Weekly Topic Play, Test, Repeat!

Happy Monday!

This week let's talk about something very valuable to any project, playtesting! This can be a great way to get much needed feedback to a developer who has been is heads down in their project. Getting a fresh pair of eyes on anything can be a productive way to find small errors and oversights, or even find that the way you’ve designed something might not be the way most people try to use it!

How highly do you value playtesting? At what point do you start inviting people to give feedback? Where do you source your testers? Are they friends and family? Or maybe your local scene has playtesting events?

Made with Unity devs Coffee Powered Machine did an interesting write up on their playtesting process on their website (Link).

“When the playtester were playing, we tried not to talk to them to avoid distracting them. We also told them not to ask us questions. If there was something that they couldn’t figure out, we would jump in only if it was absolutely necessary.”

Share your own thoughts and process with the community! If you're interested in some playtesters you might even be able to find some here!

 

It all starts with paper prototyping. If your playtesters can't find the fun in a paper prototype, the design probably isn't ready to prototype on a screen.

 

I usually get friends to play stuff when they come over to hang out. It usually went like this...



But eventually I got them trained and now they provide decent feedback most of the time. It also makes it easier for you to focus on the core of the project instead of doing smaller things around the exterior. If the test build doesn't have enough content to actually test then you're probably doing it wrong and having random people try to play it from scratch makes it clear where more work should go.

 

While I wouldn't disagree, do you find that there is an equivalent way to paper prototype games that don't necessarily translate well into physical space? I do agree that paper prototyping works extremely well and very efficiently

when working on something that ties heavily back to board / card game roots.

While that image is 100% beautifully correct when I've done tests with my friends. Do you think there is merit in building out a flushed out demo level that also emulates the style of art you're going for? Art and design are both areas that can benefit from feedback and critique.

 

To date I've only made one Mod (pretty well received too ) for a dungeon crawler that I released for public consumption. This was made through an editor that came with the game and used LUA for scripting. I must have played through it 100 times, start to finish, looking for bugs, balancing play, etc.

When I finally thought I was ready to have it play tested it was pretty easy to get testers as the game had an active Mod community. After the dust cleared there were two bugs found pre release and two found after release. All 4 bugs were directly related to players doing something I never would have thought of.

In this case though the game play was already tested (being just a Mod), so play testing was more for bug finding than entertainment value, but it sure taught me the value of play testing as a resource. Don't know if I ever get to the point of having something ready for release in Unity (much more involved than a making a Mod), but I will definitely be looking for play testers if I do.

 

Absolutely! The game world is often a simulation of a physical space, so you can usually find a fairly natural mapping. 2D games are easy. Use a gameboard or a whiteboard. My game jam teams often prototype arcade-style 2D games on a whiteboard using post-it notes for the game entities. We've even done a couple local multiplayer prototypes this way, which allowed us to refine the gameplay in very quick iteration cycles before loading up Unity.

Tracy Fullerton's Game Design Workshop has a chapter on prototyping, including some good photos that show a paper prototype of an FPS, which might be the last genre you'd think you could paper prototype.

 

I think that playtesting is one of those things that everyone thinks they can do, but really it takes a very specific sort of person, and in the end you either can or you can't.


Good play testers:
-Look to appreciate the vision that the dev is trying to convey and make suggestions to highlight it.

Bad play testers:
-Look to highjack the project and attempt to get the dev to make the game they want.


Good play testers:
-Identify what is off, why it's off, and how to fix it.

Bad play testers:
-Find things to complain about.


Good play testers:
-Allow themselves to first find the intended flow of the game and play it as a talbula rasa.

Bad play testers:
-Immediately start criticizing and pulling apart.


Good play testers:
-Go back and note geometry cracks, texture seams, stretch textures, floating objects, places the player can get that the shouldn't, and other incongruencies.

Bad play testers:
-Get bored of doing the less fun work and blow it off.


Good play testers:
-Acknowledge that they are talking to a person who worked hard on the thing they're looking at, while remembering that they put too much effort into it to let it be mediocre.

Bad play testers:
-Avoid all criticism for the sake of dev's feelings, or act like a pompous ass then act like they're doing you a favor for it.


Good play testers:
-Understand concepts like "flow", the difficulty sawtooth pattern, balance, asymmetrical game play, pallet cleansing, set pieces, basic cinematography, art, and story structure.

Bad play testers:
-Are UBER L33T GAMERZ!!!11!!!!1111!1!!1

 

If you want really, really good bug testers, find speed runners.

We had some runners of Teslagrad turn up to test World to the West, and boy did they find out of bounds bugs like nobody else. A good tester is somebody that will go into an innocent looking corner and do a special move a hundred times to figure out if that breaks something. A speed runner is somebody that does that for fun.


The nice people over at Hamar Game Collective does a really cool thing they call "test bonanza". It's an open-for-all event where everyone that has a prototype/alpha/beta/whatever can show up to test each others stuff. For indie-sized companies that doesn't panic about NDAs and such, it's perfect, and also a bunch of fun.

 

What kind of features do you all build into your games to facilitate playtesting? For example: analytics, debug logs, consoles, screenshots, and ways to make it easier for playtesters to submit feedback to you?

