Block Broker

Tetris 99, but you can bid for pieces you need and sell ones you don’t. “My kingdom for a long piece!”

So a while back I made this little experimental library called BazaarBot which I wrote about on Gamasutra. The basic idea was to create an “open source economics” engine, which is to say a glorified auction house API. It turned out to be a lot trickier than I expected to turn it towards any useful kind of end, and the research paper I based it off of has attracted some criticism. I refactored BazaarBot in response, but have never gotten around to making anything really interesting with it.

A big problem with making an econ game is what I call “The Vanilla Ice Cream” problem. The econ sim is a sort of system that requires an underlying sim where various agents are actively doing something and using and consuming resources before it makes any kind of sense to drop a market system on top. And then, you wind up spending so much time making little goblins or whatever wander around a forest and hunt pigs and chop down trees and build houses that you never get around to building your econ sim, which is all you wanted to do in the first place! What you’re really interested in is the chocolate syrup and the nuts and sprinkles, but you spend all your time just trying to churn up the boring Vanilla ice cream instead.

“Block Broker” was an attempt to find the simplest and stupidest “Vanilla Ice Cream” to layer a market system on top of. I was on a walk with my wife one day and she suggested Tetris.

It could work! Maybe!

So, you get a bunch of people playing Tetris simultaneously. Tetris 99 is a good example, but this would probably be a little less directly competitive for some reasons that will become clear. Also, the other players don’t have to be human, they could just as easily be computers, or a mix of AI and human. In any case, you’re playing Tetris, and so are a bunch of other people, and you’re all connected in a single game session somehow.

The Tetris company has standardized modern Tetris with a few key features – the two most salient for this article being the “Hold piece” mechanic that lets you save one piece for later, and the [“7-bag” random generator] which decides which piece comes next. As the name implies, every sequence of seven pieces is guaranteed to contain one of each of the seven possible Tetromnioes in the game, shuffled in random order within each “7-bag”. This means that the longest time you will ever go waiting on a long piece is 12 pieces: when it comes first in one bag, and last in the next one.

So in Block Broker, we expand on the “Hold piece” mechanic. Your “Hold piece” box becomes your “Sell piece” box, and you can adjust the price you want to sell it for. The default price is “Not for sale”, but flicking e.g. the left analog stick up and down (you use the DPAD for normal play) will allow you to adjust your asking price up or down. In addition to the “Hold piece” box, you have a “Want piece” box. You can switch what piece you want by flicking the right analog stick left and right, and you can adjust the price you’re willing to pay by flicking it up and down (or maybe use L & R to switch pieces, I dunno – controls will be essential for implementation but I won’t settle that issue now).

So you can offer your hold piece for sale, and you can ask for a certain piece from the market, and you can set a price on each. What’s the currency? We could either go with score or lines. Probably score, because scoring a Tetris should be worth more than just 4x the currency reward of a single line.

Offers are resolved in real time. When you adjust the price on your ask/sell boxes, a short timer will time out and then make those prices live. (This way you can kick a price up or down without it resolving instantly halfway to your target price). A BazaarBot-like auction house will then resolve orders – if the price you’re willing to pay for a given piece is equal to or higher than the selling price someone is offering, then their sell box is cleared, the money is transferred, and that piece will be the next one that drops for you. I’m not sure how you resolve a buy price that matches to a lower sell price – I’d start by just giving the seller the price the buyer offered, even though it’s higher than what they asked for. Other alternatives are to match the seller’s lower price (but give the highest price priority in matching), or to clear at the average of the buy & the sell price. At any rate, offering to pay more to buy should reliably get you your pieces faster than those offering to pay less, and offering to sell for less should reliably clear their inventory faster than those offering to sell for more.

Okay, so we’ve got multiplayer Tetris with buying and selling of pieces. What is this going to look like and what else needs to change?

Managing attention is the first thing that has to be handled. Tetris 99 is hard enough to play when you’re just trying to do the Tetris, managing the strategy of who to target is a whole nother level, and most people just stick to flicking the analog stick to one of the four AI targetting modes. For this reason I think this sort of game would work best when you can’t directly compete with the other players by sending garbage and trying to knock eachother out. You’ll also probalby need to add some “breathing moments” where the player has time to pay attention to the market and adjust their buy/sell orders. Natural moments might be when you make a Tetris, or perhaps there’s a regular pause every tenth drop or something for every player. If you do put competitive garbage lines in, perhaps that comes with a momentary pause.

The next this is the 7-bag randomizer might need to go. The purpose of that mechanic is to smooth out the randomization, because with a perfect RNG it’s entirely possible you could never get a line piece, and any other number of pathological cases like a ridiculously long string of nothing but S or Z pieces. It also ensures that given sufficient dexterity and skill, you can continue playing forever. With a truly random bag that makes no guarantees you could be forced into a game over regardless of skill. But that’s all because in standard Tetris, you have literally no recourse to getting pieces from anywhere else. In Block Broker, we have the market as a way to get pieces that you need.

And this is where things get interesting.

So What?

We can now create all number of interesting and/or pathological simulations. We could create a game where 99 players compete with one another, all get a mostly reasonable mix of pieces, and just see what happens when they can also trade. Or, we can take that same scenario, but each player only ever gets one kind of piece – an economy of narrow specialists – and they have to trade to do much of anything. And do some lucky jerks get nothing but line pieces all day, which they can either stack together for cheap Tetrises, or sell on the market for massive profits? This is getting kind of darkly educational all of a sudden.

The varieties are endless – what does it look like when you’re sending garbage back in forth in standard competitive Tetris fashion, versus when you’re just competing individually to get the best score and can’t hurt each other directly?

What’s it like with the 7-bag system, does that reduce the need to trade? What’s it like if we simulate “weather” conditions, and certain people get a “drought” or “bumper crop” of line and l pieces? What if someone gets a “plague” not of locusts or weevils, but of squares and S/Z pieces?

I see “Block Broker” less as a specific variant of Tetris and more as a platform for economic and psychological experiments with Tetris as a basis.