#you have a certain amount of points you can sink into two EV stats but we're not going to actually show you the numbers or anything | Explore Tumblr Posts and Blogs

cellerityweb · 6 years

Text

World of Tanks – Fixing the Core

With Update 9.18 World of Tanks experienced some major changes. Listening to the community the team improved not only the matchmaking system, but also took a necessary look at light tanks and artillery.

Months upon months of work on World of Tanks’ fundamental issues resulted in Update 9.18, the largest and most significant update the game has seen to date. With it, Wargaming improved the matchmaker and reshaped SPGs (Self Propelled Guns, or artillery – short: arty) into long-range support fire vehicles through the introduction of the stun effect1, while also stretching light tank branches to Tier X, which ended the frustrating ±3 tier split for them. Acting upon the lessons learnt, they supported the release with massive coverage on the website and forums, walking players through every change 9.18 introduced and explaining them in detail. On its way to production servers, the update went through the total of five Sandbox and three Common tests and was released in North American, two weeks ahead of the global launch. What was the reason for this multi-stage cautiousness? Did it pay off? Slava Makarov and Andrey Biletskiy join Thaine Lyman to dwell on the road to 9.18 and the improvements made following its release.

Open Discussion

Matchmaker and SPGs called for the most thorough investigation and meticulous approach, with multiple rounds of testing. We made a few rushed changes in the past and the last thing we wanted with 9.18 was to shoot ourselves in the foot again. We were taking a huge risk building something that would be totally different from what our audience is used to and we knew we’d need to continually experiment. So, we started working in Sandbox, nearly a year before the 9.18 launch, to take a few cracks at each feature and ensure we delivered a set of changes we believe in and players feel good about.

Thanks to the changes within the tech trees of light tanks gave their popularity a massive push. Two weeks after Update 9.18 almost everyone was playing SPGs or artillery, seriously messing up viable player data.

Before we even got down to fixing the game’s core, a big part of getting ready for it was reaching out to the community to discuss problems. We needed to establish a clear and honest dialogue on every single issue they brought to us. Whether we had a solution, were in the process of designing one, or didn’t consider a certain concern a problem at all, we shared the next steps and our rationale with players.

Take skill-based matchmaking, for example. A few people wanted it, which didn’t coincide with the vision the dev team has. Instead of bluntly saying »no« or ignoring their wish altogether, we were pretty clear on why we don’t see it as beneficial to the overall experience and we didn’t stop there. We created Ranked Battles for players who were longing for gameplay built around individual performance. It’s nowhere near skill-based matchmaking but it doesn’t disturb the game for those who favour classic Random Battles and lets those who want to prove their mettle against players of a comparable level.

Players who have no interest in participating in Ranked Battles can still engage regular matches, just like bfore Update 9.

Players have been overwhelmingly positive about this two-way dialogue. They’ve been welcoming it, and so have we, no matter the feedback we receive. When we fail epically, they let us have both barrels but that’s okay, because they’re passionate about the game. World of Tanks matters a lot to an incredible amount of people and there’s a lot of responsibility that comes with it. The fact that there’s this kind of passion is not something we’d ever be upset about. We’re grateful for it.

When the time came to talk about fundamental fixes, we didn’t spare any details on the changes, why we were making them and the role players have in fixing the game’s core. We were open that if players hated a certain idea and had solid reasoning to back how they felt, we would make adjustments to the initial concept. We’re making them all the time now. The 1st Sandbox was a huge success because it saved us from going down the wrong road. We received player feedback early enough to stop ideas that didn’t click, handpick pieces players liked and move forward with them. If it weren’t for Sandbox, we wouldn’t be able to make bold moves. We would only be acting on small things we were dead certain about. But that’s not how you make deeper structural changes that actually fix the holes in the boat.

Release

Matchmaking and arty have been a source of heated discussion for years now and it took us several iterations of testing to transform them into something we were happy to put into players’ hands. After a good five rounds of testing, we arrived on a batch of tweaks the Sandbox community felt worked and went on to include them into 9.18. It was the first time we were releasing an update in North America ahead of the CIS.

