Call for Playtesters: GURPS Low-Tech

[umbros]

I understand what you are saying, I was on the Martial Arts playtest and the volume of posts was almost overwhelming. I imagine that if you, as an author, had to read and to respond to all of those you wouldn't want the threads to get out of hand or duplicate themselves. Not if you wanted to have a life away from the screen anyway.

[Kromm]

For the curious, here are the final closing volumes for several playtests of important core books with which I was involved:

• GURPS Basic Set: 3,292 posts
• GURPS Bio-Tech: 2,332 posts
• GURPS Martial Arts: 3,349 posts
• GURPS Powers: 2,372 posts
• GURPS Space: 1,659 posts
• GURPS Thaumatology: 4,720 posts (!)

These are good to within about ±5%. The error is a consequence of the occasional double-post or accidentally deleted item. If we take the average above, we get 2,954 posts. Volume is definitely linear in list membership, so a 50% increase in list size naively means +1,477 posts to read. It's easy to appreciate how that could cut back on efficiency!

[Turhan's Bey Company]

Originally Posted by Kromm

For the curious, here are the final closing volumes for several playtests of important core books with which I was involved:

• GURPS Basic Set: 3,292 posts
• GURPS Bio-Tech: 2,332 posts
• GURPS Martial Arts: 3,349 posts
• GURPS Powers: 2,372 posts
• GURPS Space: 1,659 posts
• GURPS Thaumatology: 4,720 posts (!)

<LLoydBridges movie=Airplane>
Looks like I picked the wrong week to stop smoking.
</LLoydBridges>

[Kuroshima]

Originally Posted by Kromm

For the curious, here are the final closing volumes for several playtests of important core books with which I was involved:

• GURPS Basic Set: 3,292 posts
• GURPS Bio-Tech: 2,332 posts
• GURPS Martial Arts: 3,349 posts
• GURPS Powers: 2,372 posts
• GURPS Space: 1,659 posts
• GURPS Thaumatology: 4,720 posts (!)

These are good to within about ±5%. The error is a consequence of the occasional double-post or accidentally deleted item. If we take the average above, we get 2,954 posts. Volume is definitely linear in list membership, so a 50% increase in list size naively means +1,477 posts to read. It's easy to appreciate how that could cut back on efficiency!

Yeah, the MA playtest was hectic, but the THM one was real nuts, and Phil Master's health problems made it even worse. In the end, the book ended up being top notch though.

BTW, are you still selecting playtest applicants as you were? meaning, that, after selecting those on basis of merits, you pick a number of them based on how fast they applied, and then randomize the rest?

[whswhs]

Originally Posted by Kromm

Volume is definitely linear in list membership, so a 50% increase in list size naively means +1,477 posts to read.

That's interesting to know. It provides a good baseline for modeling cost/benefit tradeoffs.

On the other hand, the actual cost of dealing with posts may not be linear in the number of costs. At least one component is going to be sorting out which posts relate to which other posts—and that could go as N(N-1)/2, or roughly quadratically. Thread tracking takes it back toward linearity, perhaps, but between thread breaks and threads that cross-relate to points in other threads, I don't think it gets all the way there.

Bill Stoddard

[whswhs]

Originally Posted by Kuroshima

BTW, are you still selecting playtest applicants as you were? meaning, that, after selecting those on basis of merits, you pick a number of them based on how fast they applied, and then randomize the rest?

Not this time. We did pick about 50% of the regular playtesters on obvious merit of several types. But for the rest, we didn't randomize, nor did we pay any attention to speed of application, as long as they were in the main clump that came in over the first few days. Rather, we looked at subtle quanta of merit or desert, including "never playtested before," and often somewhat subjective. Then we negotiated about a few of the marginal choices, and moved one or two people from No to Yes, with the corollary of moving one or two people from Yes to No to keep the numbers stable. For example, I requested that one person be included because they were knowledgeable about the subject of one of my chapters that not many playtesters emphasized in their applications. . . .

Bill Stoddard

[tg_ambro]

Good Stuff to know, I am hopeful of joining an interesting playtest in the future. I have an idea for a book proposal (as I'm sure many do), so working in a playtest will help me understand the process a little better.

[pyratejohn]

Originally Posted by Kromm

If we take the average above, we get 2,954 posts.

In how many days time? Two weeks?

[Kromm]

Originally Posted by whswhs

But for the rest, we didn't randomize

The list you worked from actually did include a randomized cohort: I used Excel to pick names at random. Sorry if I didn't make that clear!

[Kromm]

Originally Posted by pyratejohn

In how many days time? Two weeks?

Playtests for large books typically run for about a month – and that's a generous definition of "month," closer to 30 weekdays or six calendar weeks. Which means that a 50% increase in list membership translates as roughly an extra 50 posts to read each workday. Even if they all took a minute to scan, that's almost an extra hour in the day . . . and most take more than a minute to truly grasp, relate to other posts, etc.