Some Lessons on Design and Testing (and Building Maps, Too!)
design design underground
Hey, Descent fans,
It's Monday, and, surprisingly, today has been better in some ways than Friday. (Or Saturday's Texas Longhorn game - but that's a different story entirely.)
The short version is that on Friday I broke the game. Really, really broke the game.
We've had a lot of feedback about the ships, the number of ships and how we make them different. We know that it will all work, but we want to make sure that each player really does experience each ship differently.
As you've seen in previous episodes of Design Underground, getting ship balance right is a very important and very extended process. So we've been slowly iterating over the last couple of months, making small tweaks to the ship stats to get each one feeling a bit more distinctive, but making sure that they feel right, individually, as well. The changes up until the last few weeks had been pretty subtle, and the end result wasn't quite where we wanted it.
Furthermore, we made another round of balance adjustments based upon how hard each ship is to hit. That's a lot of changes, and the last few rounds of stats were bigger than usual to achieve that new balance. Unfortunately, the new stats weren't feeling any different at all (other than damage numbers, which created a very noticeable difference). I made 3 rounds of changes, with each round having bigger numeric variances for each stat.
They still felt the same.
Now, some of you may recall that Jason finished putting in the predictive ship movement in the last couple of weeks, as well.
Apparently the associated changes moved all of the movement controls to a new location in the ship data. So the reason all of my stats weren't feeling any different was that they weren't different! Ugh.
Once we sorted that out, my first thought should have been that all of the previous testing had been invalid, so we needed to go back closer to square one on the numbers. In my haste (and, admittedly, frustration), I just went with the numbers that were already there, but in the wrong place. (You know: the ones that had been drastically altered under the assumption that 2 rounds of changes weren't enough.)
Need I say more, really? Let's just be generous and call it "a bit difficult to fly" in Friday's play test.
So keep in mind when you test - if your results are consistently not what you expect, you may need to consult others about your expectations and why your results aren't behaving the way you expect. Maybe you're not understanding a change that's been made, or even something that has already been in place for a while. And if you do find that your tests are invalidated, don't plow ahead anyhow. It will only lead to tears.
Enough about my bad day. It all got sorted out on Saturday, with numbers that are doing the job much better (with the added benefit that energy consumption is better now, too).
On the brighter side, at long last we've launched (and streamed live) the first Design Underground tutorial on building maps using Unreal Editor 4. This one covers a lot of basics of editor navigation and placement, but it also gives you some necessary info for creating and testing actual Descent Underground maps.
If you're interested, take a look...
If the tutorial raises any questions, or inspires thoughts on what else you'd like to know about map building, feel free to leave questions and comments in the feedback thread on the forums, so I can address them in future episodes.
Future Design Underground episodes (or whichever Underground series is up to bat) will be streamed on Mondays at 3 PM CDT/CST (so either UTC-5 or UTC-6, depending).
Meanwhile, we're getting super close on the Proving Grounds release. Everything is in testing or very, very close to done. Keep watching this space for more information.
See you Underground,