Jump to content

Fireside Chats (with a video game designer) - September 10, 2014

Recommended Posts

  • Developer

Post-PAXtum Elation


Hey folks! It took me a bit longer to get to this than I expected (due to a combination of post-PAX settling-back-in and me taking a trip to North Carolina a mere 40 hours after getting back from PAX like some kind of idiot), but here’s your PAX report, as promised.


We did it! We went to PAX! We even brought Don’t Starve Together with us! We had a really rad booth, mostly thanks to a brilliant lock-picking challenge that @Corey and @JoeW cooked up as a tie-in for Invisible, Inc, which we were also showing. Check out some photos of the booth here. There were 10 of us working the show, talking to fans, selling merch, talking to press, showing the games and generally helping to keep things running smoothly at the booth. As ever, it was an exhausting weekend (4 days of PAX is a lot of days of PAX), but it’s always super energizing to talk face to face with so many excited fans. I even met some of our lovely forum-goers :-)


I was mostly hanging out in the Don’t Starve Together area of our booth, so I didn’t see too much of the reactions to Invisible, Inc itself or the associated lock-picking challenge. The challenge did have a capped line basically all weekend: the line wrapped all the way around the booth. So, I guess that went pretty well. We were right by the expo hall entrance, which meant we we got a lot of traffic in general. All in all, I’d say it was a huge success. But how did it happen (hard work)? How did it all come together (it came in hot)? What was it like there (madness)? How did we feel after (pure exhaustion)? Read on to unravel the great mystery (it’s not actually a mystery; just some game dev stuff that we did)!


The usual bit about spoiler-tagging sections ‘cause I’m long winded as all get-out applies. You're going to want to read this one to the bitter end, trust me. Expand and enjoy.


Getting the Build Ready


As you may remember me talking about in previous posts, back somewhere around the middle of July, I sat down with the rest of the Don’t Starve Together team and discussed whether or not we could get a build ready to show at PAX Prime 2014. At that point, you could play on a server with other people but there were a lot of important pieces missing that prevented normal play. This ranged from backpacks not being usable to science machines not correctly unlocking recipes. It was very much a proof of concept; a prototype. We made a list (called a burndown list) of all the tasks that needed to be completed in order to have a build we would be proud to show off to hundreds of people at the booth (plus all the people that just stopped to watch but didn’t sit down to play).


As we worked in those first couple weeks after deciding to target PAX, Don’t Starve Together really started to take shape. Mostly this came from fixing the major bugs that prevented any prolonged play (i.e. the science machine thing mentioned above). We were making good progress on the burndown list and had another meeting. We talked about the new features we would need to add to the game to make it feel like a multiplayer game (i.e. chat, event announcements, player indicators, etc). We still felt like it could be done, so we added those things to the list and started our work on the alpha. We were out of prototype territory! From there, up until we got to beta, everything we were doing would be the alpha version.


I was away for a couple weeks at the beginning of August and when I got back, I was greeted by an enormous amount of progress made by the other folks on the team. Make no mistake, we still had our work cut out for us: insidious little items on the burndown list like “ensure stability” can take a long time to achieve (and then to test once we think they’re done, for that matter), even though they take just one measly line on the spreadsheet. About a week and a half before it was time to leave for PAX, we had some local sound design students visit the office, so we seized the opportunity to have some people play the current build while we watched--watching people playtest your game helps you find issues (both actual bugs and experiential problems) that you’d never find if you were in the driver’s seat. Technically, they were the first non-Klei employees to play the Don’t Starve Together alpha, but the build they played was nowhere near as finished as what we brought to PAX.


We kept working and cut our first release candidate (well, demo candidate: “release candidate” is just what it’s generally called, regardless of whether or not it’s truly being released to the public) build on Friday: the advance team of Klei employees seasoned in the ways of booth-building (@Corey and @Mattesque) was scheduled to leave on Tuesday and would be taking computers with the build loaded onto them down with them. We wanted a couple days of buffer time for testing to guarantee that the build was stable. Inevitably, we found issues with that build. We sent release candidate 5 down with the booth-crew on Tuesday. They actually ended up having some issues at the border due to the value of all the commercial goods (computers, monitors, etc that we’d be using to demo our games), which meant that we had to scramble to find hardware in the US that we could use for our demos. It also meant that we would be bringing the demo build with us on Thursday when the rest of us drove down. Generally, you don’t want to be changing things that close to showing a game, but it ended up giving us a chance to shore up a final few things. The build we ended up using on the floor was release candidate 13. The build we brought performed brilliantly: better than I could have hoped. Plus, it existed. Which was great.




On the Floor


This section is more a retelling of the weekend than anything else. If you’re looking for analysis and that sort of thing, head down to “Takeaways”.


