Now reading...
Choose your starter: Looking at attack and defense of multitypes analytically

Pokemon X & Y Starter wallpaper by Arkeis

Pokemon X & Y Starter wallpaper by Arkeis [via deviantArt]

Last week, we looked at Pokemon types and suggested a leadoff batter for your travels through the unknown world of Kalos.  Looks a bit green to me, but the classic question, the first question (other than name, sex, age, birthday for your creeper professor friend) is:  What Pokemon to choose?  This week, we’ll sift through the possibilities that are available and analytically determine if there’s one that will provide the best mix of offense and defense, even though we don’t know much about the starters at this point.

Read more after the break

Grass, Fire, Water.  That’s your options.  Except it’s not.  Dual types makes life difficult because instead of 18 types (with fairy now), you actually have… 324 types to deal with (an aside: I actually messed this up.  18 types x 17 possible other types + 18 more for just ‘null’ or single type = 324).  Luckily, some type combinations don’t exist… yet… T__T.  How are we going to break this down?

First off, last time, we talked about defense as a function of the possible attack type against your Pokemon’s type.  Your ‘AC’ (Armour Class) in D&D terms is really your own Pokemon’s type.  Your possible attacks may relate to your Pokemon’s type, but doesn’t necessarily.  Luckily, attacks are single types only which made analysis a bit easier because we can model the threats (except of course, fairy).

Now, we are modelling defense as well as potential attacks from starters.  We’ll have to forecast what attack types will be available to each starter as well as forecast potential trainers out there.  Luckily, Kalos looks pretty green unless Kyogre raises the water or Zapdos, Articuno, and Moltres decide that they are going to freak out again (and the world will once again, have to turn to Ash, the perpetually 12 year old boy, not the carbon compound from burning).  Though, if Game Freak had a sense of humor, the Kalos images we’ve been accustomed to might just be “back 100 years ago, Kalos was green.  Now, it’s a desolate world…”

Kalos-Region-Pokemon-X-and-Y

Once again, lets make some assumptions:  The starters will have a few good moves at all levels that will be fire/water/grass.  So we can safely assume that we get one of those three types for attack.  Second, we assume that there won’t be any mixed gym types, and no surprises.  Imagine if Blaine (R/B/Y) had a secret Venasaur to dominate the water types you brought in, just for lolz.  There are 8 gym leaders and 4 Elite Four trainers to contend with.  (The champion always variety, right?  A side note:  I built my team in R/B/Y to defeat Alakazam.)  There will also not be two duplicate gym types.  That allows for over 18,564 possible combinations of gym leaders throughout the game!  (18 types, choose 12).

 Lets break it down by defense and offense.  On the defensive side, we saw that the strongest attacks are fire types, with grass types no slouch either.  Water types, on average, have about a 10 power difference than grass (which is also about 10 power difference away from fire).  On this front, defending against fire might be your best bet.  However, we have to dig a little deeper.

We’ll most likely see some dual type starters, and perhaps not at the start.  What can we say about dual types on defense?  I computed “Net Weaknesses” by dual type of the possible starters, including types that we’ve never, ever seen.  Net Weaknesses is computed by adding together all weaknesses and resistance together.  “No effects” equates to a double resistance in this.  To simply compute this, I took the product of a standard weakness and resistance table with data inputted as 2x and 1/2x.  In this case, 2x and 1/2x will cancel out (as well as double resistance/weaknesses).  Take a logarithm to convert the result into the power, which tells you how many weaknesses or resistances there are. Here’s the formula below:

Formula for attack:defense(note:  Pi the multiplication equivalent of sigma, which is “sum.”  The backwards triangle is “for all”)

DEFENSE ↴Net Weak DEFENSE ↴Net WeakDEFENSE ↴Net Weak
FIRE/-3WATER/-2GRASS/-5
FIRE/NORMAL2WATER/NORMAL1GRASS/NORMAL4
FIRE/WATER3WATER/FIRE3GRASS/FIRE3
FIRE/ELECTRIC4WATER/ELECTRIC2GRASS/WATER3
FIRE/GRASS3WATER/GRASS3GRASS/ELECTRIC4
FIRE/ICE5WATER/ICE4GRASS/ICE8
FIRE/FIGHTING4WATER/FIGHTING4GRASS/FIGHTING6
FIRE/POISON5WATER/POISON3GRASS/POISON4
FIRE/GROUND1WATER/GROUND0GRASS/GROUND3
FIRE/FLYING2WATER/FLYING1GRASS/FLYING4
FIRE/PSYCHIC5WATER/PSYCHIC5GRASS/PSYCHIC8
FIRE/BUG4WATER/BUG3GRASS/BUG8
FIRE/ROCK6WATER/ROCK5GRASS/ROCK4
FIRE/GHOST2WATER/GHOST1GRASS/GHOST2
FIRE/DRAGON3WATER/DRAGON1GRASS/DRAGON6
FIRE/DARK2WATER/DARK2GRASS/DARK5
FIRE/STEEL2WATER/STEEL1GRASS/STEEL1
SUM56SUM41SUM78

The data’s available here from PokemonDB.  The first striking thing is that each additional type adds a liability against attack types in almost all cases for fire and water.  If you get a grass type, 8 dual grass/X types will provide you with some benefit, though starting with a net of 5 weaknesses and getting a result that’s less than 5 isn’t really remarkable.  Surprisingly, a water/ground type is about par for weaknesses and resistances.  Note, I assume that each type will attack with equal probability, which probably very true either…  And we haven’t taken into account fairy types yet, though lets be honest, I don’t expect fairy type to make this analysis moot.

Comparing the possibilities above, it’s safe to say that water’s got the least amount of net weaknesses over the entire possible spectrum.  Go water?

We noted above that water has the lower attack power on average compared to fire and grass.  Furthermore, a net weakness hides the variability in that number.  You don’t necessarily want to run around with 9 weaknesses and 9 resistances.  That makes for hard battling.  Luckily, I don’t believe that any of the types for water have that property.

In the attack arena, we can perform the same analysis, but sum the possible attacks against each defense type.  Likewise, we take the logarithm, base 2, of all interactions for each type we are interested in.

FireWaterGrass
1008157

What this shows is that 100 of the 326 possible types are weak against fire attacks (and 81 and 57 for water and grass types).  Once again, this includes all possible types and doesn’t take into account the likelihood of each possible type on defense.  So take these numbers with a grain of salt.  We can’t necessarily just say fire is the best.

Fire types do have their advantages though on attack: In a green world with way too many bug catchers, a fire type can handle a good amount of battles while walking around outside.  In a cave, you’ll probably want to switch to a good grass or water type, both of which are more plentiful in a green world with water, unless Vulpix starts to appear in tall grass.  Which brings me to the next point.

Fire types are hard to replace while looking for Pokemon.  Grass types are plentiful and water types are easier to obtain by surfing or with fishing rods.  The replaceability of fire types at the beginning of the game is low, and that may make having a fire type from the start a better choice.  Combined with decent defense and a great offensive moveset, fire types probably have a slight advantage compared to grass types and water types, even considering the secondary type of the starter is still unknown.

Sorry Harry, but I’m choosing a fire type.

 

 


Ongoing Conversation