August 15th Livestream Transcription
– Steven appears as a guest on the DCN 100th podcast (Congratulations!) for an in-depth interview on Ashes of Creation’s development.
Q: Since Intrepid has partnered with Steam, what kind of benefit do you recieve and will we see the MMORPG potentially on steam down the road?
A: We never planned to be on steam with any iteration of the MMORPG, but as you know, A1 evolved into apocalypse and its 3 modes. But as we approached our backend getting done, it made sense to hammer the servers as hard as possible, but we didnt want to devote marketing to apocalypse. So steam provides a great demographic of players and it makes sense to funnel as many players as possible to get the most amount of data, kick off matches regularly and introduce players to what ashes of creation is as an IP. This led us to change our mind with steam in regards to apocalypse. But as for the MMORPG, currently we still have no plan to do so. Platforms like Steam have a revenue split. When the MMORPG approaches its betas, we plan to ramp up marketing and are confident that we can satisfy that marketing requirement without having to utilize something like steam.
Q: Are you concerned about negative reviews on Steam for Apocalypse?
A: The MMORPG community as a whole has gone through twists and turns, so their reception is understandable. As one of those jaded people myself, I see negative comments and tend to use them as fuel. So any type of trash talk or doubt I use it as fuel.
Q: What details can you go into about the architecture changes that made you step back and redo them?
A: There were a large amount of changes to be made, ways that which Unreal does things out of the box and things we did that didnt work that well. So we want to use features of Unreal that we can, stacked with the features with our third party partners and much more to ensure the quality of the end product.
Q: Can you tell us from your internal testing what kind of increase are you seeing?
A: Regarding the test on the 20th, we have the matches designed for a 100 players. Originally server framerate was operating at 12-20fps at 100 players. This isnt good from a performance standpoint, rubberbandy and high latency. Right now with the internal and outsourced testing, we’ve had 100 players with server framerate at a steady 60fps which is its cap. This is an extreme backend change. Additionally this isnt about pushing 100 players, we have castle siege coming up and we use internal tools such as bots to test things, the bots use attacks and jump around and stuff. Last test we had over 350 bots and over 30fps, once we begin optimizing this using the data that comes in, we can take this data and start trimming the fat. So players that are only relevant to you based on direction etc. all of those things become logic that increase performance from the network and player standpoint.
Q: Are you looking at capping the fps at 60 for the MMO?
A: Im talking about the server framerate, not the client render framerate. Client framerate is dealt via the GPU with buildings/textures etc. We want the servers to transmit at a higher rate, what’s the position of players, did you score a hit, all these things need to be processed by the server. So we’re optimizing the backend. Our last benchmark was 80fps at EPIC settings with a GTX 960. Apocalypse lets us optimize all these graphical details, optimize the streaming distance, our HLOD and LOD systems, all of these things go into providing a better experience for players.
Q: Since you’re using AWS, have you used bots from outside your geographic region?
A: Yes we use bots but the way we spin up bots is we take virtual machines on the AWS and can create 50-100 bots with C5 or C4 group machines, we spin up thousands of those and throw them into our matchmaking process and balancing etc. This gives us results that allow us to tune things.
Q: What can you tell us about the server tickrate?
A: So there’s multiple tickrates. Im no engineer, but some tickrates have a very high priority and run every frame like movement. Movement is one of the heaviest tickrates that takes priority. Right now in apoc its 350 meters and weapons go to 275 meters. So movement is the highest cost when it comes to that. Tickrates vary really, depends on what we’re talking about. (We use just found out the longbow has a tickrate of 275meters)
Q: What % done is the apocalypse BR excluding bugs?
A: There’s different departments, from a design perspective its about 98% complete, from environment background its 98% complete, from a character perspective 95% complete and from an engineering perspective 90% complete. Things are going to need to change post live operations as we’ll see data trickle in and we’ll adjust our code to learn from it and move forward. So yeah the BR is relatively complete.
Q: With the BR around the corner, what excites you about the changes to the BR?
A: I am not really a BR player, however when it came to gathering data the fastest way, we went with the BR and when we QA the BR, players seem to have a lot of fun with the fantasy aspect of spells and effects. For me its the ability to see the world and what is coming, to see the immersive nature and various cultures, understanding some of the lore, these are exciting to me as I like to be story driven as a player. So that’s what excites me most as a player, but as a developer the performance and metrics we gain are great to see as we can see the progression of a product via feedback, data collection and refinement. The cool thing is as we step forward we get to take everything we learned and move it there as well. So you get a better experience in Betas due to the learning experiences in Apocalypse.
Q: From an automated bot perspective, is that something you vendored out or developed yourself?