Update 9.18 finally stretched light tank tech tree branches to Tier X, thereby ending the frustrating +3 tier split.

Communication before the Common test was huge. 9.18 became the single largest update the game has seen since launch. A lot was about to change and besides ensuring the version worked, we needed to prepare players for what was to come. We went to great lengths to ensure everyone was on the same page as to what the update would bring, how it was going to address problems with matchmaker and arty and why we were choosing this particular way to fix issues. It was just as crucial to explain that we weren’t going to get the matchmaker, arty, and lights tanks right in one update.

North America (NA) has a good combination of high PCCU and low PCCU, creating a perfect ecosystem for us to test the update. When it went live there, we all started playing alongside the local community to see how it worked and how they reacted to the changes. The first two weeks were completely mad! Ev

eryone was playing light tanks and arty, which messed with the stats a lot. In fact, the influx of SPGs and light tanks was so huge that every rule of the matchmaker failed. It wasn’t until the hype died down a little and players got back to normal patterns that relevant data started coming in.

Results and Takeaways

The biggest complaint players had about light tanks was that we actually made changes from the Common test to the release version, decreasing down their effectiveness. If you look at the state of the game to

day, there are more people than ever before playing light tanks. Their balance overall is pretty good but we’re keeping an eye on it. Players are still learning how these things work, learning the pros and cons and the way they interact with light tanks and SPGs a month into the update will differ a lot from the way they’ll be playing them 3 to 4 months later.

The first thing we knew we’d have to fix were arty Platoons2. Stopping SPGs from forming and joining them was a tricky decision, which resulted in increased forum chatter. We lifted the restriction limiting their number to just one per Platoon in 9.19.1 as we still have a strong belief that limiting the number of SPGs per Platoon is a good idea. If a Platoon of three SPGs decides to focus fire on one player, it would be a miserable experience for them, a lot like being trapped permanently. You’re just as dead as if you were one-shotted3. It’s merely a question of how long it will take to destroy your tank. Two-arty Platoons are no better. Allowing one SPG per Platoon was OK, but at that point we wanted to be conservative about it. We needed to do this to measure the effect of the arty revision. Once we did that, we acted upon player requests and brought them back. Finally, the matchmaker was almost universally appreciated as a step forward. There were some legitimate complaints that we are either acting upon or will be acting upon in future updates.

1 arty per platoon only: Groups of SPGs were way too powerful, so the team had to limit the number of artillery in battle to avoid player frustration.

Is the stun effect perfect? Is the improved matchmaker working just fine? No, there’s still plenty of work to be done in regard to both and we’re on it. Most importantly, we’re open to players and have it in ourselves to say that we know about the areas that require fine-tuning. We’re working on them and even if we haven’t found a sound solution to a particular issue just yet, we’re looking for it. Moving forward and up until we all – devs and players alike – are certain that the matchmaker, arty and light tanks are where they should be, every single patch will introduce a few improvements to these three. We’re past fixes, but there are a few improvements waiting to be made.

About the Authors:

Andrey Biletskiy

is Creative Director for World of Tanks at Wargaming

Andrey’s job is developing a detailed vision for the game and making sure various teams deliver on that vision. He’s been with Wargaming since 2014; before that he spent six years at Ubisoft and contributed to the Assassin’s Creed franchise. His motto is »Every mountain is unclimbable until someone climbs it; every ship is unsinkable until it sinks«.

Slava Makarov

is Product Director for World of Tanks at Wargaming

Slava has been with the project since its inception in 2008. He performed many roles, always staying one of key visionaries in company and a living game design encyclopedia. As the Product Director of World of Tanks he oversees the entirety of World of Tanks development and operations.

Thaine Lyman

is World of Tanks Product Director at Wargaming

Thaine joined Wargaming in 2015. Before that, he spent over 10 years at Activision where he worked as Executive Producer for the »Call of Duty« series and as VP of Production. In his current role, he works with the executive team, development and publishing teams, regional GMs, and operational teams to ensure World of Tanks provides a best-in-class experience.

The post World of Tanks – Fixing the Core appeared first on Making Games.

World of Tanks – Fixing the Core published first on https://thetruthspypage.tumblr.com/

0 notes