Seven of us arrived at the convention center on Thursday around 1 PM. The main structure of the booth had been built by the two folks who’d gone ahead, but there was plenty left to do. The last member of the PAX team would be joining us in the morning to help actually run the booth. That afternoon and evening, we wrapped the booth with the large-format artwork, set up TV stands and TVs, got trailers looping on those TVs, set up our merch area and took inventory, set up the computers for demos and made sure the demos worked correctly, and generally made sure the booth was tidy (i.e. not a minefield of garbage and wires, which it was for most of Thursday). Oh, and @Corey was there until about 3 AM making sure that the glass case for the lock-picking challenge was all set up correctly. That thing was insanely complicated and he knocked it out of the park (check out the pictures I linked above if you don’t believe me).


Friday morning at 8:50 AM (10 minutes earlier than scheduled, due to people melting in the rain), the floodgates opened and attendees started pouring into the expo hall. The allotted plush beefalo for the day were gone in minutes. Chesters weren’t far behind. Before long, we had people seated at all of our demo stations for both games (4 stations each) and a line forming for the lock-picking challenge.


Our set up for Don’t Starve Together was four computers out in the main area of the booth, all connected over LAN to a fifth computer set up in our back area, which served as the dedicated server. When a player sat down at one of the four client computers, they would start the demo by clicking a “Join Server” button. Once they connected, they would select one of four available characters (Wilson, Willow, Wickerbottom or WX-78) and spawn at one of the spawn points scattered around the world (which we had doctored a bit to be relatively small and resource-rich). A fifteen minute timer started, at the end of which, they’d be automatically killed and disconnected from the server. Every two hours, we would trigger an event that we called “end times”. You can see a video of it in action here (the “Let slip the dogs of war” message at the top marks the beginning of it). Basically, the event was 30 seconds of constant hounds spawning, followed by 30 seconds of hounds spawning and meteors falling. We wanted to kill players in a dramatic way to mask the server resetting: the alternative was for them to be abruptly disconnected without any warning or closure. Since we weren’t able to soak test very thoroughly before PAX, we settled on this periodic server reset rather than risk the game getting choppy on the floor.


The game was actually remarkably stable! We had one server crash all weekend and only one other major issue: sometimes a player’s controls would stop working when another player joined. Even the latter issue only happened a few times a day. There were definitely other small issues, but nothing glaring (mostly stuff like slightly bad UI behavior that wouldn’t be noticed unless you’re already quite familiar with the game).


Until 6 PM that day (and the rest of the days), we all stood at the booth, greeting people, talking to them about our games and helping out if people had difficulties while playing. We had a range of people come by: from people who’d never heard of Don’t Starve and were attracted by the art style to people who knew and loved Don’t Starve and were already aware of and impatient for Don’t Starve Together.


The response was overwhelmingly positive. There’s definitely some confirmation bias at work here; they wouldn’t have stopped and played the game if they weren’t at least interested, if not excited. But just the same, it was an energizing (in the long term, anyhow) weekend. Watching people play all weekend long uncovered some things that we need to change around a bit, which I’ll get into more deeply in the “Takeaways” section.


The Tournament


I actually didn’t have much to do with making the tournament build or running of the event itself: that was all @cheeriodude coming through in the clutch. Just the same, I’ll give a quick recap.


We had 40 people sign up to play in our tournament, which was the second one we’ve done. The first was at PAX East earlier this year and was just the base game. This was actually a properly multiplayer tournament: we had a special build of Don’t Starve Together for the entrants to play. Groups of four would join a server and see who could survive the longest. The victor would advance and play in a new group of four for the next round. PvP damage was increased. Hunger decay rates increased over time. The end times event would trigger in this build too, but it happened much earlier, so it lasted longer.


In the early rounds, we saw players trying to attack each other, but as the rounds went on, only the cleverer players were left and by then they’d realized that it was too easy to run away from the melee attacks for it to be worth attempting to off other players. However, there was one really awesome moment where two players were near each other at night. Torches blipped in and out as they kept Charlie at bay while dodging in for attacks in the darkness with fiercer weaponry than the torch. Generally, though, their time was better spent collecting stocks of food and materials to make various armors to keep them alive during the end times.


We had approximately as many people watching and cheering as we had participating. The crowd was loud and great! It was super fun and we’ll definitely continue to do these. We hope to release a tournament mode, too. Oh, and congrats to @eryahu for winning the tournament!




All right, outside of technical performance/stability, traffic at the booth and player-reported experiences, how did we do? Pretty great, I think! There was a solid range of behaviors: some people wandered around by themselves, others dashed for the nearest player and went in swinging and still others found other players and played cooperatively, making some pretty solid progress within the allotted time. People were visibly having fun. Even the occasional yelp of “Hey! Stop attacking me!” was good-natured and through laughter.


It’s tough to get a real sense of it in a 15 minute demo: as you’re probably aware if you’re reading this, Don’t Starve is a very slow and deliberate game. Just the same, it was very encouraging and people seemed to really enjoy the simple act of play Don’t Starve with other people around. Their mere presence had a clear and positive impact to the core experience. Don’t Starve Together is a different game. Not massively so, but it is.


It wasn’t all rainbows and sunshine, though. I don’t mean that there were any ghastly realizations about things that were fundamentally broken or flawed (seriously: all in all, I’m very happy with how it went), but watching people play definitely brought to light that course-correction was needed in certain areas. I’ll talk briefly about a couple illustrative examples.


