So I’ve historically had very significant problems with building mid scale games and playtesting. I’m looking for better ways to set this up and organize it going forward.
In higher end game dev, I believe that it’s common practice to leave the early introduction and tutorial areas for last in the dev process. Generally, these pieces need to be the most polished, and tutorials obviously need to be correct.
For very small scale indie though, I’m starting to think you really need to start here. That it’s not really possible to find playtesters or get meaningful feedback without a fairly good tutorial being in place.
Is this true? Should we as low resource indie devs, start with the tutorial? How much should you polish builds for pre-alpha playtesting?
I think this subject is really not well explored or frequently discussed. Although “playtest often” is unanimously considered good advice, the process to organizing games for pre-alpha play testing is really quite challenging.
How have you dealt with this? My experiences with attempting to run play testing rounds have been incredibly time consuming, very difficult, and often less productive than I had hoped.
Well, for playtesting I’d suggest most basic instructions regarding controls or whatnot (unless game has controls familiar from another game, such as it is FPS) then letting them figure it out. If they can’t, you take note and redesign the area so people can figure it out by themselves. Not only people will feel accomplished, after all they have figured out mechanics by themselves but you’ll get to make tutorial-less game and earn some street cred for avoiding chastising (as tutorials often are) tutorial.
Playtesting has 2 parts. The developer has to playtest the mechanics and ensure they do exactly what they want, time has to be spent honing that part. No-one else can really help with that. As regards getting playtesters for the game that can only be done when a game has a complete framework, then a playtest can be run on one finely polished level. Unless a playtester is trained in Q & A which is different from just getting the general public to play the game, then you will only get meaningful data from a polished level. People cannot extrapolate beyond what they see before them for the most part, they won’t react to a wow element of the game if it is not present, no-one will say ‘if only it had …’ In general you will simply get a ‘That was good…’ Players generally don’t want to read much when it comes to a tutorial so it’s best to build it into the test level if possible (it depends on the genre). At no stage demand much of a playtester in that regard because it won’t reflect the reaction of the buying public, you have to treat the playtesters the same as any player so if you are going to put more control in for players on release, don’t, put it in for the test version right from the go. Get the test version as close to a release candidate as possible.
This is really the kind of conclusion I’ve also come to. So is it actually unrealistic in your opinion to run meaningful playtests in a pre-alpha stage without polish?
Alternatively, is it worth while perhaps polishing prematurely in order to gather feedback early in the process?
Yes, you could polish prematurely but by that I mean instead of a level do a room which may take 5 minutes to get through and polish that. It’s not enough to judge whether the game keeps the attention of players but it will mean you can tweak for the wow effect. I noted in one of your posts you mention destructability and show a pillar being destroyed but decided to take that feature out. It could be that feature that gives that wow factor that will get players interest and give players a lot of satisfaction. A one room playtest will highlight this. Most developers are doing games because of a game they played when young, they remember that game because for them it had that wow reaction - for any game to be successful you have to give that to your target player group. Players are fairly forgiving if they like a game, so in the case of the destructability maybe all that is needed is not full destructability just some, and then really meaty sfx both particles and sound that give that feedback to a player. A small room, well polished will allow you to bring those reactions out in players.
I guess I’m also just wondering about play testing mid-sized games in earlier phases in general.
I think that my expectations for play testing is not really aligned correctly with the results of play testing (again, at early phases). Mostly the kinds of things I’m looking for are sanity checks on rougher work. The idea is to use play testing as a tool to preempt mistakes or bad design. Like “stick a mechanic in there and see how it works” - the problem being that the actionable feedback is generally very slim.
In the last 10 weeks, I’ve probably spent about half of that time on trying to polish stuff up for a bit of play testing (2 times in those 10 weeks). I wonder if that time would have just been better spent building out more functionality, and attempting to get hands on feedback less frequently.
Have other people had more productive pre-alpha playtests on mid scale projects? If so, what kind of useful feedback did you get and what kind of state what the game when you did a round of play testing?
Is the real value of ‘feedback early and often’ just that keeping playability as a constant reminder keeps you from forgetting about the player?
The problem is not neccessarily the frequency but the feedback. To get that kind of stuff you need someone trained in Q & A or at least a fellow developer that makes games and can maybe see where the bug or where a tweak may be needed Usually the general public don’t really know how to describe a problem, they have no idea what may be causing it. As regards the other feedback you’re best getting a professional to help. A year or so ago some industry professionals put a list together of fellow professional developers who were willing to give advice and mentor indie devs. Not certain but it was possibly mentioned in gamasutra. Maybe a google search would turn it up. A professional impartial developer is more likely to give better feedback and be able to describe the problem more correctly and to suggest tips and tricks and improvements to how the game feels.
I’ve had success taking my game to play test at local IGDA events. Other devs tend to give good feedback. The design forum is also a place to get dev centric feedback.