New DRK is awful. Here’s why and what could be done about it.

[Lyth]

When you add your original expression and its complement together, it doesn't "just" have the property of acting as a complete combo, it is a complete combo. Your sum is just showing you the average potency associated with replacing 2x+1 combos with 6x+3 Bloodspillers. You could just create an expression for n combos and 3n Bloodspillers and get the same result.

If you had an expression for x combos and part of another combo, then you would have retained your indexed variable (using 'i' as an index is incredibly distracting, by the way, for reasons we need not go into here). We already know from looking at specific cases that the average itself is sequence dependent, and you should be able to derive any of these from a general expression.

Again, you're not going to get a simple average without knowing the entire sequence of GCDs that you've replaced. But everything we've done to this point has been contingent on treating blood algebraically. Even in your own analysis, blood has the same value on either side of the equation.

That being said, the point isn't to figure out the average across an entire fight. For starters, when playing, we know what the average is. That's whatever dps comes out in the wash. We're also not dealing with 'true potencies', as was pointed out, because some of these things are resource costs.

The question that you're generally asking yourself is: What is the value of this next Bloodspiller? How does it compare to an alternative course of action?

[Chrono_Rising]

When you add your original expression and its complement together, it doesn't "just" have the property of acting as a complete combo, it is a complete combo. Your sum is just showing you the average potency associated with replacing 2x+1 combos with 6x+3 Bloodspillers. You could just create an expression for n combos and 3n Bloodspillers and get the same result.

Yes, it is a neat calculational trick, but since I take an average to be the sum of all possible events weighted by their probabilities both exist in my sum of terms, and that allows me to combine things in unique ways and find new relationships. The difference is purely technical, but yes because of the structure of what is gained and lost it turns out you can analyze completed combos. But this isn't controversial, it is a mathematical definition of average. I say this acts like a combo for the technical reason that those two enders of a rotation represent two different rotations, but addition doesn't care where they came from, I add in the realization that this functions like a completed combo and add the terms together realizing that those would then take on constant nonvariable values. The compliment trick just allows me to remove the variation of 18 of my 19 variables (not to mention each blood has its own value), and x canceled itself out later down the road. It is semi in the spirit of the proof of divergence of the harmonic series. If you sum up the terms in particular groupings they add up to more than a half, and a half added up infinitely many times diverges.... Well its similar in my mind.

TLDR; for this section: Yes, I could have started with n combos, but the justification for why I can do that is in the work that I provided.

If you had an expression for x combos and part of another combo, then you would have retained your indexed variable (using 'i' as an index is incredibly distracting, by the way, for reasons we need not go into here). We already know from looking at specific cases that the average itself is sequence dependent, and you should be able to derive any of these from a general expression.

Yeah, would have looked much nicer in LaTex but they don't have that as an editing option on the forums saddly. And an index set for quantities seemed at the time less cumbersome than 6 different letters for each of the 3 variable sets (18 total letter variable names LOL).

We already know from looking at specific cases that the average itself is sequence dependent, and you should be able to derive any of these from a general expression.

Yes, but I think you might have missed what I did. Yes, each specific combo ender sequence has its own average potency, but the true average potency of the population is when we take all sequences together. What I did was to tie how the individual sequences co-vary to eachother and showed that the end result is a constant. One might be higher than 191, but it is paired with another sequence which is as low as its compliment is high. That is ok, this is the meaning of average, the specific cases will vary, yes, but all sequences taken together do have a well defined average value.

Again, you're not going to get a simple average without knowing the entire sequence of GCDs that you've replaced. But everything we've done to this point has been contingent on treating blood algebraically. Even in your own analysis, blood has the same value on either side of the equation.

But that is also the difference in our calculations, and it works because the blood on both sides is the blood contribution from all possible combos and not simply soul eater, the same in meaning and value, and because I can sum over all possibilities I can use a property of the average value to make a sneaky replacement. If I add 1+2+3 the result is the same as 2+2+2, infact this is true of all averages: If I add up a list of y random numbers its the same as adding the average value y times. I did leave this step implict in what I did, so you are right to point it out. The difference is that in your calculation you cannot make the same claim as I do (adding up all the possibilities is the same as adding up all the averages the same number of times) because you only have blood as part of the soul eater combo and do not combine it with any other contributions, and that is the source of our difference. The 20 blood which is part of the soul eater combo is an average value, but the 50 blood on the left hand side of your calculation is the worth of 50 blood when replacing the soul eater combo, but those are different in value, which is why 50 blood + 20 blood =/= 70 blood in your calculation, on average blood is worth more than the blood which replaces soul eater. In mine, when I add everything together, I slide in the average value in place of adding up all the blood values, which are different and varied, I replace those sums with average values which tames the rest of my variability.

