GamePress

DPEPS

Could I recommend adding another stat to gamepress?

DPEPS (Damage + Energy per second).
Since the shake up yesterday, Charged attacks are much much more important. So rather than just spamming water gun, lick, or mudshot, it is now very important that your Pokemon gets enough energy to execute those high DPS charged attacks.

Looking at the data you guys have collected, I made a preliminary list of S tier and A tier attacks as follows:

Top tier quick moves according to DPEPS (Damage + Energy per second)

S tier (30+)
Fury cutter 37.6

A tier (25-29.99)
Bug Bite 26.67
Bite/Dragon Breath/Water Gun/Scratch 26
Pound 25.92

B Tier (20-24.99)
Psycho Cut 24.56
Lick 24
Metal Claw 32.81
Mud shot 23.64
Vine Whip 21.54
Wing Attack 21.33
Low Kick/Thunder shock 20

Still unsure of how those two big EPS bug attacks will stand up in practice.

Hope this info is helpful to some people.

Cheers!

Asked by jamesn018 years 8 months ago

Could I recommend adding another stat to gamepress?

Looking at the data you guys have collected, I made a preliminary list of S tier and A tier attacks as follows:

Top tier quick moves according to DPEPS (Damage + Energy per second)

S tier (30+)
Fury cutter 37.6

A tier (25-29.99)
Bug Bite 26.67
Bite/Dragon Breath/Water Gun/Scratch 26
Pound 25.92

B Tier (20-24.99)
Psycho Cut 24.56
Lick 24
Metal Claw 32.81
Mud shot 23.64
Vine Whip 21.54
Wing Attack 21.33
Low Kick/Thunder shock 20

Still unsure of how those two big EPS bug attacks will stand up in practice.

Hope this info is helpful to some people.

Cheers!

Asked by jamesn018 years 8 months ago

Report

Answers

I wouldn't calculate it that way even if I was a DPS-nik. For various reasons, it makes no sense.

If the point of using a move like Fury Cutter is so you can charge, for instance, Cross Chop (not that I think two pokemon have that move combination, only they are the highest "EPS" and "DPS" moves, respectively*), then you'd expect the uniform "value" of Cross Chop to be more than that of Fury Cutter. The "DPS"+"EPS" (30.0+0.0) of Cross Chop should exceed Fury Cutter's (7.5+30.0). That it doesn't should reveal the problem. Before the secret update, there is a reason posters and sites generally prioritized the fastest and strongest fast moves.

(*There are a currently a lot of sites and posters going about combat statistics the wrong way. The focus on DPS is not absolutely misguided as it is a useful general metric, but it is creating some misplaced uniformity.)

My apologies if I didn't explain it clearer Gofessor Oak. DPEPS as a stat would only have relevant values for quick moves.

While this stat is still quite rudimentary, much similar to OBPS (On Base Plus Slugging) in baseball, it fits a current niche in the aftermath of the recent mass nerfing on quick moves.

Also, while it is still unclear how much exact value is gathered from EPS compared to DPS at this time, DPEPS should still have value.

A more complicated stat that could be derived would involve the total DPS of an entire chain of quick moves and the subsequent charged move. This stat would in theory be individualized for each possible quick + charged moveset.

Lastly, can Gamepress please add the damage windows for attacks? This is especially important for guaging the best gym defenders. All the data is readily available on thesilphroad.com, it would just be nice to be able to see it all on gamepress.

Sorry if my first post was unclear. I hope this makes sense now. You are also correct that there is no pokemon with fury cutter and cross chop.

To be clear I only intoned that your initial formulation had an obvious problem, I never said it was impossible that a Damage+Charge value formulation of whatever nature couldn't be composed, or even that I didn't have such a formula in mind. It was just a logical point: if there is now a point in choosing a lower DPS fast move in the current metagame so that a much higher DPS charge move can be used, then the focus is still on DPS same as it was in the prior environment, only now there are potentially new "routes" to higher DPS that look more heavily at charge moves. But, in such cases, since DPS is still the primary thing valued, you can't say charge is equal to DPS.

Another issue to be dealt with is the double weighting of damage in my previous formula since the DPS value would be present in both DPS itself as a statistic and also in EPS since more energy is generated with more damage. STAB is also an intervening variable of importance here that would have to be in the formula as a binary variable.

So to be clear, I think it is a little more complicated than just all bogging down to DPS. There's sort of two aggregate variables that would lead to a TDPS (Total DPS) for a chain of quick moves followed by a charge move. The DPEPS for the quick move, the DPS for the charged move, and finally the TDPS for the entire chain.

I'm glad we're on the same page now gofessor Oak. Also, you mentioned that the formula wouldn't work for various reasons. I'm curious as to what other reasons could be analyzed to create a more robust formula and what kind of formula you have in mind.

Cheers!

No, the energy generated is inherent to the move. A level 11 Scyther using Fury Cutter will not generate more energy than the same Scyther but level 12 using the move. "EPS" is actually generated the same way "DPS" is, by dividing the energy generated from a move (as opposed to its base damage) by its cooldown. DPS can of course be "weighted" because different pokemon have different attacks stats and/or different CPM (and it can also be "weighted" in other ways, such as by figuring DPS*SiB, although such involves figuring in the effect of usually unassociated statistics).

