The data are supplied in a data frame, pkmn, with one row per currently-released Pokémon. This is a decoded version of the GAME_MASTER file included in the game package, with a few columns added on the end.
library(DT)
library(pkmngor)
datatable(pkmn, extensions='FixedColumns', options=list(fixedColumns=list(leftColumns=2), scrollX = TRUE))We can use tools from the dplyr package to explore and extract subsets of the data in a “tidy” way.
library(dplyr)For example, explore the chances that each Pokémon will attack at a certain point during the catch encounter. Slakoth is among the lowest, as might be expected from the idle creature.
pkmn %>% select(pokemonId, encounter.attackProbability) %>% arrange(encounter.attackProbability) %>% head## # A tibble: 6 x 2
## pokemonId encounter.attackProbability
## <chr> <dbl>
## 1 SLAKOTH 0.0100
## 2 SPINDA 0.0100
## 3 ABRA 0.0500
## 4 SLOWPOKE 0.0500
## 5 SLOWBRO 0.0500
## 6 SUDOWOODO 0.0500But what about the highest? I knew Tyranitar was pretty aggressive. But I was surprised to see it was Vigoroth, the first evolution of Slakoth. I look forward to encountering one of these in the wild!
pkmn %>% select(pokemonId, encounter.attackProbability) %>% arrange(desc(encounter.attackProbability)) %>% head## # A tibble: 6 x 2
## pokemonId encounter.attackProbability
## <chr> <dbl>
## 1 VIGOROTH 0.700
## 2 SHARPEDO 0.500
## 3 BEEDRILL 0.400
## 4 GYARADOS 0.400
## 5 EXPLOUD 0.400
## 6 PRIMEAPE 0.300There are ten variables in this data file that appear to govern how Pokémon behave during the encounter. The Silph Road have studied some of these in detail.
#[15] "encounter.movementType"
#[16] "encounter.movementTimerS"
#[17] "encounter.jumpTimeS"
#[18] "encounter.attackTimerS"
#[19] "encounter.attackProbability"
#[20] "encounter.dodgeProbability"
#[21] "encounter.dodgeDurationS"
#[22] "encounter.dodgeDistance"
#[24] "encounter.minPokemonActionFrequencyS"
#[25] "encounter.maxPokemonActionFrequencyS"What about the lowest base capture rates?
pkmn %>% select(pokemonId, encounter.baseCaptureRate) %>% arrange(encounter.baseCaptureRate) %>% head(20)## # A tibble: 20 x 2
## pokemonId encounter.baseCaptureRate
## <chr> <dbl>
## 1 RAIKOU 0.0200
## 2 ENTEI 0.0200
## 3 SUICUNE 0.0200
## 4 LUGIA 0.0200
## 5 HO_OH 0.0200
## 6 REGIROCK 0.0200
## 7 REGICE 0.0200
## 8 REGISTEEL 0.0200
## 9 LATIAS 0.0200
## 10 LATIOS 0.0200
## 11 KYOGRE 0.0200
## 12 GROUDON 0.0200
## 13 RAYQUAZA 0.0200
## 14 JIRACHI 0.0200
## 15 DEOXYS 0.0200
## 16 ARTICUNO 0.0300
## 17 ZAPDOS 0.0300
## 18 MOLTRES 0.0300
## 19 VENUSAUR 0.0500
## 20 CHARIZARD 0.0500Or the highest base capture rates?
pkmn %>% select(pokemonId, encounter.baseCaptureRate) %>%
filter(!is.na(encounter.baseCaptureRate)) %>% arrange(desc(encounter.baseCaptureRate)) %>% head(20)## # A tibble: 20 x 2
## pokemonId encounter.baseCaptureRate
## <chr> <dbl>
## 1 RELICANTH 0.900
## 2 MAGIKARP 0.700
## 3 FEEBAS 0.700
## 4 ODDISH 0.600
## 5 CATERPIE 0.500
## 6 WEEDLE 0.500
## 7 PIDGEY 0.500
## 8 RATTATA 0.500
## 9 SPEAROW 0.500
## 10 EKANS 0.500
## 11 SANDSHREW 0.500
## 12 NIDORAN_FEMALE 0.500
## 13 NIDORAN_MALE 0.500
## 14 JIGGLYPUFF 0.500
## 15 ZUBAT 0.500
## 16 VENONAT 0.500
## 17 DIGLETT 0.500
## 18 MEOWTH 0.500
## 19 PSYDUCK 0.500
## 20 MANKEY 0.500Find the base capture rate for a particular Pokémon. Surprised Slakoth isn’t higher - I’ll be surprised the first time I see it jumping out of a Pokéball.
pkmn %>% filter(pokemonId == "SLAKOTH") %>% select(encounter.baseCaptureRate) # 0.4## # A tibble: 1 x 1
## encounter.baseCaptureRate
## <dbl>
## 1 0.400Fury Cutter is a great fast move. Which Pokémon can learn it?
pkmn %>% filter(quickMoves.V1 == "FURY_CUTTER_FAST" | quickMoves.V2 == "FURY_CUTTER_FAST") %>% select(pokemonId)## # A tibble: 9 x 1
## pokemonId
## <chr>
## 1 PARASECT
## 2 FARFETCHD
## 3 SCYTHER
## 4 GLIGAR
## 5 SCIZOR
## 6 SCEPTILE
## 7 NINJASK
## 8 ZANGOOSE
## 9 ARMALDO