Loops and Progression

An intriguing tweet from Dan Cook led me to checking out a game design report named "Applications of Cyclical Progression", which I really liked, although I have one reaction.

The doc focuses on the design trope of "linear progression": for example, many games from Candy Crush to RPGs have been structured as an ever-growing content sequence (matches, levels, story, etc). Players go through the sequence of challenges, experience the story, go through the ups and downs of the story arc - and in the process collect more wealth, build up the character, and so on, on their path to victory.

It's a good trope, but the problem is: if players like it, they will play through it, get through all the content, and then what? Assuming the developer wants to keep this going, they're faced with a content cadence problem: they need to make more content, and good content, at a pace that's fast enough to keep player happy, with a good quality so it's not boring or repetitive. But because the progression is linear, the player has already leveled up pretty far, gained a ton of wealth, stats and so on. So what can we do for a player who's super advanced and has everything? Do we have to keep escalating abilities and challenges to ridiculous proportions, like in Star Trek V where captain Kirk went to the center of the galaxy, fought with God, and won?

So that's the problem description. The report then digs deep into one particular family of solutions: converting linear progression into looping progression. For example, compare Diablo to chess. In Diablo, there is a main story line, and the game keeps changing and evolving, as the different challenges and areas push the player through the story. But in chess, after each match, the player returns back to the same game board! They play against other players, better players, different players, but the game itself doesn't change. Over time they may improve their skills, and their experience may change and get deeper, but it doesn't come from experiencing more content - rather, from experiencing the same game, over and over, but differently.

On this point, I'm very much in agreement with the authors of this doc. I'm a fan of games that you replay over and over, as you learn and get better (simulation, strategy, builders, etc), and I think this whole approach makes a lot of sense. It's a different kind of an experience, a different kind of enjoyment, when we come back to similar situations and find new depth and new challenges in them, as we ourselves get better.

However, if you'd allow me one nitpick about the report: in presenting a laundry list of challenges and cyclical solutions, I think the authors missed out on an opportunity to organize this knowledge. It doesn't take away from the report, but would make it easier for the reader to apply it.

Specifically, I think cyclical progression is sensitive to design decisions along at least two axes:

  • Is the game single-player or multiplayer? and
  • Do we want to add cycles in form of metagame loops or gameplay loops?

Here's what I mean specifically:

  • A multiplayer game is going to be inherently easier to cyclicalize. We can give two people a ball to kick around and some rules, and voila! We have a fun cyclical game. :) In this case, we don't have to do much work to make the game cyclical - but we do have to work on making it fun for players are losing and therefore aren't having fun.
  • A single-player game is more about individual achievement and building things in the game. This kind of a game is going to be harder to cyclicalize, because it means asking the player to start over, so there need to be reasons and rewards for doing that. But on the plus side, these cycles feel much more inherently interesting, because it's the player's own limitations that hold them back, and overcoming them is fun. It's like playing chess against a perfectly matched opponent.

Then the second axis is that of the type of loops:

  • Metagame is the kind of decision making that happens outside of game sessions: after you played a hand of cards or a single game of chess, you go back and get more cards and reconsider your deck, or maybe go study some openings and standard tactics, and so on. Sports (and sports spectatorship) are also full of metagame elements - what happens on the field is already gripping, and then what happens in between games is extra interesting. This is where mechanics like leaderboards, competitions, team- or deck-building, etc. can form the basis of very appealing, long-term cyclical gameplay.
  • Gameplay loops are the kind of decisions at different frequencies that people do when playing a single game session. For example, in The Sims, we have to decide every few seconds what each family member will do next; every few minutes what friends to invite over; every tens of minutes on how to redecorate the house; and maybe every hour or so whether to add new family members into the mix, and completely shake everything up. When games layer these loops together, it generates the famous "one more turn" feel, where the player wants to keep playing, because they have a ton of irons in the fire and they want to see everything through to the end, which will never come because they keep putting even more irons in.

So depending on where we land on this 2D matrix, I think the approaches could be very different:

  • Multiplayer metagame: such as boosting excitement and planning around leaderboards, matches, improving one's odds from one match to the next, and so on (e.g. via card trading and deck building in CCGs)
  • Muitiplayer gameplay: such as adding more long-term strategic elements and layering them on micro gameplay (e.g. macro strategy layering on difficult micro play in competitive RTSs)
  • Single-player gameplay: such as adding low-frequency and very-low-frequency decisions that may require the player to reconsider and rebuilt huge chunks of what they've already built (e.g. rare technologies in Civilization which open up entirely new types of military units)
  • Single-player metagame: such as boosting excitement around single-player achievements and sharing them with other players (e.g. "share your house" feature in The Sims)

I do think that, depending on whether our game is 1p or mp, and based on whether it's composed of short or long sessions, the strategies for cyclicizing a game will look very different.

But regardless of this, that report is very detailed, and very much worth checking out. And in the end, I think the focus on cyclical design is completely right on. As players, once we find something we like, we want to keep playing. And as designers, we should support this!

@rzubek
Saturday, January 11, 2020