this balanced view.

Very noticeable and balanced in that light.  And it stemmed the end problem pre stamina.

As soon as I hit SOs?

Gloves were off.  I could double my acc, dam and end red for the same amount of slots by replacing the DOs with SOs. 

In practice I went one SO acc, x3 dam, 1 SO end.  Double in end heavy attacks or one end and one recharge.

Azrael.

[HEATSTROKE]

I think to a degree BrotherMutant is underestimating the incarnate abilities compared to adding slots.

For arguments sake lets say you had the ability to add six slots across your build. Maybe you could max out a power or two. But you would still have to deal with ED and the decrease in benefit due to its diminishing return effect.

Taking my NRG/NRG Blaster who had Musculature Core Paragon as his Alpha. Now Six slots would be nice but in all honesty most of my powers were already six slotted. So I could add some slots to a few other things..

But the Alpha increases my damage 45% across EVERY power and 2/3 of its effects are ignored by ED.

That means my Power goes from a 98.5% buff in damage to a 130.85% Buff in Damage. Now this damage buff is across EVERY attack power.

Power Bolt, Power Blast, Power Burst, Snipe, Energy Punch, Bonesmasher, Total Focus, Explosive Blast. Nova even lowly Power Push all get a HUGE boost in damage.

That to me is worth more than a few slots..

[Codewalker]

The trick would be to somehow "highlight" the extra slot so they know which will NOT be part of their exemplar build.

That would require a substantial redesign of how enhancement slots work. The game did not keep any record of what level you got each slot at. Instead, it only remembered "Power Bolt has 4 extra slots added". When it came time to level it just added up all the slots you've placed so far and compared to the number you're supposed to have at that level in order to determine how many new ones you got to place.

[Arcana]

I always thought you could only hit 9 times in a row too. Really, I could hit every time if luck was on my side? That is awesome. Ok, so if that works with one dmg vs one endred or one rc, what happens when you have two slots to choose from. Should I still go dmg for both? a mix of dmg/endred or dmg/rc or endred/rc? At that point, my calculations get nuts.

It also starts to get hazy what you're trying to get.  The math can only tell you what you get, but not what you should be getting.  Assume the first slot is acc and the second is dmg.  What should the third slot be?  Well, the only way to compare is to see what happens, but to summarize a bit:

Stacking damage means proportionately speaking the second one has less benefit than the first.  Not in absolute terms, but in relative terms.  Adding 33 to 100 is proportionately more than adding 33 to 133, and we have to look at it proportionately rather than linearly because that's the only way to compare apples to apples: when we stack recharge or endred on top those will have proportionately similar results in both cases: speeding things up by 33% has the same effect if the attack does 100 or 133 damage.

Or if you don't buy the proportionality argument (many forumites didn't for various reasons of varying bizarreness) you can just do the explicit math.  An attack already enhanced with 33% damage deals (in our example) 133 damage every 10 seconds costing 10 end.  Your three options now are:

Slot damage: 166 damage every 10 seconds at 10 end; 16.6 dps, 16.6 dpe, 1 eps
Slot endred: 133 damage every 10 seconds at 7.5 end; 13.3 dps, 17.7 dpe, 0.75 eps
Slot recharge: 133 damage every 8 seconds at 10 end; 16.6 dps, 13.3 dpe, 1.25 eps (assumes 2s cast time)

The clear loser is recharge.  The numbers suggest that damage looks a little better than endred: you get a lot better dps for only slightly less dpe, assuming you are mostly concerned with constraint efficiency.  Interestingly, people used to say that recharge had diminishing returns because of a math error: they claimed that since the first one reduces recharge by X and the second by less than X, that was a diminished return.  That's actually false, but recharge actually has diminishing returns for a completely different reason: constant cast time.  Notice how fast it hits: it devalues recharge on the *first* slot of recharge, not even the second.

