VGC11 Defensive Typing Ranks

What are the best defensive typings in the VGC11 meta?

Measuring Type Effectiveness

As last time we discussed coverage rankings for offensive moves, it is only natural that we now look on the defensive side of things to ask ourselves: what are the best defensive typings? I was inspired to think about an analytical way to approach this after seeing this wolfey video:

Amazingly, wolfey approximates what the data tells us about type resistances through his own intuition and reasoning on the types. Mortals can approximate his results by looking at the average resistance for each type combination, which we’ll do later. We won’t get all the nuances he does, but as we’ll see it’s a good rough guide. But before we get to that, let’s talk about fitting the meta into this and move beyond asking what is good in general to what is good right now?

Factoring in the Meta

Using a simple average resistance approximation makes for a good starting place. It will also have a certain gravity to it so that if the meta ever gets off balance, we could probably expect it to return to its “natural” equilibrium. We simply enumerate all the possible combinations of offensive typing vs single and dual defensive typings and compute the average resistance for each defensive combination.

But how can we look deeper into the defensive typings in the current meta? Let’s look at all the factors we need to consider:

• Which Pokemon are being used most heavily?
• What are the most common attack moves on those Pokemon?
  • What is the base power of the move?
  • What is the accuracy of the move?
  • Is the move STAB on the given Pokemon?
  • What is the attack or special attack power of the Pokemon using the move?

And that’s not even everything, really. We could make adjustments here and there for things like when items are heavily used to alter power. Certainly Dynamax matters, but honestly I have no idea how to account for dynamax moves. For the meantime, though, the factors above are the ones I consider in my calculations.

So now we take each possible defensive typing combination and compute the weighted average resistance. This weight is

\[Pokemon Utilization * Move Utilization * Base Power * STAB * (Attack/100or SpAttack/100)\]

These weights ensure that we’re not counting Shadow Sneak from a Mimikyu as equally worth our defensive attention as a Glacial Lance from Calyrex Ice Rider.

We do a sumproduct over the defensive typing combination by sum(weight*Resistance Factor) and compute the weighted average by dividing by the sum of the weights.

Note that I only used the moves of Pokemon with greater than 5% usage in the meta right now. I had to hand-curate the data and it was kind of a pain, so I’ll be more inclusive when I get this process streamlined.

So let’s examine the weights and make sure there’s nothing crazy (I had a typo earlier that made mud shot 50x more important than Behemoth Blade):

Table of Move Weights

So these are the most dominant moves in the meta combining the strength of the Pokemon that use them and the other factors we mentioned.

We can also aggregate these weights by move type to see which types are most dominant offensively.

Dominant Meta Attack Types

Note this answers a different question from my previous post on Offensive Power Rankings by type. While the previous post looked at what would be good into the meta type-wise, this now answers the question of what is dominant in the current meta. I am not necessarily making the claim that the meta is sub-optimal: it’s possible these move types have risen to the top for other factors such as Base Power, STAB, etc that offset their typing handicap. However, we should also not assume that the meta is perfectly efficient. There may exist opportunities for typing arbitrage within the current meta.

Physical/Special Split

We can also aggregate on attack type and see that the offensive meta is leaning Physical right now:

So which Pokemon are best equipped to deal with these threats? In the next section we’ll look at the typing component of this question.

Evaluating Defensive Typings Against the Meta

Here’s the result of computing the weighted average resistance for each type, sorted by best defensive typing for the meta. Keep in mind, we’re evaluating the defensive typing, so the defense stat of a Pokemon is not taken into account and will have to be done by the user. I would like to eventually work towards a Pokemon-level defensive ranking but small steps.

Using the Table

Perhaps the most interesting result from this is that we can quantify where different typings stand in the current meta. Using the table above, you can try a few things to get a read on the field:

• Sort by Meta-Weighted Resistance to get the currently best defensive typings. It’s electric/flying! Maybe Zapdos deserves some more love

• Sort and Reverse Sort by Average Resistance to see why Fairy/Steel was so good in the first place and why Aurorus is truly terrible. We can also see that the meta has balanced for Zacian, bringing Fairy/Steel from .74 all the way back 1!

Meta Warp: The Battlefield Got Weird!

However, the most interesting part to me is the Meta Skew, which is derived simply by subtracting the average resistance from the weighted average resistance. High positive values indicate the meta has evolved to make them much less viable than they would be normally.

On the flip side, the high negative values indicate the meta has opened up to allow better than average viability for other Pokemon. Whether these Pokemon have the other factors needed to work themselves into the meta remains to be seen. For example, Dragon/Grass has high skew making it more attractive than it normally would be, but at 1.07 it’s still a fairly middling option.

I’m not good enough to know how good these Pokemon are in general, but if you’re looking for an opening in the meta, hopefully negative skew can help you!

EDITS

An early version of this post had Coalossal as worst Meta-Weighted typing, but this was due to a data entry error on my part that gave Mud Shot waaay too much weight.