What does DPS mean Pokemon go?

DPS, or Damage over Time, is a common stat that we use to determine how good or bad a certain move is. It is simply calculated by raw damage divided by animation time. From this definition, it does make sense for us to use DPS as a power rating metric, and for the most part it worked when Pokemon Go launched. However, we must be careful when religiously using DPS. While it is useful, the recent balance changes that makes charged moves much more beneficial to use require more analysis from us than simply using raw DPS stats.

This post aims to give an intuitive feel of what DPS means and how it is used in calculations. I do realize that since this is The Silph Road, many are mathematically strong and already know what I'm about to say in this post. But for those who do not, I hope that this will help.

Let's click on this link about Snorlax in Pokemon Go and investigate its moves' DPS. Scrolling down, you'll see that Lick has 10.00 DPS compared to Zen Headbutt's 10.48 DPS, so Zen Headbutt has higher DPS. If you're curious, here are now they are calculated:

Now, below, you should see Snorlax's charged moves, Body Slam, Earthquake, and Hyper Beam. Here, we'll focus on the two highest DPS charged moves, Body Slam with 25.64 DPS compared to Hyper Beam with 24.00 DPS. It seems that Body Slam has higher DPS. If you're curious, here are now they are calculated:

Finally, take another scroll and look at the recommended offensive moveset: Lick and Hyper Beam. Wait, what? How did the two lower DPS moves combine together to great the better offensive moveset? Maybe the site got it wrong, so let's refer to another source. Take a look at dneal12's Attacking Optimal Moveset post. You'll also see that the best Snorlax moveset is indeed Lick and Hyper Beam. So what kind of magic happened and how did they arrive to this conclusion?

Before we move on, it's good to review the impact of the recent balance changes. When Pokemon Go was first released, everyone was saying to spam tap and ignore charged moves most of the time, right? This is true because charged moves were mostly not worth using, except in rare cases such as Aqua Tail. Because of this, we could simply consider fast moves' DPS and it would be accurate. Now, the reverse happens. In most cases, it's worth it to use charged moves ASAP for highest damage, unless you're saving it strategically. Because of this change, one more stat becomes vitally important: energy per second (EPS). It makes sense, right? Since charged moves are more important, we also need to worry about the road it takes to use charged moves, so we factor in energy generation. Now let's look at Lick and Zen Headbutt's EPS.

Now we know that Lick generates more EPS than Zen Headbutt. Despite Lick having lower DPS, it generates more energy which allows for faster use of charged moves. This is the reason why Lick is used in Snorlax's optimal attacking moveset, not Zen Headbutt.

For Charged moves, DPS is actually...largely irrelevant, and we care more about raw power. Let's think about this for a second. Looking at the DPS of charged moves is like seeing what would happen if you used Body Slam over and over again. As we all know, this is not how Pokemon Go works, because charged moves are not spammable. If so, then what is the better damage indication if it's not DPS? For that, we would need to consider what truly matters, which is what makes using charged moves possible: energy. Let's say we use Lick for our fast move, which has 12 EPS. One full charge bar will allow us to use Hyper Beam once (120 pwr) or Body Slam twice (80 pwr). Now I will calculate what I call eDPS, or energy dependent DPS, based on 100 energy.

Here we can see that when we factor in the energy required to execute charged moves, Hyper Beam comes out on top. You may be concerned that the eDPS of charged moves are lower than the DPS of fast moves, but both are not directly comparable--that's what DPS is for. In short, for charged moves, DPS is used to see if you should use a charged move over spamming fast moves, while eDPS compares charged moves against each other. Based on this, Lick + Hyper Beam is the superior moveset.

Hold on. Let's talk about another scenario. Remember that from the time it takes to use Hyper Beam once, you can use Body Slam twice, right? What if you attacked enough to generate 150 energy? In this scenario, you can actually use Body Slam three times, but Hyper Beam still only once! Here are the calculations:

So there we go! In this case, Body Slam actually performs better than Hyper Beam. In fact, it does seem like half of the time (before you can use the 1st, 3rd, 5th, etc... Body Slam), Hyper Beam is better, and the other half of the time (after you can use the 1st, 3rd, 5th, etc... Body Slam), Body Slam is better. While my calculations are simplistic and the true damage formula is more complicated, it's not far off. It's been calculated that for battles lasting 40s or shorter, Body Slam does more damage for 18 of those seconds, which is 45% of the time. After 40s, Hyper Beam does completely dominate, but battles rarely last that long. Therefore, about 45% of the time, Lick + Body Slam actually does more damage than Lick + Hyper Beam.

