Typical Movesets used in September

[X-Act]

I'll provide the following example to clarify how the program works.

Suppose we have a hypothetical Pokemon with the following move usages:

Move 1: 83.3%
Move 2: 83.3%
Move 3: 66.7%
Move 4: 66.7%
Move 5: 50%
Move 6: 50%

It seems obvious from the above data that a moveset containing Moves 1, 2, 3 and 4 would have been used by at least one person.

However, consider the following:

User 1's Moveset: Move 1, Move 2, Move 3, Move 5
User 2's Moveset: Move 1, Move 2, Move 3, Move 5
User 3's Moveset: Move 1, Move 2, Move 4, Move 5
User 4's Moveset: Move 1, Move 2, Move 4, Move 6
User 5's Moveset: Move 1, Move 3, Move 4, Move 6
User 6's Moveset: Move 2, Move 3, Move 4, Move 6

If you count, Moves 1 and 2 are used by 5 out of 6 players (83.3%), Moves 3 and 4 are used by 4 out of 6 players (66.7%) and Moves 5 and 6 are used by 3 out of 6 players (50%), so they correspond exactly to our move usage percentages. However, none of the users used Moves 1, 2, 3 and 4 in the same moveset! The reason is that the summation of their percentages is 300%... which doesn't exceed 300%!

This is a worst-case example, but, as you can see, it can happen. The typical moveset for this Pokemon would be:

- Move 1
- Move 2
- Move 3
- Move 4
 
Comments:
* Move 5 can replace any of the four moves listed.

In this case, we see that such a moveset (Move 1, Move 2, Move 3, Move 5 or Move 1, Move 2, Move 4, Move 5) is in the movesets posted. And, not only that: however you write your movesets, you'll be certain that at least one moveset will satisfy the typical moveset posted above.

 

[X-Act]

Inspired by Hipmonlee's post, I've modified the algorithm slightly as it had an error (it was actually needlessly complicated). Now it is displaying the correct movesets. Sorry for the confusion.

Basically, what Hipmonlee said was right! (Check out Abomasnow's moveset now.)

 

[Caelum]

Excellent work X-Act.

This is more a question for Doug, but its related to X-Acts stuff. Is it possible under the current collection method to obtain the necessary data to match move sets & EVs?

Edit: Oops, forgot to 5 star this. Great thread, as usually X-Act.

 

[X-Act]

Just to be clear, the algorithm now is very simple:

1. Sum up the move percentages one by one until the percentage exceeds 300%. Put the move names summed up in a list.
2. Output the first four moves produced in the above list as the typical moveset.
3. Output the remaining moves, if any, from the list as options that can replace any of the four moves in the typical moveset.

Very simple... yet produces the correct typical moveset!

 

[Jenigmat]

Nice job X-Act, you're fucking amazing. Lombre's quite interesting, 3 usages and 9 different moves.

 

[nardd]

very cool stuff, thanks a bunch. Should be helpful in seeing what's out there when making teams and such- especially if you are making one of those "anti-metagame" teams

 

[TPX_Vengeance]

39. KINGDRA (Usages: 18521)

Typical moveset:
- Waterfall
- Dragon Dance
- Surf
- Rain Dance

Funny, what one month and one new move can do for a pokemon.

 

[reachzero]

Did anyone else think it was interesting that Leech Seed was not on the typical Skymin moveset? Perhaps Subseed Skymin is not as common as we thought.