Design Lessons from a Few Weeks of MtG:Arena

Well, I ended up taking last week off for a move, and now it’s almost the end of the week after that, so let’s try and get something in shall we? I’ve been playing a fair bit of MtG:Arena, so I wanted to jot down some of what I’ve learned as a player, and make some associated notes for how those lessons can be applied to broader game design

Let’s start broad - I’ve always said that MtG is essentially a game design sandbox.  A couple of specific goals, but myriad tools that can be used together in endless combinations to expand those goals and to create different types of engines for achieving them.  And I’ve come to realize that deckbuilding closely mirrors which rules go into a design.  As a youth, I had a very limited cardpool, and had seen very few games in play, so I knew very little about momentum, control, or engine building.  The deck du jour at the time was the premade Sliver deck and it showed off all that Magic and its colour pie were capable of - but all *I* took from it was the tribal aspect.  I liked creatures and buffing them up and trying to give them trample.  I built red/green decks with a lot of sorceries since I didn’t much understand instants / the stack, and I didn’t put enough lands in because I didn’t actually know that gamestores typically give you those for free.  I had no sense of resource management, nor did I take into consideration the decks of my fellow players, their meta, or how to manage it.  Around that time was also when I was starting to think about game design, and while I did have a nascent sense of Experience First design, there was no Second - I knew the experience I wanted to achieve and had no knowledge of mechanics or how to go about creating the experience, save through cheap, unbalanced imitation.

Now I’m older and - well, just older.  But I’ve seen more examples of things working and not working, and I pay more attention to the why and how of things.  And the decks that I’ve had the most success with, and the least success against, tend to be fairly short decklists, as they’re not about whittling down a surplus of options, but identifying some small but broad core synergies, and carefully adding one option (up to 4x) at a time and checking to see if the core system still runs optimally.  A deck with no singletons is like a game design with no edge cases or exceptions - it’s elegant, and will allow fewer interruptions to the flow of learning and playing the game, and encourages the emergent play in when you take an action, not what action to take.

I also used to always think that mulliganing was bad - it was never worth one fewer option to draw a new hand.  Maybe I had that mixed up with the Monty Hall problem.  But while I now accept a starting hand that doesn’t have the exact foundation for my engine that I might desire, I do have dealbreakers in place.  If there are fewer than two lands, and no cards with a cost of less than four, then I cannot bet on drawing another land or a cheaper card.  I cannot risk that loss of early game momentum.  I literally won’t be playing the game on an even footing, and that’s the only game I have a chance of being competitive in, and the only game I can win.  This reinforces for me something I’ve seen a few times now on game design Twitter - if everyone (who wins) is taking the same three actions, then those three actions should be folded into setup.  And it’s not worth your while to design for low-probability events, when you can get both players in non-or-less-luck-based control of their strategy at the same time.

Alright, that’s probably enough for now, though I should do another one of these about the types of match or event you can play in Arena, and maybe one about how meta-based counter strategies can help your design stand out.  Hopefully I’m able to get one of my foretold Out of Manas written tomorrow for Mechanic Monday.  We’ll see! Stay safe, protect trans kids!