WoaM 1 year anniversary - a developer's retrospection

Started by Andail, Sat 18/04/2020 18:48:02

Previous topic - Next topic

Andail

With Whispers of a Machine turning one full year, I thought it was time to sum up my experiences, and most importantly, share with you some things I'd do differently, because learning from others' mistakes is cheap and fast.
So here are my
Top 4 things I'd do differently with Whispers of a Machine

4.
Not use AGS. Yes I did it. I posted in the AGS forum and said I wouldn't use AGS. Now, the AGS community is the best in the world. I've been a part of it for 19 years and I've made so many great friends here (although I haven't been very active lately, admittedly), so I really hope it will always be a nice play to hang around, even though its members may eventually find other ways to create their games. But what took us (read Joel) months in AGS, namely porting to various platforms, Unity can do with the click of a button. Of course, if your happy with a PC only release it's less of a problem.

3.
Not split the writing part 50/50. Again I must begin with a don't-get-me-wrong-here: Joel is an amazing writer, and has a true knack for writing dialogue - that's not the issue. The issue for us was that every single word had to be minced, ruminated, regurgitated and compromised about before it'd end up in the actual game. Even though the game surely ended up better thanks to our shared writing duties, the process itself was really streinous. I would recommend others to divide the responsibilites more, let one person design the framework/background lore, and one person write the dialogue, and another write descriptions/messages etc, if you need to have more than one person writing the game at all (Ideally one person would be in charge of writing the game and the others would merely offer feedback and the occasional suggestion.)

2.
Not branch puzzles or game content. While it's certainly one of the selling points of the game, I still wonder if the multiple-solutions puzzles were worth it. Like #4, this is a matter of saving time that could be used for other things, so it's not that I think the result was bad per see. The main issue here is that only a fraction of players will ever play more than once (and some will turn to youtube to see other solutions or endings) so for most people it holds little value. Designing and implementing some of these puzzles took ludicrious amounts of time. Some puzzles even let you use two "skills" (or augmentations) in conjunction, and since there are three skills on level three and three skills on level two, you'll do the math and find that what for most players seemed like one solution actually had a high number of possible solutions that they'll never notice, much less care about. I can't help but think of all the time we could have saved on just settling with one solution per puzzle, and use the time to expand the game's length instead of width, if you catch my drift.

1.
Construct the game world so that there's less travelling back and forth. One thing we decided early on was to design a game world that you'd navigate manually, from location to location, like Lure of the Temptress or early Sierra games, instead of having the locations scattered around an abstract map and navigated using a fast-travel screen. The drawback is, obviously, that you force the player to walk around a lot, just to get from A to B. I do believe you can design the game so that you don't encourage the player to revisit all the previous locations every time they get stuck on a puzzle, but I'm not sure we succeeded here. One contributing factor was that you could ask every NPC about every possible topic, and that meant that as soon as you got stuck somewhere, one way to kind of brute-force your way forward was to go back and talk to everyone about every new topic, for the umpteenth time. Unless the writing is absolutely stellar, this routine will quickly turn pretty tedious. One good workaround that many developers use (Dave Gilbert, from the top of my head), is to construct nodes of locations where most of if not all items or clues are gathered. So instead of unlocking one location after another, and have the puzzle-chains stretch across various old locations, you unlock a new area, with a handful of new locations at a time, and put the necessary clues, items and NPC in those locations.
This whole issue has made me question whether letting the player ask everyone about everything is really necessary, but I won't make it a paragraph on its own.

There's probably more I'd do differently, but on a whole I kind of like the game so I don't wanna drag it through the mud completely.
Thanks for listening and see you around!

Ali

Congratulations on the anniversary! I just bought it, having had it on my wishlist for a while.

This is a great mini post-mortem, with tips that would be applicable to lots of game writers. I'd love to know what other devs end up regretting (in a positive way).

milkanannan

Hoping to play it soon. Have you done any podcast interviews or long-talks about the development process? I really enjoy listening to those types of conversations where developers go into the nuts and bolts of game making.

Durq

This was a great read. Thanks for posting, Andail.

SMF spam blocked by CleanTalk