{"id":237,"date":"2017-09-22T22:33:00","date_gmt":"2017-09-22T22:33:00","guid":{"rendered":"http:\/\/www.gamasutra.com\/view\/news\/306344"},"modified":"2017-09-22T22:33:00","modified_gmt":"2017-09-22T22:33:00","slug":"raw-data-devs-share-their-missteps-so-you-wont-make-them-too","status":"publish","type":"post","link":"https:\/\/sickgaming.net\/blog\/2017\/09\/22\/raw-data-devs-share-their-missteps-so-you-wont-make-them-too\/","title":{"rendered":"Raw Data devs share their missteps so you won’t make them too"},"content":{"rendered":"

Survios\u2019 Raw Data<\/em> has been one of the standout successes in virtual reality game design. But at VRDC Fall 2017 in San Francisco today, a pair of Survios devs took the stage to talk about how development went wrong along the way\u00a0— and what they\u2019ve learned from the experience.<\/p>\n

\u201cToday we\u2019re going to tell you everything that went wrong with Raw Data<\/em>,\u201d said Survios design director Mike McTyre, alongside Survios CT Alex Silkin.<\/p>\n

The game launched in Early Access last June, and went through what McTyre describes as seven major updates to date. While it\u2019s available on Oculus Rift and the HTC Vive (with a PlayStation VR version coming soon), the game\u2019s playerbase heavily skewed towards the HTC Vive — to the tune of roughly 80 percent.<\/p>\n

Here are a few interesting excerpts from their talk: During development, the team was initially going to keep the game\u2019s map count very small — maybe as few as just two, said Silkin, with some color variants.<\/p>\n

But when the game got popular, the team thought to increase the scope and add more maps — but that turned out to be kind of a waste, because at this point just 21 percent of players make it past the fifth mission.<\/p>\n

First impressions matter — most people play just the early bits of single-player, and not PvP or co-op<\/strong><\/p>\n

\u201cFirst impressions matter,\u201d said Silkin. \u201cA lot of players just played the first few missions and got the impression \u2018oh, this is a game wher eI just stand in a room and shoot.\u2019\u201d<\/p>\n

He says this kind of hurt the game, because the team spent a lot of time and effort making more complicated maps later in the game — maps that most players never see.<\/p>\n

The number two most requested feature for the Raw Data<\/em> devs to add, according to McTyre, was player-vs-player combat — but even though the dev team put a bunch of effort, again, into making a PvP mode they could be proud of, McTyre says only 13 percent of players have ever actually played PvP.<\/p>\n

\u201cOut of all our\u00a0players, 80 percent basically, have only ever played single-player,\u201d said McTyre. \u201cThat\u2019s an important takeaway for any other VR product, especially since co-op was such a selling point of the game.\u201d<\/p>\n

Despite that, only 16.5 percent of players have ever loaded up co-op play. And again, only 12 percent of players have ever played PvP.<\/p>\n

\u201cThis is an important takeaway for devs: while VR is growing, and there\u2019s a lot of users, creating a PvP-focused product, or trying to build a PvP-focused community, is going to be a challenge for you,\u201d said McIntyre. \u201cEven if you\u2019re going to try it, I would encourage you to have some kind of single-player experience as well.\u201d<\/p>\n

All of these numbers come from the analytics data Survios has been pulling on Raw Data<\/em>, which the pair said have been vital to ensuring they\u2019re understanding how people play the game. For example, they were shocked to see that only 1 percent of players ever used the game\u2019s defensive abilities, or that at one point 50 percent of players were failing the first mission.<\/p>\n

\u201cHalf of them are VR enthusiasts,\u201d said McTyre. \u201cThey were just people who loved VR, and that was the biggest takeaway for us…why we had to change things and nerf missions and that kind of thing.\u201d<\/p>\n

Also, \u201cone of the big lessons we learned was about the frequency of our updates,\u201d said McTyre. \u201cOriginally we were too ambitious\u2026.and we were gonna shoot for every two weeks.\u201d<\/p>\n