Not yet. If you do the same calculations but ignore animation time (since gyms attack at the same speed regardless of attack animation), then Zen Headbutt + Hyper Beam is theoretically the best defensive moveset. Why? It's because at a fixed 1.5s attack cooldown, Zen Headbutt generates more EPS.

And as we all know, this means you will be be using more charged moves when defending with Zen Headbutt. If you have Lick, use it offensively; if you have Zen Headbutt, use it defensively.

Note quite! Now we leave the realm of quantitative analysis and move into the realm of qualitative analysis. Although Zen Headbutt + Hyper Beam is theoretically the better defensive moveset, many have noted that in practice, Hyper Beam is much easier to dodge than Body Slam or Earthquake. This means that potentially, Hyper Beam does no damage and we just wasted the 5 seconds of animation time for a dodge! With this in mind, many people prefer to have Zen Headbutt + Body Slam on their gym guardian Snorlax, because it's much harder to dodge Body Slam and challengers will take more damage overall than Hyper Beam. Of course, we have no clue how many challengers are good dodgers and how many just ignore the dodge mechanic, so we cannot make any conclusive decision. If challengers dodge, Body Slam is better; if challengers do not dodge, Hyper Beam is better. Other factors we should consider include:

If the battle lasts too short, Hyper Beam might not even fire before Snorlax falls.

Hyper Beam can result in an "overkill" and you could have already won earlier with Body Slam.

So, what's the best moveset? I know some of you hate to hear this, but I hope I've convinced you by now that it all depends. With the addition of charged moves into the mix, certain movesets have "peaks" that should be considered over other movesets. Note that DPS and eDPS are both theoretical metrics that assume the battle lasts infinitely, which in practice is not true. Therefore, we truly do need to take them with a grain of salt and consider what is best for your situation. Here's a quick summary:

Offensively, Lick + Hyper Beam is better, but just barely. About 55% of the time, you do more damage with this moveset.

Offensively, Lick + Body Slam is perfectly usable. You will do more damage about 45% of the time for battles under 40s.

Defensively, Zen Headbutt + Hyper Beam is better for challengers that do not dodge.

Defensively, Zen Headbutt + Body Slam is better for challengers that dodge frequently.

I know we've been talking about just Snorlax, but this applies to all Pokemon. Very often I see posts about a near perfect IV Pokemon but will be scrapped because of an "inferior" moveset. Well, let's not do that. In practice, this game is currently very balanced, and many movesets, although inferior, have some peaks where it's better than the best moveset. Other movesets are more suited for defending gyms instead of attacking, so you can use it that way also. Don't be quick to transfer Pokemon based on "DPS" or "Best Offensive Moveset", and instead learn how different situations where you could best use that Pokemon.

Thanks for reading, and I hope this has been helpful.

Players who battle in Pokemon GO care about damage output. While it’s easy to figure out which fast moves are better than others, charged moves are trickier to rank. Metrics classically used to compare charged moves suffer from huge flaws, and they cause misinformation due to their prevalent use among Pokemon GO resources.

Damage per second (DPS) is the earliest and most intuitive metric to compare charged moves. The concept of DPS is familiar to anyone who’s played an MMORPG or battle arena game, and it’s a highly useful way to compare fast moves. The concept is simple: moves that do the most damage in the shortest amount of time are the best.

The problem with DPS is that it overvalues low-power, short duration charged moves. DPS is only accurate if you’re using a move all the time. Charged moves cannot be used all the time by design; you have to charge them up first with a lower DPS fast move. Ranking by DPS leads to gross inaccuracies like Thunder > Wild Charge or Brave Bird > Sky Attack. 1-bar moves clutter the top of the list, while 3-bar moves concentrate near the bottom.

Damage per energy (DPE) attempts to counteract the flaws of ranking charged moves by DPS. The reason why moves such as Thunder and Brave Bird were poor despite their high DPS was that they made inefficient use of energy. In contrast, moves such as Shadow Ball and Leaf Blade performed well in practice despite their middling DPS because they did a lot of damage for only a partial energy charge.

