The Best Starting Pokémon, According To Science

The Best Starting Pokémon, According To Science
To sign up for our daily newsletter covering the latest news, features and reviews, head HERE. For a running feed of all our stories, follow us on Twitter HERE. Or you can bookmark the Kotaku Australia homepage to visit whenever you need a news fix.

It’s the agonising choice and never-ending debate of a generation: Bulbasaur, Charmander or Squirtle? Picking one of the Kanto starter Pokémon is usually a gut choice, based on past experience or which one looks cooler. But what if there was a way to know, objectively, the best one? Well, now we do, thanks to science.

As mentioned earlier, the WolframAlpha database just added all of the data of more than 600 Pokémon (Generations I through V). Scientific American‘s Kyle Hill decided to comb that database and match up the starter Pokémon against the gym leaders in Pokémon Red/Blue to determine the overall best choice — if by “best” we’re talking about the Pokémon who allows you to beat the game the quickest.

Spoiler alert: It’s Squirtle.

This despite Bulbasaur’s initial statistical advantage (and Charmander, too.) “Bulbasaur might do great at the beginning of the game, it also will have the hardest time of all the Pokémon against the rest of the leaders,” Hill writes. “Squirtle, on the other hand, has simultaneously the most advantages and the least amount of disadvantages against the leaders. Once again, Squirtle comes out on top.”

Here’s the matrix Hill compiled of the three Pokémon (and their evolved forms) and how they rate next to the Pokémon of gym leaders. Hill also decides to ask if Magikarp really is the worst of all the Pokémon. “The short answer is no.” The long answer is at the link.

Squirtle, I (Should) Choose You! Settling a Great Pokémon Debate with Science [Scientific American]

Picture: tloz_herooftime


  • Where on earth did that guy get his information about Squirtle’s base stats? They’re completely wrong. Its stats are in the 40s-60s range, not 20s-80s

    • Wolfram Alpha. I’m not sure about the credibility of the stats but says 40-60 like you’ve said and I’m more inclined to believe them.

      • Yeah, I think they launched it in time for XY. I checked it out last week and was a little surprised at the lack of information (or presence of misinformation) when there are sites like Serebii and Smogon that already have hugely detailed databases.

        I guess it was done as a bit of fun – which is allowed – but I certainly couldn’t recommend it to anyone who actually wants information on base stats and ranges, etc.

        I was also a little bit annoyed that Hill matched up individual types of dual type Pokémon – resulting in matchups being classified as “Super/Not very”. It was clearly written for people that play Pokémon, yet manages to flub one of the core concepts of the game. If the target audience had been people who are not familiar with the game, then I could understand glossing over some of the details in favour of presenting the numbers behind it. It just seems messy.

        Oh well, good on him anyway.

      • Yes, and pokemon fans are right about what they’re talking about. And the thing is, you’re not a pokemon fan, therefore we are vastly more intelligent than you on the subject. :/ I can’t believe you can’t see your own intense stupidity here.

  • if by “best” we’re talking about the Pokémon who allows you to beat the game the quickest.


    Just kidding, competitive pokemon is beyond me.

  • Um, fire based are the hardest to find in the original games, grass & water can be found quite easily.

  • Truth be told – squirtle although looking awesome is not really relevant. Out of all the pokemon there’s by far more water type pokes then any other typing.

    I actually see now that in X/Y mega venusaur (removing 2 of it’s weaknesses with thick fat) followd by Mega Charizard X are the best starters. – Also that speedboost blazekins’ pretty badass but that’s another story…

    • Blaziken is by far the best starter if we’re counting other gens. I’m not sure about what has changed in X/Y but Blaziken has been uber for quite a while now.

      • There was an event one given out in X and Y – It mega evolves and has the ability “Speed Boost” which raises it’s speed every turn.

  • It annoys me that he created a matrix listing the evolutions of each starter, even though the results are the same between the basic Pokemon (eg. Bulbasaur) and the Stage 2 Evolution (eg. Venusaur). Redundant columns are redundant.

  • I’m truly disturbed by how many people here seem to have chosen the two obvious worst choices for their starter, neglecting the raw awesome of Charmander. You know. The motherfucking dino-dragon who BREATHES FIRE.

    Awesome of that magnitude distorts all statistics into meaninglessness.

  • This is a simple answer doesn’t need stats

    One of these pokemon evolves into Ivysaur [the greatest pokemon ever] the other two do not.
    That being said Ivysaur evolves into a really fucking stupid pokemon So I guess that doesn’t help..

    • I still prefer Charizard Y, even if everyone laughs at me for it. Insane Fire damage is just too much fun – that and Charizard with an instant Solar Beam is hilarious.

  • My special attack EV’D mega charizard Y begs to differ. Not only can it breathe fire, it’s ability drought adds to my STAB flame attacks. I am tossing up throwing overheat in to replace inferno for extra awesome

  • For an article that claims to be lathered with science, that linked article is such short-sighted look at things. Apart from the fact that they’re mixing and matching partially-incorrect info (using stats that are not only wrong, but taken from Gen 2 (special split) and applied to gen 1.), they seem to be cherry-picking.

    Firstly, resistances are ignored in the type-matchups. Unless you’re one-hitting everything, resistances are important: eg. Water gets slammed by Lt Surge, Grass resists, etc. EDIT: they also neglected secondary typing, which is relevant mostly for Charizard, although Venusaur has access to Sludge Bomb in Gen 3 (if we’re talking FR/LG)

    Secondly, looking at the stage-to-stage gains from evolution are totally irrelevant (even if they DID use the right stats). What matters there is the raw stat comparison between each pokemon of their respective stage. eg. stat differences between Blastoise, Charizard and Venusaur.

    The whole Magikarp argument? Two major problems there. Firstly, BST (Base Stat Total) on its own is totally useless. Most importantly, Physical attackers will never use S.Atk and vice-versa. Eg. Ralts is “second-worst”, but has decent S.Atk and low Atk. On the other hand, Magikarp is really fast, but has useless Atk (figuratively speaking, 10 is utter trash) and useless SAtk (literally – no Special moves in sight).

    That brings up the second issue here – movepools. Magikarp has 4 moves in the later Gens, 2 in the first. One is the token functionless move Splash. The other is the 35 BP (Base Power) un-STABed (Same Type Attack Bonus) Tackle. On the other hand, Ralts’ movepool is gargantuan, including Confusion at Lv6 – a 50 BP STAB move, running of Base 45 S.Atk. Before considering stats, Confusion has more than double the raw power of Tackle in this case. After stats, it’s somewhere in the order of 5.5x. Similar case exists for Sunkern)

    I could go on and on about the number of validity issue here, but I won’t.

    People performing “science” need to know what they’re talking about, first. Sounds to me like they just got their hands on a big batch of numbers and wanted to try and prove something with them. Also, they speak as though this is the first time the numbers have been available. They’ve been around for YEARS, and from what I can tell, the numbers they have aren’t even all right. What they’re trying to do is called Theorymon-ing. And they’re doing it wrong.

Show more comments

Comments are closed.

Log in to comment on this story!