TLDR; for this section: My sums add up all possibilities and so the addends can be replaces with the average value everywhere, in yours you are only adding a very specific blood contribution and then averaging and this doesn't work the same way. 1+2+3 = 2+2+2 but if I replace 1 random value with the average blood cost I might not preserve the sum... 1+2+3 =/=1+2+2

That being said, the point isn't to figure out the average across an entire fight. For starters, when playing, we know what the average is. That's whatever dps comes out in the wash. We're also not dealing with 'true potencies', as was pointed out, because some of these things are resource costs.

The question that you're generally asking yourself is: What is the value of this next Bloodspiller? How does it compare to an alternative course of action?

If you look close enough the proof starts with arbitrary combo length, and ends without a combo length, meaning that this isn't just a long term average. If you average together losing 1 hardslash, one syphon strike, one soul eater, hardslash syphon strike, syphon strike soul eater, and soul eater hardslash you will get the same average provided you do not analyze whether or not the blood gained is useable or not. Because there is no reference to limiting values and the combo length eventually drops out of the equation, the averaging works out to exactly 50*650/170 provided you average all the possible combo lengths pushed out.

However, and this is the big statistics lesson: The average gained is 191 is not the same as THIS bloodspiller is worth... It represents an average of a range of values. The average male height is 5 foot 9 inches, that doesn't mean every male is 5 foot 9 inches***. If you want to know how much a specific bloodspiller in your combo is worth, well MaraD has the right of it, you need to analyze the use of that bloodspiller.

If you want to just know on average how much potency your bloodspiller gains you, the answer is the average value, about 191 potency over your 3 step combo.

Personally here is what I think is important, Bloodspiller is always a gain and this is a good thing. Dark Knight has no way to manipulate its GCD use anymore, so just use it and don't overcap. We no longer have the ability to force an extra bloodspiller into our rotation, they come up and we use them (preferable in buff windows). I'm a stickler here because other jobs have a guage, and has a similar calculation, and there it has a difficult to manage positive feedback term that could get out of hand if analyzed improperly, and it is very easy to underestimate its value in a similar way... I'm sure you can appreciate where that might go.

***Full disclaimer I have no idea what the average male height is and made up a number that sounded reasonable.

The question that you're generally asking yourself is: What is the value of this next Bloodspiller? How does it compare to an alternative course of action?

And this is the last thing I'm going to respond to before bed. I would argue here that the correct answer is either (1) analyze your rotation for the right number or (2) 191 potency. Why? Because if I'm only interested in this bloodspiller than I really need to analyze this bloodspiller. But if I ask this question every time I use bloodspiller, then my question is going to become the average value. This is because not only is 191 the average value of the population, it is also a horizontal asymptote or long term trending value. If I ask this question enough then I start getting into the relm of asking questions about the population value. However, keep in mind the original question was not framed around analyzing a specific bloodspiller use, it was about the average bloodspiller potency over the GCD in which you claimed it would be 202 potency, I have presented arguments, examples, and now proofs to show that it is not 202 and is 191, to which we are now saying we are more interested in specific uses of bloodspiller, not the average. But this new question is boardline goalpost moving and the 202 potency does not answer this new question raised either.



Based on Final Potency Bloodspiller appears on average to be a gain of 50*800/170.

[shao32]

well there it is, here we have the oficial numbers and nothing change, how the skills works is the same like in the pax, is hilarously watch all the nerfs, idk if im going to be able to complete the story spaming soul eater without kill my self and pick GNB.

[SummerScorcher]

Abyssal drain still 200 potency....

[T-Owl]

Should we rather start talking about GNB here, considering that nobody here will keep playing DRK with ShB?

[Gula]

Should we rather start talking about GNB here, considering that nobody here will keep playing DRK with ShB?

DRK since HW and forever. Only cowards can't face their inner darkness.

[T-Owl]

DRK since HW and forever. Only cowards can't face their inner darkness.

I guess the job could be practical if you want to do something in the meantime while playing, like reading, cooking or finally writing the epic book-series you always wanted to write. I mean, it is slow enough that you can do plenty of things inbetween the GCDs.

[Gula]

I guess the job could be practical if you want to do something in the meantime while playing, like reading, cooking or finally writing the epic book-series you always wanted to write. I mean, it is slow enough that you can do plenty of things inbetween the GCDs.

That's just like, your opinion man.

Are you honestly trying to dissuade me from my enjoyment of the game I bought and continue to pay for?

[DRKoftheAzure]

Yeah, DRKs are going to be using Stalwart Salt combo before every Saltleater combo... and yes those typos were intentional... At least Gnashing Fang combo will always be up by the 3rd or 4th Solid Barrel, so I wil be having fun on GNB

[Jandor]

That's just like, your opinion man.

Dark Knights speed isn't really an opinion.