So if you buy that analysis, you're down to an attack having acc, dmg, dmg.  Now it gets a bit hazier in terms of what you consider important, because the advantages of damage slotting will drop significantly with the third one.  It looks like this:


Slot damage: 199 damage every 10 seconds at 10 end; 19.9 dps, 19.9 dpe, 1 eps
Slot endred: 166 damage every 10 seconds at 7.5 end; 16.6 dps, 22.1 dpe, 0.75 eps
Slot recharge: 166 damage every 8 seconds at 10 end; 20.8 dps, 16.6 dpe, 1.25 eps (assumes 2s cast time)

Compare damage vs endred slotting.  The endred case deals 16.6/19.9 = 83.4% of the dps, but at 75% of the cost.  That can sound like a good deal.  On the other hand, slotting recharge is the undisputed dps leader at 20.8 dps (although its only 5% more), but at huge endurance costs that might not be sustainable.  Damage slotting is now somewhat of a middle ground case: not as much dps as recharge, but not as much cost as recharge, and vice versa for endred.

My overall judgment was that on a quantitative basis the slotting of endred seems to have the biggest payoff: 83% of the dps for 75% of the cost is a quantitative good deal.  But its not a clear-cut overwhelmingly obvious choice.  It continues to be somewhat judgment call oriented with additional slots.

And while I am sure "playstyle" didn't really affect anything, you have to admit that if fighting high level cons where missing was a pretty decent possibility (obviously not the 95% chance to hit cap), having the ability to cast a power more times than your high damage slotting COULD allow me to surpass you in total damage out, yes? I mean, it sounds like you are suggesting ideal conditions where we both hit our target an equal number of times. We cannot guarantee that (as far as I know) so how could you account for that?

Mathematically.  Base chance tohit is 75%.  That means you'd expect 75 hits out of 100 on average, but it also means you can calculate the average damage of such a power: it will be 75% of its nominal damage.  If we have an attack that deals 100 damage and we slot it with damage and accuracy, we can calculate what the statistical average case will be.  For damage slotting, you'd expect the average damage per attack to be 0.75 * 100 * 1.33 = 99.75 points of damage.  For acc slotting, because 1.33 * 0.75 = 0.9975 and the tohit ceiling is 95%, its expected average damage will be 0.95 * 100 = 95.  So damage slotting actually generates slightly higher results, because of the tohit ceiling.  When you're attacking things where your base chance to hit is lower (higher level, high defense, etc) and that ceiling is not applicable, the results will be less divergent.  At the 50% mark (and any value more than a couple points lower than 75%) you'd have 0.5 * (100 * 1.33) = 66.5 for damage slotting and (0.5 * 1.33) * 100 = 66.5 for accuracy slotting which is obviously identical just from the math expressions.

So below base 75% chance to hit, acc and damage slotting generates identical results for expected average damage.  But where there's a tie for damage, there are tie-breakers.  More accuracy means more hits for less damage, more damage slotting means less hits for more damage, but they come out to the same damage.  But the total hit rate also increases the amount of secondary effects that land, and that isn't also boosted by boosting damage.  So away from the tohit ceiling, acc boosts damage *and* secondary effects (by hitting more often) while damage slotting boosts damage by the same amount but does nothing for secondary effects.

Now you have to decide what is more important to you: slotting that always buffs damage but never buffs secondary effects, or slotting that sometimes buffs damage less but also buffs secondary effects.  That's difficult to quantify in most cases and is somewhat of a judgment call, although most players lean towards accuracy in this case for various reasons as do I.

[Arcana]

That would require a substantial redesign of how enhancement slots work. The game did not keep any record of what level you got each slot at. Instead, it only remembered "Power Bolt has 4 extra slots added". When it came time to level it just added up all the slots you've placed so far and compared to the number you're supposed to have at that level in order to determine how many new ones you got to place.

Which is partially why adding more slots became a nightmare for the devs.  There were dependencies like this that relied on the number of slots being a very specific amount at specific levels, and changing that number broke many of them in different places, even the client itself.

There was a bug TopDoc found in the way the client validated things that in my opinion ranked up there with the infinite peacebringer slotting problem, and I was shocked that a) he didn't get Bug Hunter for it and b) that it was even possible at all.  Even by my standards of knowing how broken certain things were, some of the stuff TopDoc broke were astounding.

[Arcana]

I think to a degree BrotherMutant is underestimating the incarnate abilities compared to adding slots.

For arguments sake lets say you had the ability to add six slots across your build. Maybe you could max out a power or two. But you would still have to deal with ED and the decrease in benefit due to its diminishing return effect.

Taking my NRG/NRG Blaster who had Musculature Core Paragon as his Alpha. Now Six slots would be nice but in all honesty most of my powers were already six slotted. So I could add some slots to a few other things..

But the Alpha increases my damage 45% across EVERY power and 2/3 of its effects are ignored by ED.