 

I do two forms of playtesting. One is automated, which makes sure that components of the game are as close as possible to be free of bugs. The human playable version is mostly done by three people that are my target audience. However, a full playthrough simply wouldn't work, so they get pre-created save games (exported from the automated part) to play test if the story elements work. Doing it like this doesn't exhaust a beta tester to run through a game again.

I love that idea where the game should be working on paper. In my case, I prototyped the story and choices as a pure text-based game prior doing the heavy lifting. But I must say, many people like the art they have seen so far - not because it could compete with all those great artists and studios out there, but because it just "works".

In regard of good or bad testers, you do not always want a speed runner. If you just want to eliminate bugs in the game, have rather a coder or a good friend to walk through if you don't have the time to build a framework. Often I get asked to try out someones game - and the providing party rarely provides me a tool or a form to give feedback. That road goes both directions - most people (paid or voluntary) want to help you to make a better game. If you can't absorb criticism - that is not about the tester. If you ask a tester which can not give you constructive feedback because you failed to determine what you expect them to do, that is not about the tester. If the tester makes your wall have scratches with blood and make you considering a felony that involves an axe... that one is most of the time about the tester

A simple report sending tool adds a screenshot to an online tool for report a bug and includes the saved snapshot of that part of the game. I have never exceeded the amount of human play testers beyond two people though (too much love will kill you).

 

Dungbeetle!
Paid
Free

One of the other programmers on our team built that. It's an in-game bug reporter and in-editor bug tracker. It's been a godsend - the tester clicks a button to get a screenshot and a text input field to describe the bug, and then we get the bug directly in a window in Unity, with a bunch of sorting tools.

You're right! Speedrunners-like people approach the game in a certain way and find a certain type of bugs. You also need people who can give good feedback on why doing something isn't fun, or why they don't understand the interface.

So addendum: try to get a varied set of testers. Getting info from other game makers can be helpful, but be sure to get people from outside the industry too, as game people might have a too meta approach to spot some obvious things, or be too familiar with tropes to not spot things.


Story time: Pretty early in our game (see sig), one of our characters gets a shovel. The shovel can be used to create holes in the ground which you can enter, and then you can move around underground like a mole. We struggled a lot with teaching the player how that works. Mostly because testing tutorials is horribly hard - once you've had a tester try the tutorial, that tester is forever unable to test that tutorial again.

I think we end up with holding the players hand too much because by the time we hit a teaching method that works, we've pretty much spent every single tester we can get our hands on, and iterating more would require scavenging up new people from God knows where.

 

Cool! Thanks for letting us know about it.

 

To me, a "good playtester" is someone closely matching my target audience who is available to play the game. They don't need any specific skills because the solutions and the design are my job. All I need from them is willingness to be exposed to the game, and their reactions and/or feedback when they play it. From there it's up to me to identify issues and decide on solutions.

A bad playtester is anyone who doesn't fit the target audience, because their reactions and feedback will not be useful for me in making decisions.

 

That... looks amazing!

 

I think it's important to have play testers that will be as a honest as possible especially about the fun factor.

I had various friends and relatives test my first game (which was NOT fun--something I learned from the market later). They were overly kind and complimentary. My 10 year old daughter, on the other hand, said: "I don't get it. It seems kind of boring." I ignored the feedback and chalked it up to her age and the fact that the game (a FPS) was different than things she would normally play. I should have listened to her. And if I would have went beyond friends/family, I certainly would have heard the same thing loud and clear. Then I would have gone back to work on it instead of going forward with the ill-fated launch. Live and learn.

 

Ermagherd!

 

Playtesting is really important because there are aspects of, not only mechanics, but polish that you might not think about as you focus on getting the game done. I definitely wish I had put Party Animals out into the wild somehow before I tried Kickstarter and Greenlight. I got crushed with negativity, and it was all my fault.

So, very costly lesson learned, I've polished the game up in response to the criticism and plan on putting it on itch.io's refinery soon. They allow you to limit the number of free downloads, so it's very good as a closed beta platform. You can also set up forums for the game to gather feedback, so I think not getting on Steam right away was a blessing in disguise. Greenlight will probably be gone very shortly, so getting on Steam will just be a matter of paying whatever the fee is. I still think putting any game on smaller stores and making sure it is the best product you can create will be the correct way to go. A failed Steam launch will get in front of a lot more eyeballs than a slow to start Itch.io soft launch would.

 

Well if there's a Unity service for it, perhaps as thin wrapper around a reliable 3rd party that can just pull things off cloud build, then, well.

I guess it's only cost effective for both parties if there's enough people actually needing it. I would use it if the price was right.

 

I've got way too many stories on play testing. So, here's the TL;DR. Open with: "We need your help. We're looking to improve our game. We'd just like to just watch you play so we can see where the game needs to improve." Then, hand 'em a device, say nothing more, and watch. Don't answer questions, don't provide any guidance, and just watch. Or more simply stated:

"Shut up and let 'em play"

Haven't seen anything more effective.

Gigi

PS - If you want feedback on your game, Feedback Friday #54 is open right now in this very forum. Give feedback to someone else, or get feedback on your own creations.