The first example is the ghosts mechanic and associated UI: the two separate meters is too complicated and the humanity penalty doesn’t add enough depth to warrant the increased complexity and confusing UI. We actually removed the penalty from the demo build because with the short play time and the negative feedback loop it creates (each penalty makes it easier to die again and get more penalty), it seemed like a poor fit. I expected to feel like the ghost mechanic was gimped without it, but I didn’t miss it at all. In fact, we had cranked up the speed at which your life force drains to ensure that players wouldn’t stay as ghosts for too long: we just wanted to offer a taste of that mechanic, rather than sell it as a main part of the game. After PAX, we had to remove a bunch of hacks that we’d put in place to make the demo how we wanted it. The absence of a penalty and the increased life force drain rate are two hacks that I left in. Right now, it’s tuned such that you have about a minute and a half (before any haunting) before your life force is all gone. The shorter ghost duration and lack of penalty for longer spells as a ghost felt about right. I think that with the greater time pressure and less to process (i.e. just one meter and no penalty), the ghost mechanic will be more easily parse-able while still feeling tense.


The second example is the player indicator UI icons. For the demo, they were set to be always on. This was extremely useful for players that had only a short amount of time with the game: it allowed them to quickly locate other players and start having the multiplayer experience they came for. It was absolutely the correct thing for the demo at PAX. But I pretty strongly believe that it’s not the right thing for the “real” version of Don’t Starve Together. The main problem with them as they existed in the PAX build is that they are a UI element that is A) very precise and B) strongly in the meta layer. Don’t Starve is very light on that sort of stuff, and I’d like to keep it that way as best as I can. When there are meta UI elements, they’re either imprecise (hunger meter, etc) or strictly functional (crafting tabs). To fix this, it either needs to be toned down or to have some fictional justification. To tone it down means to make it so that they only appear after you’ve seen a player on your screen and so that they disappear if they get too far away. To fictionalize it, we would need to add a beacon item that you can give to another player, or something along those lines. I’m not sure yet which solution I prefer--I’ll likely have to just implement both and see which suits the game better.


So, yeah. It went well overall and we learned some valuable things. Of course, we got more notes than what I’ve shared here. But in the interest of (relative) brevity, I’ll leave it at that.


After PAX


After we sat down and talked about what we saw at PAX, we collectively realized something pretty exciting: it’s already fun! Just being in a Don’t Starve world, playing Don’t Starve in whatever way you choose, with other people around doing the same (and expanding your play options) is fun. Sure, we’ve got more work to do before it’s done (i.e. has parity with single player in terms of what modes/content is there) and plenty of opportunities to add stuff. In the extremely short term, the work is along the lines of “implement a proper server browser with real art”. In the longer term, it’s making sure we deliver that parity and any game modes or content we want to add.


But that realization bolstered our confidence about where we’re at and what the road ahead looks like. In conjunction with what I said above about treating the PAX build as our alpha version (in the “Getting the Build Ready” section), we were convinced it was time to move into beta and start sharing the Don’t Starve Together experience with more people around the world.


With that, one last (but certainly not least) thing: we’ll be opening up sign ups for the closed beta of Don’t Starve Together this week. But since you’ve suffered through my lengthy posts time and time again, here’s a hot scoop for you: the application is actually up right now! Ehem. We’ll be making a formal announcement for the general public soon.


As always, thanks for reading and keep the discussion going!

Link to comment
Share on other sites

Perhaps regarding the issue with player indication icons on the UI being too meta, you could layer on some significance and multi-functionality to existing items with currently limited usage within the base game.


The Compass is a good example, as it requires specific enough materials that you have to know about it and actively gather one of the materials (papyrus is generally a resource I don't collect unless I'm specifically trying to make a Bird Cage or other item) yet its only real functionality is to check your current orientation, so you can re-right yourself if you rotate the screen too much and forget the initial angle you were playing from. That in itself is a rather minor functionality, and it took me a while to realize this was the only real purpose of the Compass aside from looking neat. Perhaps the Compass could be reworked, or even be involved in the crafting of a new item recipe, some scientifically/magically conductive device to guide yourself to other players, like a magnetic detection field of some sort.


Perhaps you can use this Magnetic Tracker (I'm just going to call it that for now) as some strange guiding mechanism that could function like the Divining Rod, only effective when equipped in your hand, except it would vaguely direct you to other players who also hold a Magnetic Tracker in their inventory, increasing the intensity of an audio/visual frequency as you become closer to one another. This could be a great way to reunite with allies, or detect potential threats, all without the need of a blatant UI indicator, not to say that is also a nice option to have for those who want it.


Anyways, fantastic update as always, thanks so much for keeping us in the know Seth, and thanks for that "hot scoop" ;)

Link to comment
Share on other sites


This topic is now archived and is closed to further replies.

Please be aware that the content of this thread may be outdated and no longer applicable.

  • Create New...