State of Production #7
April 10, 2020
My name is Derek, and I am the Head of Production at New World Interactive. I hope that everyone has been staying inside, and staying safe during these unprecedented times. With COVID-19 affecting everyone across the globe, we sincerely hope that all of our players within the community have been minimally impacted thus far, and continue to remain healthy. While there’s no telling how long this pandemic will continue to drastically impact all of our lives, the best we can do as game creators is to take full advantage of the opportunities we have in our field to continue fine-tuning our game, and to supply an added reason to stay indoors! We’ve got a somewhat light SoP this month, with the 1.6 update going live last month, as well as our hotfix patches that were deployed within the past week. Without further ado, let’s jump right into it.
Key Points of State of Production #7
- COVID-19 Statement and Update from New World Interactive
- 1.6.2 Patch Overview
- Free week retrospective
- Fixing spawn issue on Frontline and Power Plant exploit fixes
- Potential improvement for SKMM timetable and
- Clipping Fixes + Tavor 7 1x Kobra alignment fix
- Smoke fix timetable
- Tutorial Overhaul
- Refactoring the Progression System for New Players
- Discussion about monetization + new skins
- Looking ahead to 1.7 Update
COVID-19 Statement and Update from New World Interactive
For those of you who have followed the SoPs over time, you’re likely aware that one of the big reasons we came to Calgary was in an effort to centralize the studio and the team. We originally operated in a remote-only capacity, given our modder roots, and only in the last few years have we been able to create a unified site for collaboration together.
As many others have undoubtedly experienced over the past few weeks, we here at New World Interactive were not exempted from the impact of COVID-19. Our official statement on social media can be viewed here. I’m pleased to report that the studio was able to be proactive and responsive to the pandemic situations. Our team was able to act decisively and expeditiously, and we rolled out a work-from-home approach within 24 hours. Thus far, we have had no team members directly impacted by the virus, and we were also able to not only ship our 1.6 update relatively uninterrupted, but 2 additional hotfixes as well in that time frame. Additionally, the company made a sizable donation to the Calgary Food Bank, to ensure that those who are unable to provide for themselves are not left feeling gone and forgotten. I’m extremely proud of the company, the infrastructure team, the leadership team, and most importantly, all our employees for helping to ensure that we were able to get everyone into a fully remote capacity within a single work day. I also applaud the efforts of our company to give back to the community in a time where so few people have been provided so little relief.
Not only were we able to look after our team, we were also able to expand it. In the past month, we’ve hired several new team members. We’ve added additional headcount to our QA team, our Support Desk team, and our Community team, which should bolster our ability to catalog feedback, troubleshoot technical issues, and engage further in community development and outreach efforts. We’ve also recruited an additional backend programmer, and an AI programmer, which will be instrumental for our future updates. This should allow us to be even more actionable and agile when it comes to our front-facing communication efforts!
1.6.2 Patch Overview
The 1.6.2 patch contains a range of fixes for issues discovered after the 1.6 release as well as a number of new modding features requested by the community. Primarily we tried to clean up issues that were either unable to make it into earlier patches due to time constraints or hadn’t yet been detected. The patch itself was scheduled for release just after the free week as we didn’t want to interrupt players in the event with a download. Full notes for our most recent patches can be found here for 1.6.1 and here for 1.6.2
Free week retrospective
After the launch of the 1.6 update and it’s free week, I thought it would be frugal to invite our Communications Director, Jon Higgins, to share some business analytics and metrics with the community.
The March 1.6 Update and Free Week was one of our most successful events to date, welcoming over 160,000 new players to the game, with 2.25 million game launches the game peaked on Sunday afternoon with close to 12,000 players online concurrently.
We would like to thank all of the Sandstorm community who welcomed the new players and helped us continue to build a bigger and better community for the game. Even after the Free Weekend ended our team observed a strong increase in our number of daily active players in the week that followed with a 42% increase in daily active users compared to the end of February.
As we will go on to mention later in this post our team is working hard on improving the first time player experience to improve retention rates as well as increasing the average session duration to boost up the number of concurrent players online. As with the original Insurgency we expect to see our player metrics continue to grow over time as we add more content and features to the game through our series of free content updates and we thank everyone who either welcomed new players or helped spread the word of the free trial period.
I’d also like to give a special shoutout to our fantastic Sandstorm Partners who helped grow our hours watched and average viewership on Twitch by over 500% compared to the previous 30 days, and increase the number of total broadcasters and hours broadcast by 70 – 80%. If you are a video or livestream content creator and want to join our partnership program you can apply at sandstorm.game/partners.
Fixing spawn issue on Frontline and Power Plant exploit fixes
I’ve invited our Level Design Director, Jeroen van Werkhoven, to provide some commentary on adjustments being made to correct spawn issues experienced on Frontline, as well as some of the Power Plant exploit fixes.