That speaks to the "double-weighing" that you mentioned--"double-weighing" in that form doesn't occur, because EPM (energy per move) is fixed. There is "double-weighing" though that you're not registering for the fact that the dps of a chain can never exceed the dps of one or the other move. It can in fact be equal in the case where it is found the higher DPS move is (still) the fast move though. If the DPS of two moves were equal, it wouldn't matter what the charge generated by the fast move were, unless the charge move offered a temporal type advantage.

Move optimality mostly does all bog down to dps in this game (although, technically, you're actually sort of right that it doesn't, for two-ish statistical reasons that I won't explain), the "complication" viz-a-viz the former system and now is only that charge moves are now the primary carriers of heavy DPS (wherein the former system charge and charge moves could often be irrelevant): i.e., we (usually) have to work with combinations instead of monolithic chains of fast moves.

Decent analysis and I see where you are going.

I'm still a little confused as to your original analysis of using cross chop (or really any charged move) in a DPEPS formula. Was this a mistake on your part or am I still not understanding something there?

Regarding the EPS for moves, I guess I was still going off the previous assumption that damage generated variable rates of energy. Now that I know this is fixed it makes life a lot simpler in regards to this formula.

As a side note, what does the acronym SiB stand for (Seconds in Battle)?

Edit: CPM stands for CP (Combat Power) multiplier

My point regarding Fury Cutter, Cross Chop, and "DP/EPS" is only that you're generating energy to unleash a higher DPS move. "Energy" is not a value in and of itself if the charge move is irrelevant, and, as you're generating energy with the point of only increasing damage, energy can't be weighted the same as damage such that the value of a fast move would exceed that of a charge move. (This is the area where we are liable to the above-mentioned "double-weighing", as you can't just treat the energy-generating move's produced energy as one and the same with the damage it is meant to unleash.) Charge moves interact with energy the same as do fast moves. It makes more sense to have a common analysis able to give fast and charge moves alike a value, not to have a separate analysis for fast moves merely adding "DPS" to "EPS".

Yeah, SiB stands for "seconds in battle", it's my personal counterpoint to all the DPS focus on this site when comparing pokemon like Exeggutor and Alakazam (although this happened before the move changes).

Here we go.

https://www.reddit.com/r/TheSilphRoad/comments/4uffha/pokemon_dps_total_damage_calculator/

After looking at the adjusted values of the EPS for the bug move outliers, I'm seeing a correlation between the simple metric of DPEPS and what turn out to be the tiers of quick moves.

Got some more ideas to make the metric a little more accurate and applicable to all move combos and have been following the various proposals for best move sets, whether from models including a pokemon's hp, one's accounting for integrating dodges into attack cycles, to the one proposed by GamePress.

Regardless, I still view DPEPS as a simple metric that people without the math abilities to understand formulas such as the following:

Variables:
In the following formulas [QM], [QMS], and [ChM] refer to Quick Move, Quick Move Series, and Charge Move, respectively.

A Quick Move series is the number of quick moves required to satisfy the CE of a Charge Move.

EPS = Energy Generated Per Second by a Quick Move
CD = Cool Down; The time it takes after initiating an attack before another action can be taken (either dodging or attacking again)
CE = Charged Energy needed to use a charge move

[QMS]CD (Series Cool Down) = x*[QM]CD
where x must be a positive integer

[QMS]C (Quick Move Series Charge) = [QMS]CD * EPS
where [QMS]C > [ChM]CE

[QMS]Atk (Quick Move Series Attack) = [QMS]C*[QM]Atk (Quick Move Attack)

MSDPS (Move Series Damage Per Second) = ([QMS]Atk + [ChM]Atk)/([QMS]CD + [ChM]CD)

Things to be added to this formula:

STAB should be integrated into [QM]Atk and [ChM]Atk w/ a binary variable

2. CRIT chance should be integrated into the [QM]Atk and [ChM]Atk (currently unaware of crit hit damage percentage)

Can still understand.

This a great idea that I agree with, but there isn't a good way to calculate it. As Gofessor said, it deends on the charge move of the Pokemon. We can use an average value like 60 for a charge move to calculate it, but it wouldn't be very useful because all Pokemon have different charge moves and such.

Creativi Timothy, correct. However, knowing the DPEPS of an attack should theoretically let you know how efficient it is at BOTH dealing damage and getting you closer to your charged attack. Seeing as all of the top DPS charged attacks are single bar (except Body Slam, Dragon Claw, and Sludge Bomb) and have a DPS range of 20.59 to 30 (Sludge Wave and Cross Chop respectively), having a pokemon with a high DPEPS and a single bar charged attack would seem to be the best way to ensure heavy damage when attacking gyms (or fighting PvP in future updates).

Case in point, I have noticed that my Parasects with bug bite/fury cutter and solar beam are very good at inflicting heavy damage very quickly. I can usually get off 2 solar beams before the gym defender is able to pop off one full bar charged attack. Unfortunately, due to their low stamina base stat they also take the damage quickly as well.

I expect Nidoking w/ fury cutter and earthquake to be a top tier pokemon if DPEPS is a valid stat.

As I previously stated, I am still unsure of the complete value of EPS vs. DPS, but with the increase in damage to charged attacks, whatever the previous ratio was would have tilted further in the direction of EPS having value.

That being said, I am still unsure of how DPEPS should be weighted in regards to DPS and EPS. There are many factors here that I am still trying to consider.

Any help on this idea would be great.

Cheers!