Tuesday, November 24, 2015

The Problems Compound postmortem, part 2: testing

I'd like to talk about testing, first, both in general and specific to The Problems Compound.

But first, my previous entry is a hopefully humorous pop quiz, if you followed from planet-if.

From my perspective, it's hard to ask testers. If I've asked them before, there's always the "oh man, I'm asking them for another tricky thing to test" angle, but if I've never asked, I wonder what I'm getting them into. It's--well, it's hard to ask, and this actually hits on themes I didn't develop enough in my game. About finding it hard to ask, or figuring you had or were something pretty smart but no way to develop it.

So I'm grateful for the efforts my testers put in. Especially given the things I wanted to address in the game. So, mentioning everyone in no particular order, I'd like to bring up my work with Juhana Leinonen on BitBucket as something that went really well & which I'll discuss later in a post on organization, as well as September-ish interactions with Brian Rushton and Hugo Labrande, who helped despite preparing games of their own. And Wade Clarke, Marco Innocenti and Matt Weiner came back for more punishment even though they know what testing a game of mine can mean: unimplemented stuff, odd bugs, and things that should connect but don't.

I tried to thank testers individually for the big things they inspired and, in a few cases, should have inspired if I had started earlier. In the first week of the competition, I was mostly happy with my plans for post-comp release. I knew it was missing something, but I thought back to how I could've made it a bit more fun for my testers. And thinking this instead of "HOW WILL PEOPLE RECEIVE THIS" is a big step up. Specifics are at the end. But I want to discuss testing in general and what it can mean and how it can be more than just a chore. After having so many testers help me, there are a lot of themes.

First, I found that testing a game helps me get back in the creative flow. I've been spitting stuff out pretty regularly, and it, well, requires effort on my part, and on the part of people who take the time to look at my works. If I'm testing, I feel a bit less hesitant asking for something in turn. That's not to say reciprocity's forced or should be. When people are helping, it should be because they want to. But I do want to have something resembling balance. And I think this may be a tip for a new author who feels like they don't fit in: offer to help on a game. See what happens. See what questions you have. You may learn something, and the best things are what you didn't expect to. Maybe you will even know something they don't, which can help them.

I'm pleased with having tested 7 IFComp games, which matches up with the 7 testers I had. I had other people to consult with, and they helped me get on the right track. I hope I can continue to test even when I run out of ideas on my own, or my ideas aren't ready. And I am also pleased with the balance of new and old testers I had. This isn't just a guess at things, that they should work this way. This paper looks at Broadway musicals and the connectedness of the authors vs how well they did. The best ones had connectedness in the middle. Roughly speaking, it seems you need people used to your own quirks, who will be able to intuit what you mean, and people not used to them, so they can say "hey, most people won't do this." Of course, both types of people can still say "you need to change this" or "I'd like to see this, too," but a balance of familiarity and newness helps. I sort of wish I'd dug a bit deeper to get someone who didn't know Inform, but those who did gave a lot of "someone might not understand this..." and they immediately understood some bugs were just 1-line fixes. Which helped them move on. Which was very good.

And that leads to another balance. First, you want to point out what's concretely wrong. Second, you want to point out what could be better. Of course, when you bring the red pen, you also want to say, "yes, this worked for me" in lots of places. So there's a lot of trust here that a tester won't say GEEZ WHAT IS THIS, which is incredibly demoralizing, and my testers had the chance to do this. At some points, I can imagine they wanted to go all caps, but they were being diplomatic without making me feel there was an "I'm being diplomatic" context.

So I'd like to point you to Richard Thaler and Cass Sunstein's Nudge, which, regardless of its political implications, has been critical for me. Or at least my interpretation of it is. We can and should look for small suggestions that make big changes. How do we find small suggestions that help make big changes? It's not trivial, and it's more than just sniping at typos, which is also important, especially since a tester who finds those gives me something to warm up with--something I know I can fix--before hitting probably more subjective things. But it's also not guaranteed you'll find something big that emerges. So I like to encourage my testers to say "Hey! Wait! I know what was bugging me!" Not that I want stuff to be bugging them. Just...I've had the penny drop about others' games, and I'd like to leave the door open.

This also ties in to the theme of my game, in that ... well ... most of the antagonists, if they tested my game and sent a transcript, I would have to put it in a corner and go work out to burn off my frustrations. No matter how right they may have been--and there are a lot of LOOK AT ME I'M RIGHT types. And I bet they would have found things! There's a balance between being right and being helpful in testing, and even when it's achieved, you wonder if a tester you're good friends with is just being diplomatic. It's tough to balance encouragement and saying, whoah, this needs to be fixed. There's also a certain sacrifice that goes into testing--you may have a good organizing idea in general, and you wish you had it for your game, but--the other guy first.

But overall I appreciate the positivity, and I hope I do so without sounding like a self help book. Sometimes we need to provide general support, and I got that, but I also got pointers to what to fix. People saying, as I see it, this is the weakest link. And perhaps the most rewarding thing was, whether these people needed to hold their tongue and be diplomatic or not, this helped me. And it ties into the game: the nastier people in the Compound are "trying" to "help" Alec, but their help is rubbish. Looking ahead to the post about the game: I encountered a lot of this as a teen and was unable to deal with it, and it poisoned the sort of general encouragement that I did get. Getting the general encouragement as well as reasonable goals to fix undid that. But I took time emotionally to go back through that. Time I just barely had. So this is a bit meta.

Perhaps another way to look at it is from a sports fan's perspective. The testers act as defense. But they can also act an offensive line to open running lanes or give you time to throw a long bomb so you aren't blindsided by bugs. They may be the midfielder providing that cross for a free header, or the point guard who penetrates and dishes to you for a wide-open shot. Or they may be the catcher telling you what pitch to throw, or even the hockey player who passes the puck not to you, but to where you need to be (I sort of stole this from a famous Wayne Gretzky quite.) I've found visualizing that helps me give advice and also take advice without it feeling too technical, or like a self help book.

I'm sure we've heard "IT'S NOT WHERE YOU ARE BUT WHERE YOU'RE GOING" which is sort of annoying, but being able to frame that in terms of something I want to do, or something a friend wants to do, has helped a lot.

Fortunately, I got a lot of that from my testers, to help with where I was going despite a mess of bugs. The scoring method was implemented late. I figured it wouldn't be hard to do, but I left in an infinite-point bug which was a one-line fix. Other things included asking about the bad guy, which felt like the sort of tap-in you can do on the last day. I even tested talking to everyone! Well, except the jerks.

But other things went well. Testers suggested things to say, or how to describe Fritz, and even gave me ideas for different ways to get ticketies. I "upgraded" Sly Moore and his puzzle based on hearing from my testers, and I know I have brain fades where I describe too little rather than too much. So having people check off on that as well as other things helps. My testers had to suffer through oddness with Meal Square being open, then not, then--well, it was tricky. Even the cookie ending was busted for a bit! And sometimes I just realized something needed to change based on seeing transcripts where people got stuck a few times--or even when no tester said "I particularly like this bit." And they all did find nice things to say, and that added up to me being able to find weak spots that didn't have anything nice enough--yet.

And the underground area...well, I think there's still stuff to shake out there, the hows and whys. Objectively, it was a bad idea to put it in there. I didn't write a full case-checker until October 2nd after David Welbourn pointed out something had alternate interpretations. I didn't describe fully WHY the underground area should be there. But testers were good sports about it and they helped me. It always stinks to have to say "Your banging your head against the wall helped. Thanks for doing so. It should have been much easier." So I am grateful for their time and thought.

No comments:

Post a Comment