Lessons in Playtesting

This is a question for Vincent, as well as anyone who has been playing AW since Vincent released the playtest document.

The playtest document itself was a lengthy, very complete text and game.

What changes, what lessons were learned during this playtest phase? How did it affect the final form of the published game?

Were rules changed? Which ones, and why?

Were principles changed? Which ones, and why?

Was the format of the text itself changed, or just updated and clarified?

How did the (semi) open playtest phase change the final product, what lessons were learned?

Big questions! More to answer than I have time for, but I'll try, post by post. Nudge me if it looks like I've forgotten.

Today's:

There were 2 public pre-final documents, essentially 2 playtest documents. There was the playtest document proper, which I released in early 2009 I think, then the preview edition, which I released in early 2010. I consider the preview edition to have been good enough for a 2010 indie game, but I wasn't content with that.

Without doing any calculation, just guessing based on my memories, the playtest document proper accounted for probably, what, 50% of the words in the final book? And the preview edition was about 85% of the final book? Something like that.

Next up: rule & principle changes.

Thanks, Vincent!

I'm really looking forward to hearing more.

How much of that remaining 50% (and 15%) in new words was clarifications, examples, etc, and how much would you say was new material you suddenly realized was needed? (Like, I dunno, the Harm move, for example.)

It's hard to say how much! In a lot of cases, it's even hard for me to tell the difference.

Here's an example. In playtest, there were 2 MC principles: make Apocalypse World seem real, make the characters' lives not boring. As I was writing and filling out the text, I realized that I also needed to explain about playing to find out what happens, so I was looking for a place in the book to include that section. Only THEN did I realize that it would make a perfectly good 3rd agenda.

Same with being a fan of the PCs and maybe another MC principle or two, and turning the PC's move back as an MC move, and threat impulses, and who knows what else. Being a fan of the PCs in particular I would never have thought to say explicitly without some key external playtests where the MC wasn't.

The harm moves, the Hx-for-harm-and-healing moves, how vehicles take harm, those were all new to the final text. There were tons of changes to the basic moves between the playtest and the preview texts, entire moves cut and new moves added, plus changes to the wording of probably every move. I think that there was only one change to the basic moves between the preview and the final. I also revised every playbook at least once. I think the battlebabe went through half a dozen revisions.

Of examples and clarifications, the error-and-correction examples were new to the final, and I think they single-handedly improve the quality of the text. I got the idea for them from David Berg, who made these great little comics about GM-player communication in his game-in-development Delve. (Oh crap, that credit might not be in the ludography. Crap.)

And of course the advanced fuckery chapter was brand new to the final text.

Found David's comics, here.

Found David's comics, here.

That's great.

I love in the "Establishing Setting" comic, how the guy closes his eyes and thinks really hard, and an image of the houses pops into his head.

That discussion about rooftops and pull-ups brings back every painful, pointless game session we had in high school. We would have debates like that for hours. Once, in a Shadowrun game, someone took cover behind a table and the resulting discussion of table position, material thickness, and angles of line-of-sight ground the game to dust.

On-topic: I distinctly remember finding "Be a fan of the PCs" and "turn their move back on them" in the preview edition and being very very happy they were there.

Also, the Gunlugger's secret cache of arms went away, which was also good.

Vincent,

So, when some of us e-mailed you for the password to the original playtest document, to download from the "secret wobsite", was that the playtest or the preview?

Because...

There were tons of changes to the basic moves between the playtest and the preview texts, entire moves cut and new moves added, plus changes to the wording of probably every move.

I'm looking right now at the printout of the Brainer playbook, which I printed from the original playtest document I downloaded from your site back then, and only one move is different from the final version--the seduce/manipulate move. "Go aggro" also has an additional option: "Force their hand or suck it up."

But otherwise things look identical.

So, does that mean my "old" copy was the preview version?

Not that it really matters, I just want to make sure I'm following along with your thought process here.

More importantly, I'm curious how much of the rewording and rewriting (e.g. "I revised every playbook...") was *cosmetic* versus *formal*?

How many of the changes were just questions of better wording (like adding "Force their hand or suck it up" to the going aggro move), and how many were actual, significant, changes in System?

How much playtesting did you do before you released the playtest document?

I guess I'm really curious how close you shot to the mark with the original rules, and how much of the shape of the game, as well as the format for the moves, was formed in your head, as the sculptor imagines the finished work inside the uncarved block of stone, as opposed to bits that you found you had to change in playtesting.