Hi everyone! Over the past several months, we had an issue with Frontline spawning players under the map. It was a strange issue where we thought at some point it was fixed, but under infrequent circumstances, the bug still occurred. We found out, through isolating the problem and establishing reproduction steps, that if a player dies right after the last objective has been captured and the player respawns; the player will respawn at the origin of the map. Once we found the reproduction steps, our programming team was able to fix the issue. Moving forward, we will keep monitoring the game and our forums to make sure the problem no longer occurs.
Level exploits are a severe threat to the fun factor and balance of a level. As a level design team, we always try to be proactive and fix as many exploits as possible before a map gets released into the “live” version of the game. For Power Plant, our most recently released map, we did several QA passes to look for exploits specifically, and used our knowledge from previously created maps to find critical exploits faster. Using this method, we were able to release Power Plant with fewer exploits than any previously released map. There were still a few issues we had to solve in a hotfix, which we will keep learning from, so that we can do an even better job moving forward.
Potential improvement for SKMM timetable
Since the release of Insurgency: Sandstorm there has been a lingering issue in the code base that handled our Skeletal Mesh Merge (SKMM) process. SKMM is a process that merges multiple Skeletal Meshes (different 3D assets) into a single Skeletal Mesh using code. This reduces the rendering cost resulting in improved performance. The issue in this system was causing crashes for players. One of the greatest frustrations for our development team is the inability to easily replicate and address the root cause of these issues. With the 1.6 update, we included additional tools in the game that help our developers understand how and when the SKMM issue was occurring. With each subsequent patch, we have included incremental refinement of the system to reduce and prevent occurrences of the issue. Our goal is to eliminate occurrences of these issues within the 1.7 update cycle.
With every addition of content to the game, we do our best to review all combinations of cosmetics to ensure they appear visually appealing with limited clipping (one object appearing to stick through another). With any game engine, there are some limitations in what can and can’t be done to reduce or eliminate these visual issues. After the 1.6 update was released, our team did an additional review on cosmetics to highlight some of the more prominent instances of cosmetics clipping. Some of these were included in the 1.6.2 patch, notably a few issues with the Neck Gaiter High security cosmetic.
Additionally, we’ve fixed the Tavor 7’s alignment issue with the 1x Kobra optic. There is however a small issue where the weapon’s recoil can cause a clipping issue. We’ll try to address this in a future update.
Smoke Visibility Issues
There are three distinct smoke systems in Insurgency: Sandstorm. One for Fire Support (Smoke and Chemical Mortars), one for smoke grenades, and one for under-barrel-fired smoke grenades. While these systems all use the similar looking assets, the way they are displayed, managed, and removed from the screen is slightly different in each case. In the 1.6 update, we made adjustments to the Insurgent Smoke Mortar and Chemical Mortar Fire Support to increase the duration smoke would stay on screen. This included some minor adjustments to the particles to make them work properly with the longer duration.
Since this update, we have seen reports from players about two issues: players inside smoke clouds being able to see players outside the smoke and smoke disappearing prematurely.
While these issues both impact smoke, they have different root causes. The particles we use for smoke are larger, which creates better looking smoke effects; however, a limitation we’ve run into with the Unreal Engine is that these particles are forced to rotate based on their point of origin. That means that when a player stands at the edge of the particles and rotates their cameras, it’s possible to make the particles rotate so they are not visible to the player. This effect is only possible for players inside the smoke, which is why players outside the smoke do not see it vanish when rotating. To combat this problem in the 1.6.2 version of the game, we added an additional emitter to the artillery and mortar particle systems to help obstruct the view of players who are within the perimeter of the smoke. This will reduce, but not eliminate the issue.
To completely fix visibility of players inside smoke, we will need to implement a smoke overlay. This is a post process overlay that would cover the screen of those contained within the smoke’s volume. Once implemented, players will have a consistently obstructed view while inside smoke regardless of which direction they rotate their camera.
The issue of smoke prematurely disappearing has only recently been diagnosed. It is an edge case that has become more prominent due to the evolving nature of our gameplay. In rare instances, it’s possible for smoke particles to be despawned to prevent too many particles from being created at the same time. We’re currently exploring the option of exempting all forms of smoke from this system and thoroughly testing to ensure that the change doesn’t have any unintended consequences.
We’ve added a proper implementation of an overlay for players in smoke clouds and exemption from our particle cleanup system for smoke to our development backlog. These features will likely be released with our 1.7 update.
Over the past several months, we’ve consistently received negative feedback about the state of the tutorial in our game. The wide range of issues, ranging from stability problems, to bugs and dysfunction, to the length, to the low completion rates of the tutorial, have been heard loud and clear. I’ve invited our Lead Game Designer on Sandstorm Michael Tsarouhas, to provide some detail on our planned tutorial overhaul.
Hi guys, I’m Michael Tsarouhas, also known as Mikee, the lead game designer on Sandstorm. Having a strong tutorial for a hardcore game is very important, and so we’ll be taking some time to improve ours. This should go a long way in onboarding new players, especially when it comes to consoles, sales, and free weekends where a player might be coming in with zero knowledge of what Insurgency is.