That means my Power goes from a 98.5% buff in damage to a 130.85% Buff in Damage. Now this damage buff is across EVERY attack power.

Power Bolt, Power Blast, Power Burst, Snipe, Energy Punch, Bonesmasher, Total Focus, Explosive Blast. Nova even lowly Power Push all get a HUGE boost in damage.

That to me is worth more than a few slots..

I think he would say that you could add those features to the extra enhancement slots, but for a variety of complex reasons including but not limited to the one Codewalker mentions that would be problematic.

But I think even above all technical issues there's the simple fact that the devs would never want an entire Incarnate power to be just more of what we already have.  Getting more slots is pedestrian, no matter how useful.  The point of the incarnate system was to add something different to the game.  By its very definition it was not going to incorporate just things we already had.  The deciding vote on why the incarnate system didn't give us more slots is simply that we had a way to get slots, called the normal leveling system.  If you wanted more slots, the correct place to have that fight was to argue the leveling system should grant more slots (which the devs in fact thought might be a good idea).  It was not in the incarnate system, where if it wasn't new in some significant way, it wasn't worth doing at all.

In a certain sense, asking the devs for the incarnate system to grant more slots is like asking why Praetoria doesn't have its own Skyway City.  Its not about whether it would be useful, its about the fact that Praetoria was going to have nothing but new, unique zones.  Asking for a near replica of an existing zone misses the point of Praetoria existing at all.

[pinballdave]

One thing I wished the devs would have implemented is some sort of character bound IO recipe system.  I think that once a character acquired a specific IO recipe then they should be able to use that recipe multiple times instead of having to find/purchase more recipes of the same exact IO.  I also think that when considering that we could have multiple builds, our enhancements should have been able to span across each of our builds instead of requiring us to acquire a whole new builds worth of enhancements for each build.  Both of these ideas would likely reduce grinding time and affect the auction house.

This got lost in the shuffle. The game did employ the memorization of generic IO recipes for crafters. You would buy and craft x recipes for say level 25 and y recipes for level 30. Upon completion you would earn a badge and not need to buy a level 25 or 30 recipe to craft another generic IO of that level. You would still need the salvage.

Set IO recipes could not be learned. The levels to memorize were from 15? 10? clear up to 50.

[Arcana]

And I won't point out that the more I attack with my -defense rad power, the better my chances to hit the next time. ;)

That's actually worth pointing out, because those are the special cases where the general rules of slotting and enhancement might not apply.  But we're talking mostly about general rules of thumb, not special cases like attacks with defense or resistance debuffs.  In those cases, the analysis of the best slotting becomes more involved, and more complicated.

[Sinistar]

My main blaster had a T4 Cardiac Core and T4 Musculature Core for her Alpha slot.

I'd slot which ever I felt was best suited for the Trial/TF.  usually went with the damage boost.

[Arcana]

One thing I wished the devs would have implemented is some sort of character bound IO recipe system.  I think that once a character acquired a specific IO recipe then they should be able to use that recipe multiple times instead of having to find/purchase more recipes of the same exact IO.  I also think that when considering that we could have multiple builds, our enhancements should have been able to span across each of our builds instead of requiring us to acquire a whole new builds worth of enhancements for each build.  Both of these ideas would likely reduce grinding time and affect the auction house.

I wish the devs didn't make up a ton of names for things first, then try to figure out how to make them useful.  But that's another story.

[brothermutant]

Thanks Arcana for an always enlightened way of explaining things. Very helpful and indeed a good way to look at how to check which way is the "best" for power slotting choices.

And its not that I am unaware that an additional slot at tier 4 would be miniscule in comparison to what the normal Alpha slot did. Its that I didn't like what the Alpha slot did to toons as a whole. To me, when I pugged and exemped down to find most people were crippled because they relied on their Alpha slot to cover dmg or acc or end issues, that was lazy planning...by both the player and the devs. They gave us too much, if you will, with that one slot. I felt a more gradual increase to your powers would have been more enjoyable.

And I wouldn't suggest that just a slot be the reward. I would actually suggest powers like the Lore Pets or the Nuke, etc be a choice for which Alpha tree to go for, but have a free slot be given to the player once tier 4 was reached. And I didn't like the level shifts either. Thought they should have just made an extra level instead of the 50+1. That was weird to me that I was 50+1 instead of 51.