I suppose a better way to ask that question would be:

From start to finish, what parts of Apocalypse World surprised *you*? How much were you writing down the image of the game in your head, and how much were you reacting to unexpected emergent behaviours in the system?

Did you discover/learn anything really unexpected that has changed your view of play and/or game design through this process? What kinds of things?

As always, thank you for sharing! Your internet output is a fantastic resource and a community-builder for all of us.

Other readers: what significant changes did you notice from version to version? How did they affect gameplay for you?

The Savvyhead's workspace rules seem like a big addition from the playtest document I have; maybe not in intent, but in execution.

In the playtest doc, it says this:

The savvyhead gets this one.

Handy: you can repair, rebuild, refill, or reset any busted or used-up thing. You can repair it cheaply, quickly and well, choose 2. (As they say.)

MC, you’re responsible for making it expensive, time-consuming or a poor job, whichever the savvyhead chooses.


Which clearly got a rewrite for the final (and preview) document.

There were also the "toyboxes" and whatnot - and the Angel kit items being individual things for healing particular sections of harm.

Yeah, and the Gunlugger's cache of arms that seemed to point to some kinda ammo system or something that eventually got scrapped.

[ASIDE]

Hmm, that old Saavyhead move (or an updated move like it) would have been very useful in one of the sessions I ran, where the Saavyhead was always wanting to fix or jury-rig things in the field, away from his workspace, and just having him Act Under Fire to do it was deeply unsatisfying.  Guess I can steal the old move as the bones for a new custom move.  Thanks!

"When you want to fix or jury-rig something without being fully prepared with tools and supplies, roll+sharp. On a 10+, you do it quickly, it functions exactly as you'd hoped, and it is sturdy enough to last a while.  On a 7-9, pick 1."

[/ASIDE]

Hmm, that old Saavyhead move (or an updated move like it) would have been very useful in one of the sessions I ran, where the Saavyhead was always wanting to fix or jury-rig things in the field, away from his workspace, and just having him Act Under Fire to do it was deeply unsatisfying.  Guess I can steal the old move as the bones for a new custom move.  Thanks!

"When you want to fix or jury-rig something without being fully prepared with tools and supplies, roll+sharp. On a 10+, you do it quickly, it functions exactly as you'd hoped, and it is sturdy enough to last a while.  On a 7-9, pick 1."

The Savvyhead could theoretically create a "workspace" that they carried with them. Imagine: weird-ass electronica, transmitters & receivers, booby traps. It's perfectly feasible to imagine that as being a demolitions expert with his toolkit.

I'd just create a custom move: "The savvyhead has a workspace of 3 components, yeah? For every component that they don't have with them, I can choose another setback for the project at hand."

There were also the "toyboxes" and whatnot - and the Angel kit items being individual things for healing particular sections of harm.

Hey, what was the deal with these things? Why were they removed? They seemed pretty cool, at least on paper, right?

What was the catch here?

Fast, good, cheap is a great custom move to add to a game. I used it like this:

"When you're executing a complex multi-stage project, and you've got the time, resources, and knowledge, and we don't want to play out the details of it, choose two of Fast, Cheap, and Good. The MC will generate trouble based on your choices."

In the context of our game, the Savvyhead was fortifying the town in anticipation of a full-on assault by the cannibal hordes. He had been given the run of the town by the hardholder, and had drawn up plans for the fortifications, but I figured it wouldn't be Apocalypse World (or real life) if everything went exactly according to plan. So choose 2: Fast, Cheap, Good.

He chose fast and good. So I describe: "Ok, it's going well. The walls are strengthened, the watch-towers are up, and it's all only taken a few days. Joan's gang have been making sure you get everything you need, but to get it done this quick, they haven't been able to be polite about it. A couple of people have been shot, a lot of people have been forced out of their homes, and there's a shitload of bad feelings and grumbling. The threat of being eaten by cannibals is mostly keeping folks in line though."

That's all describing the outcome of the move. Then, genrating trouble, I make a move for one of my threats:

"Pierre, the owner of the brothel in town, comes up to you. Joan's gang have been acting like they own the place, ripping off fixtures, using it as a watering hole, and getting free tricks off Pierre's girls. Pierre is not happy. He says "It stops now, or maybe you wake up very cold one morning, you get me?" Here's the thing - you need Joan's gang onside a heck of a lot more than you need Pierre happy."

Nice, Simon.  In some ways, I like the roll-less move even better.  Joe, I'll have to think about how to apply a workspace to a job that needs doing immediately; I'm still unsure if using a move structured that way makes sense for quick tasks as well as large projects.