Somebody Save Me

Chris over at Only a Game published an interesting perspective on save games. He wrote a hypothetical dialogue between a game producer and his engineering, art, and QA teams on what type of save game system they should include on their game.

I think he raises some interesting points, but I think he misses the mark on some others. So I am responding with a dialogue of my own that I think more accurately captures what’s at stake. One of Chris’s core assumptions is that only “hardcore” players care about being able to save anywhere. I disagree with this assumption. In fact, in my experience, it’s the exact opposite. Hardcore players have been trained to put up with all sorts of stupid behavior from their games. The people who want to save anywhere are the casual gamers. Here’s a counterexample:

WIFE: OK, sweetie, we’re late for Thanksgiving dinner. We have to go now.

HUSBAND: Sure thing, hon, I’ll be right there. I just have to get to the next save point. If I quit now, I’ll lose the past 20 minutes I’ve played and have to do it again.

WIFE: How long will it be until you reach the next save point?

HUSBAND: Well, usually the save points are spaced about 15 minutes apart. Of course, sometimes they’re 30 seconds apart. And then sometimes you’ll have a deathmarch where you have to trudge through the lava level for an hour.

WIFE: An hour?

HUSBAND: But that’s unusual. It should just be a few minutes. Well, assuming I don’t accidentally walk past it and miss it.

WIFE: Why can’t you just save the game?

HUSBAND: Well, the game designers worry that they might alienate part of their audience with complex concepts like saving the game.


HUSBAND: Why are you looking at me that way?

WIFE: My grandmother, who is 80 years old, knows that she has to hit “save” before quitting her word processor. There are people who seriously think that saving a game is a hard concept?

HUSBAND: Well, it’s a very complex issue, sweetheart.

WIFE: And these are the people who are making me late for Thanksgiving dinner?


HUSBAND: Be right there, kids! I just have to get past the Plains of Pandemonium and defeat…

[WIFE shuts off console.]

HUSBAND: …Hubris.

It’s not that I think Chris is a bad guy. He’s looking at this from a software developer’s perspective and discussing the tradeoffs. I’m looking at this from a consumer perspective. The ability to walk away from a game at (nearly) any moment, to save at roughly any time and return later without substantial penalty, is of paramount importance. It is, if you will, part of the production values of a game. I believe that a commercial game that does not provide that ability is – as far as this aspect of production is concerned – unacceptably primitive, and unprofessional.

In other words, I’m not inclined to be generous about this. Traditional console game attitudes towards save points are antediluvian, wrong, and broken, and the entire idiom of the “save point” must be destroyed. If you want to save some money to create an unprofessional game, go ahead. But I think we should be creating the sort of consumer environment where people who try to sell games like that are mocked, just like people who produce commercial movies with poor production values are mocked.

Chris raises some objections to “save anywhere” that I think are somewhat tangential to the problem. First off, let’s recognize that by “save anywhere” we don’t really mean “store 100% of the game state at absolutely any time, including when I’m about to take a bullet to the head.” Rather, it is a fuzzier definition that amounts to “allow me to save frequently enough that if I have to get up and do something more important, I can do so after no more than at most a minute.” Or, more succintly, “allow me to save often enough that I don’t feel compelled to hop on a flight to LA and hook the game’s producer up to the face-slapping machine.” I don’t care (and won’t complain) if the game had to reset some bits of state to ensure it wasn’t in an unwinnable state, as long as it doesn’t require me to replay the last 20 minutes of gameplay, which, let’s be frank, probably weren’t all that much fun the first time through anyway.

Second, there’s cost of implementation. On the one hand, Chris talks about how save anywhere creates more work for QA, but on the other hand he agrees that a “let me stop playing now” checkpoint save is pretty much a requirement for reasonable games. Don’t such saves present approximately the same QA load?

Lastly, Chris worries that the user interface aspects of managing many save games might be too confusing for users to handle. He’s right. However, this is completely orthogonal to the issue of whether such UIs are on top of “save anywhere” or “save at save point” systems. Developers can (and have) created save point systems that are just as confusing as save anywhere systems. User interface is an important aspect of any game, but it shouldn’t be used as an excuse to not do the right thing.

User interfaces are funny things. There are the patent aspects of user interfaces (this is a button, these are checkboxes, this is how I navigate menus), and there are latent aspects. Books, for example, have a fabulous save interface: put the book down, and then when you want to start reading it again, pick it up, find the page, and read. DVDs, likewise, use chapter markers and fast forward to allow the viewer to pick up where she left off. No matter how transparent a checkpoint save you provide, you still run the risk of providing a simple patent UI, but a horrible latent UI: every gamer, no matter what their level of sophistication, is capable of forming the thought “Why is the game making me play this? I already finished this part.”

And when the players have that thought, it’s the game designers, and their reputations, who are stuck in an unwinnable state.