I am still playing Magic the Gathering in the form of Magic Duels on iOS every day. Today a new expansion, Aether Revolt, came out and I'm looking forward to playing with the new cards. While I spent some money on Magic Duels at the start, my daily playing gives me enough gold that I can buy a complete new expansion the day it comes out without having to spend any real money. So far, so good.

Like most other games, in Magic Duels I mostly play PvE against the artificial intelligence. I don't know how many people play Magic Duels on the iOS, but if you choose a PvP game you frequently wait a long time before you find an opponent, so long it sometimes even times out. And then you run into the usual problems of PvP, with people throwing games as soon as they run into the first difficulties. So while theoretically a PvP game could give you a much bigger challenge than a PvE game, in practice it frequently doesn't.

So where is the fun in playing against moderately challenging AI opponents? Well, for me Magic the Gathering was never about building the unbeatable deck and refining it to absolute perfection. There are a lot of people on the tournament circuit out there who do that. The result is always the same, an environment in which only a small handful of decks is viable. Many other deck ideas are more fun, but less efficient, and thus get weeded out. By playing against a moderate player in the form of the Magic Duels AI, the fun decks become viable options. So I can amuse myself all day building crazy decks and trying them out.

In the basic set of the early editions of Magic the Gathering there was a creature card called Grizzly Bears. It never was a particularly good card, costing 2 mana for a 2/2 vanilla creature. But it resulted in an interesting thought experiment: What if your opponent was playing a deck consisting of 20 forests and 40 grizzly bears? (Not a tournament legal deck by any standards, it's a thought experiment). What that opponent would play every turn and how he would attack is rather predictable, so you know you would face the first grizzly bear in turn 2, another in turn 3, possibly two more in round 4, etc. The argument of the thought experiment was that if the fanciful combo deck that you just imagined can't beat this imaginary Grizzly Bears deck, you should throw your deck away and try a different idea. The Grizzly Bears deck isn't a real opponent, but it provides a minimum challenge that your deck must be able to beat in order to be not completely ridiculous.

For me the AI decks in Magic Duels are somewhat improved Grizzly Bears decks. They usually are all about playing more and more, bigger and bigger creatures. There is sometimes a bit of removal, but never any mass removal that would completely change the environment. And there are sometimes cute tricks with enchantments and artifacts, but never devastating combos that one-shot kill you. They are far from the decks a tournament player would play, but they provide a good challenge between easy and medium (depending on the difficulty level you choose). And that gives me the perfect environment to try out fun decks.

By the way, even a tournament deck can lose a game against a Grizzly Bears deck, due to the inherent randomness of Magic the Gathering. If you draw only lands, or not enough lands, or lands that don't fit the color of your spells, or any other combination of cards that just don't fit together, even a mediocre AI with a mediocre deck will win the game. In Magic Duels I have some doubts about the random numbers generator, which appears to produce clumping far more often than statistical probability would predict. In any case, a deck I build as a fun idea is far from being an automatic win, even if I win far more than half of my games. There is never "no challenge at all".