[ANY ENGINE] WHAT'S YOUR BUILD/PLAYTEST LOOP LIKE FOR YOU, MENTALLY?

Posts

Pages: 1
Like are you constantly playtesting, noticing, and fixing minute errors, or do you just blast through a large area at a time even knowing it's missing features or events or has bugs, and then take on testing it all at once, or (more likely) somewhere in between?

I am towards the first extreme I described, but trying to be more like the second.
It depends on what I'm doing. If I'm coding something like a CMS, then every little feature whether it's exiting the menu or opening a tab, I always playtest for individual things first, and fix them, that way it's more consistent.

I feel like making a CMS and never playtesting it until it's mostly finished might cause me to miss or forget something. Then you end up having more bugs or errors to keep track of, instead of ironing them out as you go.
I try to test the most relevant thing I've done. I only do longer testing if I have a test build to send out to actual testers (or the game's going to release). I think the most wasteful testing is when you've fully implemented something and you're sort of just messing around with your own game and not really going after bug fixes per say but very minor "would be nice" improvements.
It depends on what I’m testing.

Usually in the prototype stages it’s the former, but if I’m working on balancing and area or something it’s more of the latter.
The one thing that MV has over the other engines is that the editor isn't locked while playtesting. So I'll fix as I go with that, while needing to take notes / remember everything that's broken on a map at a time when using VX Ace.

Text not fitting in message windows is the bane of my existence when it comes to RM.
watermark
Got me my shiny new MZ
3283
I like a tip from a famous author (I think Dan Brown said it? I forgot...) who said that he finishes all chapters before diving into the details of each. The reason for this is that if you "polish too early", you may end up polishing a portion that you should've thrown away because it's not working in the overall picture. And it's much harder to "kill your darling" the more time you spent on it.

In game terms, I think it best to get the whole framework working first. Make something you can play from start to finish using filler maps and dialogue ("Hero says something romantic here." "Dwarf dies." etc.). Then if it all makes sense, start filling out the necessary maps and events.
All of the above. Sometimes I'll create an entire dungeon and not test it until it's finished. Other times, as other users have mentioned above, I'll test every little thing (especially if I'm working on a custom menu or puzzle).

I find that it's better to test entire dungeons at once, because if I create and test/play-through the dungeon piece by piece, it's hard to see if the dungeon actually flows well, which is important to me.

I also do complete playthroughs of my game constantly. This is where I tend to find the most bugs. In these playthroughs, I usually save, close out and fix each bug I find before I continue playing.
author=coelocanth
The one thing that MV has over the other engines is that the editor isn't locked while playtesting. So I'll fix as I go with that, while needing to take notes / remember everything that's broken on a map at a time when using VX Ace.

Text not fitting in message windows is the bane of my existence when it comes to RM.

Wait...IT ISN'T? In MV I can use the editor WHILST the test play window is open?
AtiyaTheSeeker
In all fairness, bird shrapnel isn't as deadly as wood shrapnel
5424
I usually playtest after small chunks, but I try to get the dialogue fully churned out for story sections first. Which can be a pain if I run into issues during playtesting, and have to take notes of typos and where they occur. Usually I playtest shortly after I integrate things onto maps though, once a decent swath has been cut. I've had times where I haven't finished all of a dungeon room before testing it, but usually I make a town with buildings but no doors and most NPCs to test first.

Much of the time I try to get basic databasing done before any playtesting first, either adding or dummying out content as it becomes relevant. Usually subtracting. Really databasing stuff is where much of my game-planning occurs, and without having a game engine to test in context it can be rough. Especially because I'm a perfectionist and tight combat is important to me.

tl;dr Don't follow my example as a game dev. XD

Although I will make this suggestion if one gets stuck on making a game, especially as a pantser or discovery artist. I plot out creative works with vague bits of what I want to see, and make it vague on how to get there so I make it up as I go. Not sure how relevant or helpful this might be, but is my two copper pieces on the subject.

author=watermark
(insert post full of good-ass advice here, check the post a little ways above this one you lazy twit)

That... that is some good-ass advice. I should keep it in mind for my next project. :o
Pages: 1