\u201cThat…really didn\u2019t work out,\u201d he admitted.\u00a0The team switched gears to try for a once-per-month update schedule, but in the end they had to settle on a once-every-3-months update cadence so they could put significant effort into each update.<\/p>\n

\u201cWe were in this Early Access period and we had all these users, and we wanted them to come back and experience the new content,\u201d said McTyre. \u201cBut what we realized early on was if we did small updates — just one map, just one feature — it wasn\u2019t bringing the users back. So we quickly realized that we neeed to have longer update periods and bigger updates to bring people back. And when we did that, we saw bigger spikes in players returning because the updates were meaningful.\u201d<\/p>\n

People play the same game different ways on different platforms\/headsets<\/strong><\/p>\n

Fun fact: Most Vive Raw Data<\/em> players (79 percent) prefer the game\u2019s \u201csticky\u201d control scheme over the \u201cHold\u201d or \u201cToggle\u201d schemes, whereas on the Rift, the majority (72 percent) prefer the \u201cHold\u201d scheme.<\/p>\n

\u201cIf you\u2019re releasing your game on different platforms, they have different controllers,\u201d said McTyre. \u201cSo don\u2019t assume that just because one of your control schemes is the most popular on one platform, that it will be the most popular across all platforms.\u201d<\/p>\n

And while the game was initially intended to only have teleportation-based locomotion, the team’s decision to add PvP meant they had to figure out how to get\u00a0joystick-based locomotion into the game mid-development.<\/p>\n

\u201cWe just couldn\u2019t wrap our minds around making Raw Data<\/em> PvP with teleportation,\u201d said Silkin. It wouldn\u2019t be fun if an opponent could just teleport away at will, so the team decided (despite heavy skepticism) to try using joystick-based movement.<\/p>\n

\u201cIdeally this is something you should plan for; we added it mid development cycle, so it broke a bunch of things,\u201d said Silkin. Since they hadn\u2019t scoped for joystick locomotion, the enemy AI could be easily broken by players just backpedaling, and had to be fixed.<\/p>\n

(And despite the fact that a vocal group of players clamored for joystick locomotion, McTyre says only 20 percent of players actually turn it on.)<\/p>\n

Incidentally, to do joystick locomotion on PSVR, which has no joystick, the team tried a \u201cbackpedal\u201d button — and it works well enough that the game will ship with it.<\/p>\n

\u201cOur biggest challenge to date was porting to PSVR, because we didn\u2019t plan for it,\u201d added McTyre. Silkin agreed and explained that because of that, there was a lot of reworking that had to be done.<\/p>\n

\u201cWe had a lot of issue with performance, mostly; we\u2019re CPU-bound,\u201d said Silkin. \u201cSony is very strict about performance; they don\u2019t want you to get their customers sick.\u201d<\/p>\n

\u201cOther importants things about PlayStation, besides not having a joystick, is that there\u2019s no grab buton,\u201d said McTyre. \u201cSo we actually had to change up our control scheme quite a bit.\u201d<\/p>\n

\u201cOne nice advantage? There\u2019s a lot of buttons!\u201d McTyre added. \u201cWe were very happy to have a lot of buttons to play with.\u201d<\/p>\n

To\u00a0optimize for the PS4 and PSVR, the Raw Data<\/em>\u00a0team switched over to a forward renderer and did some aggressive changes to the game\u2019s levle of detail — not only on assets, but on enemy animations (decreasing quality when they\u2019re far away, or behind the player), enemy count, and other aspects of the game.<\/p>\n

The game\u2019s biggest cost was enemies moving around the world, \u201cso we just kinda refactored our systems to not have as many things attached\u201d to things in the world. After a lot of time and effort, they had something that could stand by itself on PSVR.<\/p>\n

In closing, the pair offered fellow devs some tips for selling your VR game:<\/p>\n