The problem with DPE is that it undervalues 1-bar moves and wasn’t good for comparing moves with a different number of energy bars. For example, how does Dig (100 power, -50 energy, 4.7 s) compare to Earthquake (120 power, -100 energy, 3.6 s)? They’re about equally bad, but DPE can’t tell you how much of an impact Dig’s long duration has on its damage output.

We see from the above examples that DPS and DPE both have value in determining charged move quality. The natural next step would be to create a metric that assigns weight to both parameters. In the absence of strong evidence either way, I experimentally assigned equal weight to DPS and DPE, then studied the results. Here is a partial table of moves ranked by DPS * DPE:

Name	POWER	DURATION	ENERGY	DPS*DPE
Precipice Blades Origin Pulse	130	1.7	-100	99.412
Frenzy Plant	100	2.6	-50	76.923
Shadow Ball	100	3	-50	66.667
Solar Beam	180	4.9	-100	66.122
Overheat	160	4	-100	64
Sky Attack	80	2	-50	64
Draco Meteor	150	3.6	-100	62.5
Wild Charge	90	2.6	-50	62.308
Outrage	110	3.9	-50	62.051
Leaf Blade	70	2.4	-33	61.869
Avalanche Dynamic Punch	90	2.7	-50	60
Dazzling Gleam	100	3.5	-50	57.143
Focus Blast	140	3.5	-100	56
Play Rough	90	2.9	-50	55.862
Blizzard	130	3.1	-100	54.516
Zap Cannon	140	3.7	-100	52.973
Discharge	65	2.5	-33	51.212
Hydro Pump	130	3.3	-100	51.212
Thunderbolt	80	2.5	-50	51.2
Surf	65	1.7	-50	49.705
Ice Beam	90	3.3	-50	49.091
Foul Play	70	2	-50	49
Psyshock	65	2.7	-33	47.419
Rock Slide	80	2.7	-50	47.407
Fire Blast	140	4.2	-100	46.667
Heavy Slam	70	2.1	-50	46.667
Crunch	70	3.2	-33	46.402
Dragon Claw	50	1.7	-33	44.563
Flamethrower	70	2.2	-50	44.545
Close Combat Stone Edge	100	2.3	-100	43.478
Thunder	100	2.4	-100	41.667
Brave Bird	90	2	-100	40.5
Earthquake	120	3.6	-100	40
X-Scissor	45	1.6	-33	38.352
Flash Cannon	100	2.7	-100	37.037
Psychic	100	2.8	-100	35.714

Full table available in the 'Charged moves' tab

Yes. Charged moves known to be good, such as Shadow Ball and Leaf Blade, are ranked highly, while charged moves known to be poor, such as Thunder and Brave Bird, are ranked low (although there are many much worse charged moves). This is an improvement from pure DPS or DPE ranking.

Are charged moves properly ranked relative to each other?

Yes and no. The list indicates that Frenzy Plant > Solar Beam, Wild Charge > Zap Cannon, and Foul Play > Crunch, as examples. But there are also incorrect relative rankings, such as Blizzard > Ice Beam, Zap Cannon > Thunderbolt, and Fire Blast > Flamethrower. This is still an improvement from pure DPS or DPE ranking, however.

DPS * DPE still overvalues 1-bar charged moves, specifically those with very high DPS. This is illustrated by Precipice Blades and Origin Pulse topping the list by a huge margin, although in practice, they perform at the level of Frenzy Plant. Moves such as Fire Blast and Zap Cannon are widely considered to be inferior to their 2-bar counterparts, but at least this metric puts them within just a couple of points of each other.

Practical quirks in battle, such as fainting with wasted energy or prior to a charged move’s damage window, are not accounted for in this metric. One could include penalty multipliers for 1-bar moves and late damage windows, but that would introduce a lot of complexity for a little improvement in accuracy.

Finding a metric that 100% correctly ranks charged moves is almost impossible, because DPS and DPE carry different weight depending on a Pokemon’s fast move. However, I believe that DPS * DPE is a far superior ranking metric to either DPS or DPE alone, which is what almost all Pokemon GO resources use. While weighting can be adjusted to improve accuracy, DPS * DPE has the advantages of being elegant and easy to conceptualize.

GamePress took a step to move away from the charged move ranking by DPS paradigm by removing the DPS column from our charged move page. We also plan on removing charged move DPS from individual Pokemon pages. We will primarily be using DPS * DPE as an aggregate metric to express charged move quality.

Thanks for reading.