Sunday, February 13, 2011

Advanced Pacing

I have a problem with most games these days, and that problem is pacing.

Not that the pacing in the game is bad. Oh, that's a huge problem in indie games, but experienced teams tend to have pretty good pacing in their games.

No, my problem is the pacing outside of the game. When you quit playing and come back. I'll call this "external pacing", just for the moment.

I've talked about this before, but today I'd like to talk about it in more detail.

At first glance, the main problem with external pacing is that you forget everything you were doing. This is a huge problem with complex games such as civ-style games and tactical RPGs.

Today's casual/social games show us one way to deal with it: reduce the amount of complexity in the game. If there's a giant sticker on the game telling you what to do next, you can always go and do it, even if you're gone for a month. However, this solution relies on the game not allowing the player to create anything significant.

In most open-world games, for example, your sandbox play is limited to collecting orbs or other almost zero-state situations. If you're in the middle of collecting orbs when you quit the game, and you forget what you were doing, it hardly matters. Orb-collecting will wait patiently for you to remember that it exists while you do other things. It is just as easy to begin collecting orbs again as it would have been if you had continued collecting orbs without the break. Hell, you could give it to a completely different player, and he would collect orbs just as effectively: there's no real unique state to the orb collection game. None that matters.

This is carried into every facet of the game. If you advance along the main quest line, it doesn't matter if you quit. When you return, the quest picks up with no more than five minutes of lost headway, and it picks up again with a crystal-clear "Do this next!" HUD element.

This is a trait of casual games, too. In fact, you can probably say it is the defining trait of casual games. Nearly all social network games are the same way, but with fun scheduling drama as well.

The point is, this is an extremely common practice.

A common practice I don't like.

If this "low state" method is used, it does allow players to drop and pick up the game as they see fit. But it means that players can't change the state of the game except A) in ways that don't matter or B) along the specific path carved by the designers.

Contrast this to a game of Civilization, where the state of the game is almost entirely determined by the player. If the player leaves and comes back, it's entirely possible he'll forget that he needed to build up Naples some more, or that Paris was about to be attacked. On the other hand, although it's hard to pick it up again after quitting, there's actually some meaningful play. It's not just meaningless zero-state play as you slide along the designers moving sidewalk.

Fundamentally, the more of the game's state I am allowed to modify, the more involving I find the play in a game. A game like Call of Duty doesn't appeal to me much because there's not a whole lot of ways to express myself. You can polish your skills so you kill the enemies better or worse, but in terms of the game world, there's no difference. Whether you just squeak by or whether you ninja them all to death doesn't matter.

I realize that's just my preference. This post is not about whether that's good or bad, it's about games that do want to include a lot more agency. For people who do consider agency important, the problematic fact is that external pacing is killer. Any game with real agency - such as Civilization - suffers dramatically from the breaks the player takes.

Some workarounds have been introduced from one side or the other. Call of Duty and the like shows us one method that works towards agency from an agency-free game: trophies. If you do specific cool things in the game, you earn permanent trophies. I guess it appeals to some people, but I actually find it a net negative. My hatred of the continuous popups telling me how far I am from another pointless trophy is renowned. I have quit games because of it.

Coming from an agency-heavy game to try to make external pacing work, we can look at MMORPGs. Most MMORPGs restrict the agency you have to very specific and numerical axes, while also keeping a strict record of all the things you are trying to do. It doesn't get along well with "open play" where you might just be doing something because it's fun, but it does do pretty well with most other situations.

The record is obviously helpful, since a player can just pull it up and read what the eight things they were doing were. However, the agency axis restriction is also very valuable. By restricting the player to expressing himself through equipment only, and not being allowed to make unique equipment, the MMORPG creates a very clean, mostly linear expression path. Players feel like they are expressing themselves by choosing hat A instead of hat B, but the way they express themselves does not require them to remember anything the next time they log in. Similarly, the skill sets are pretty linear too: you're not likely to screw your build up any worse after a break than if you had kept playing.

Personally, I don't like MMORPGs, either. Sure, a lot of good games have linear paths you follow, but those are normally backed by non-linear, relatively open play in other parts of the game, such as platforming or party customization. MMORPGs simply slap down a million fragments of linear paths to replace both the normal linear path and the normal non-linear play, and suddenly it's agency. I find that uninteresting. Less interesting than a game which is just linear, actually.

So, taking potshots at my personal dislikes aside, I thought about how to make a game with real agency that doesn't suffer from quitting and returning to it.

After some looking around, I decided the key is that you don't have to have all your agency in the same place. You can package it up so that it exists in different domains, each one of which can be reasonably managed when it comes to quitting and returning.

As an example, in Banjo-Kazooie: Nuts & Bolts. You can express yourself by building a vehicle. Once built, the vehicle can then be used in the rest of the game. The vehicle part is not completely integrated: you're likely to finish the vehicle construction as an atomic operation, without quitting and reloading. However, even if you were to quit and return to an unfinished vehicle, it is reasonably easy to remember what you were doing and noncritical even if you do forget.

So the gameplay is partitioned - agency-heavy car building does not mix with exploring town for doodads which does not mix with missions. Each kind of play is reasonably agency-heavy, but not so intertwined that it suffers from quitting and reloading.

I'm not saying Nuts & Bolts is a great game, I'm saying it's an example of what I'm talking about. In the future, I imagine we'll see a game like Nuts & Bolts also have play elements that interact with you via the web, allowing for scheduling gameplay as well.

The idea here is to allow you to build up unique and interesting elements, express yourself as you see fit... and then compile it. Partition it. Instead of remembering everything all the time, you only have to remember a little bit... but your hours of play so far have been very agency-rich, made your world very unique.

There are certain kinds of games I think might be hard to do. Like tactical RPGs where you create two dozen characters. Even if you try to partition the character design into its own unit, you still have two dozen completely unique characters, each with their own specialty, each with their own optimal growth path. Who can even remember which ones you were going to use in the next fight?

It may be we see a descendant of this sort of game that does use play partitions, though. For example, you still make two dozen characters. But instead of being lumped into one monolithic list, you have to split them up into small units of 3 or so, each of which is located elsewhere, with their own concerns. Any given play session might be about any of these teams, allowing you to remember each of them and rediscover their strengths, without feeling like you're discarding 21 other characters.

Anyhow, those are my thoughts. What are yours?


Ellipsis said...

An alternative approach may be to try and provide enough transparency to allow a player to reconstruct his goals and strategies from the current game state, even if the actions that lead to that game state were complex and important. For instance, in a game of chess, every move is important, and the play is very complex, but you don't need to remember the entire history of moves because the information you really need is all present on the board for you to surmise.

Obviously this becomes increasingly difficult with games where the action doesn't all take place on a single board or screen, but you can imagine applying this idea as a kind of standard for a game - ideal external pacing means that a player could pick up a game from any possible state, and if they know what to look at/for, they can figure out relatively quickly what they need to know to take their next step. One of the things that keeps this from happening in a game lik Civ is the fact that all information in the game state is information you might need to know now AND it's distributed throughout all of your cities and units. There's no way to "look at the board" and get an idea of what the current game state looks like.

Craig Perko said...

I agree that perfect information (at least for your side) is a reasonably good way to do it. You do have to be careful not to overwhelm the player: chess can use perfect information because all the pieces always move the same way and interact with the game board in the same way. Chess, as a skill, is mastering those aspects of the game.

A game where you customize your chess pieces to have unique abilities would require you to refamiliarize yourself with what your pieces can do, and remember what the enemy's pieces can do, and then remember how those abilities actually interact with the board.