What Are Pokemon?

  • Pocket Monsters, or Pokemon for short, are creatures that are typically captured and trained by humans for companionship and/or used in fighting competitions.
  • Since 1996, there have been many generations of Pokemon– each generation adds hundreds of new Pokemon to the total number of Pokemon.
  • A single Pokemon can have elemental-like “types” - such as Water, Fire, or Grass - and a Pokemon can have two types (also referred to as “a double-typed Pokemon”), such as “Ground-Rock” typing
  • A Pokemon also has individual statistics (“stats”), such as Weight, Height, Speed, and Attack which we will analyze.

Pokemon Data

We Had the following data sets from 2 separate sources that had different information about each Pokemon.

##   X.      Name Type.1 Type.2 HP Attack Defense Sp..Atk Sp..Def Speed
## 1  1 Bulbasaur  Grass Poison 45     49      49      65      65    45
##   id identifier species_id height weight base_experience
## 1  1  bulbasaur          1      7     69              64

Combining The Data Frames

# subset to have only pokemon number, name, height, and weight 
pkmnweihei = subset(pkmnweihei, is_default == 1 & species_id <= 721, 
                    select = c(species_id, identifier, height, weight))
# subset to have only pokemon number, name, Type1, Type2, Generation
pkmnstat = subset(pkmnstat, 
                  select = c(X., Name, Type.1, Type.2, Generation, Speed))
# remove duplicate pokedex numbers (alternate forms)
pkmnstat = filter(pkmnstat, 
                  !duplicated(pkmnstat[,"X."]) | (X.==413 | X.==479))
# merge by pokedex number
pkmnstat = rename(pkmnstat, species_id = X.)
pkmn = merge(pkmnstat, pkmnweihei, by = "species_id", all = TRUE)
pkmn = subset(pkmn, select = -c(identifier))
head(pkmn,1)
##   species_id      Name Type.1 Type.2 Generation Speed height weight
## 1          1 Bulbasaur  Grass Poison          1    45      7     69

Type Breakdown Across Generations

- Across the generations, there tend to be more single-typed Pokemon than double-typed, with the exception of Generation 4 and 6.

Single Type Breakdowns Across All Pokemon

- The most common single types were Water and Normal, while the least common was Flying.

Do Weight and Height Correlate With Speed?

- There seems to be no notable relationship between these three variables.

Comparing Weight Across Generations

- Across all generations, there seem to be similar medians that seem skewed right with notable outliers in each generation.

Comparing Height Across Generations

- Similar to weight, all generations are skewed left on height, with notable outliers for each. Interestingly, Generation 6 had a lower median than the rest of the generations.

Analysis of Weight Across All Generations

##   Min LowerQuartile Median UpperQuartile Max
## 1 0.1             9     28         60.65 950
## [1] "Mean: 56.35"
## [1] "Standard Deviation:  89.83"
  • We see that the average is a lot higher than the median, which is caused by the outliers in the data. Moreover, the large standard deviation shows a wide range of variables due to these outliers. Lastly, we can see a trend of the data being skewed left from the closer min, LQ and median.