In our redesign, we’ve made a great effort to simplify, fill in gaps of knowledge, and fix bugs. To start, we’ve removed the second half of the level, meaning the Tutorial Level ends after the weapon cache is destroyed. The late level fire support mechanic section was felt to be too advanced and irrelevant for a first time player, and should instead be learned during gameplay. The ending vehicle section that follows was felt to be mechanically intuitive enough to not need explanation. We’ve also removed all enemy AI in the Tutorial Level, and replaced them with stationary targets. This ensures a new player can get familiar with Sandstorm’s special and unconventional features like free-aim, the lack of a crosshair, magazine retention reloads, etc without the stress of getting shot at. We also added a grenade section, since not every FPS out there nowadays (especially on console) treats explosives as separate weapons as we do.
Refactoring the Progression System for New Players
Furthering our efforts at improving our new player experience, we’re making some fairly substantial adjustments to the way classes and weapons are accessed. I’ve asked Mikee to provide some context and flavor, as well as to describe the rationale behind our approach to these adjustments from the design perspective.
Outside of the tutorial, we are planning to set player level limits on classes in an effort to improve the first time player experience and address feedback on Commander and Observer misuse. Rifleman will be the first class unlocked by default at level 0, followed shortly afterward by the Breacher at level 2 which in our current projection should take about 15 minutes of gameplay. Classes will slowly start to take more time to unlock, and the classes unlocked will become more and more specialized as you go along, with Observer and Commander coming at the end. By around level 20 (roughly 9 to 12 hours of gameplay in any official matchmaking mode) a player will have all 8 classes unlocked. The goal with putting Observer and Commander at the very end of this track is to mitigate new players taking up these slots and not playing their roles properly, and to make sure they’re only open to players who are at least mildly experienced. If you’re at or above level 20, this change won’t affect you. This feature will be able to be switched on and off on community servers if you prefer to have all classes available regardless of player level. Rifleman will be given infinite slots in Hardcore Co-op in line with this change, and Competitive classes will be unaffected.
We acknowledge this is a pretty significant change, and we don’t make it lightly. We want to be clear and honest with our community about why we are doing this. This change is primarily directed at onboarding new players better and preventing misuse of limited slot specialized classes like Commander and Observer. Right now we get a lot of feedback about how Insurgency: Sandstorm is difficult to learn and get adjusted to. We feel a large part of this is because of the amount of information overload players get. In addition to all of our esoteric game mechanics like Fire Support and those mentioned in the tutorial blurb above, we have a total absence of player progression, save for cosmetic unlocks. Absolutely all gameplay content in the game is unlocked from the start, including classes, weapons, weapon upgrades, explosives, etc. On the one hand this is great because it allows for easy experimentation and prevents unfair advantages between players if some meta weapon or equipment is only available after a couple dozen hours of gameplay. On the other hand, being a new player and seeing not 4 or 5, but 8 different classes with dozens of weapons, weapon upgrades, and other equipment can be very overwhelming and create a lot of choice anxiety. It could also lead to a player making a choice that isn’t a good first time experience in a game like Sandstorm. At events like E3, Gamescom, and PAX, we’ve seen players hop on the computer, go straight to the Gunner class, and try to run and gun all over the map with an MG3 like it’s a much faster paced FPS, only to get immediately domed and confused at why weapon recoil in the game seems so unforgiving. By easing players in over a relatively short period of 10 hours, we can better ensure they take their time at learning each class, try out each class’s weapons and weapon upgrades (all of which are still available per class and don’t need to be unlocked), and don’t stumble into Commander and play without ever once drawing their binoculars to the dismay of their team.
Discussion about paid DLC
Over the past 10 years we’ve realized that our strategy of free updating and extended game support has been a pretty big differentiator for NWI. When we looked more closely at this approach, we realized it has compromised our company’s ability to stay independent as we consider bigger projects with larger budgets. Therefore, we have decided to make a minor pivot in our philosophy and introduce some cosmetic-only paid DLC into Sandstorm that will be oriented to improving customization. For those of you that want a more personalized experience and a unique look in Sandstorm, our in-game purchases will offer you the option to build your characters in special ways you couldn’t otherwise do with regular cosmetics. You’ll be able to customize your appearance in a way that will be more distinct and more enjoyable, and also (in time) be able to apply custom-made weapon skin. Why are we doing this? The bottom line is we want to be able to achieve two outcomes. The first is extending the life of our game and being able to financially manage ongoing support of the game for all of you. The second reason being that it is critical for us to explore how to best incorporate recurrent revenue as a means of removing stress from the development team, as well as to ensure we can create the types of content we currently aren’t able to provide with our team size.
Looking ahead to Update 1.7
Finally to wrap things up, now that the dust has settled on our Power Plant update we wanted to give you a little overview of what is to come in the Summer. The next update will officially be NIGHTFALL!
Thank you for supporting us and helping us to continue building great games that you all love to play.