A: We have tool engineers, they run varying degrees of tests from load balancing to simulating player packets, network traffic, website features, all of those things are tests that we get to permeate. Our tools engineers after December 2018 (when we didnt have those tools in place), we were so excited to get live data from the community that we were a bit hasty back in december. When we moved forward as a Studio, we were able to course correct and this contributes to better serve the MMORPG development.
Q: You guys have this dev discussion series on the forums, has any of the player responses affected development in any way? If so, how?
A: No we’re not getting rid of elves. With regards to the dev diaries, we take all feedback into consideration. Where the feedback tends to shine is where post-discussion, Lt.Toast circulates that around the team and they have a look at the stuff and some of it comes into the round table discussions. So insight is gathered from the player input.
Q: So regarding mayors running cities, can mayors dedicate tasks to players or does the mayor have to run it solo.
A: So the mayor doesnt have the ability to delegate roles. But there are roles that become available that are achieved via title and rank progression. So achieving high priest of the temple or head merchant and achieving this gives them certain duties. So these titles come with roles in sieges that give you buffs and additional skills. So world events where nodes advance or grand temples being built, when they trigger events such as a dragon attack or zombie invasion, these title players become more heroic and may gain additional damage mitigation or additional damage or battle cries to motivate the citizens around them. Those are some of the ways we approach additional roles for players that arent necessarily the mayor. In the game we want to have a separation of state and guild. We dont want guilds to control nodes, guilds can fight for castles, guild halls, guild wars. The node system is focused around the inter-guild community. We want players to be adversaries or friends through not just the guild system.
Q: Are the sub jobs also differently selected like node leaders or some other way?
A: There are many quests that revolve around advancements or performance which dictates the leader.
Q: If the mayor is elected and you have other groups that have leadership in different tasks, do they have checks and balances that work against the mayor or does tha mayor have defacto balance?
A: There are some give or take roles that exist within the node system. This simulates what is the politics of the mayor, the mayor has to build relations with these title players to progress certain systems. So if there’s non-agreement between players present, it might stifle growth. So when different parties control different branches, you can halt progress which might be a good thing. A stop gap that requires compromise. That’s the type of functionality we want to duplicate in game.
Q: Are there any tools to help me fight a mutiny within my own guild? Or these different positions within the cities, are there a set number of them? Are they locked by node type?
A: The title positions in a node are tied to 2 things, the development level of a node – which dictates the possibility to build certain buildings which when built bring certain titles into play. So this is where ultimate power lies with the mayor as they choose which buildings go up. So a guild could come to a mayor and say they will help build a temple and the mayor might agree and ask for their support in return so that whoever is given the title as a result of the temple being built, that they work with me (the mayor).
Q: Early on you mentioned the ability to track and record data as a foundation of the game, are there plans to share that with the players like in the library or character tab? To see what things the player is done so we can decide whether to recruit them or not.
A: We do have some grand plans that may not be live at launch but will come online post-launch, such as a character profile system that they can turn on to boast or let others see, linked into the website and linked into the mobile companion in the future. There’s a desire to do so, yes, but it takes a backseat to the more fundamental aspects of launch. Regarding the mobile app – We’ve discussed the ability for players to manage their property/processing/node activities/guild features via the mobile companion app. It’s not a “Dont you have phones” its more of a “Hey this is a great tool for when you cant be at your computer.”
Q: Will the website profile you’re considering, will it show gear as well?
A: We have to be careful about showing gear and some of that stuff because those things can partition communities. That’s not to say the partition cannot happen, so guild leaders can absolutely ask people to send a screenshot, but what we want to encourage as a whole is for people to play together for mutual benefit and progression. There’s lots of types of people you wouldnt meet if it wasnt for an MMORPG, so we dont want to make the divide greater between players. Kinda the reason I made an MMORPG, to stick to my ideals. If it works great, but if it doesnt we’ll take the feedback and re-assess.
Q: Can you give us a leak for the 100th episode?
A: Maybe I can give it to you over Discord, one moment. So you could show the concept first and then the final product.
Q: As Steven is coming from a different background in the gaming industry – What’s a time when Jeff/Margaret had to stop Steven’s idea?
A: Very early on – I catch on pretty quickly – the stupidest thing I had thought of was the hotmic that happened at one of the PAX events. That was Bacon’s fault though, it was totally his fault! Lets just say there was another company talking some not so good things about me and I was like “man this company is talking some mad stuff about me” and bacon was like “uhhhh, we’re live” and I was like “WHAT”.
Q: Since this is your first time developing an MMORPG, what has been the biggest lesson so far?
A: If I could do it all over from start to finish, I would have a serious contemplation of whether I would have revealed it early on or whether I would reveal it closer to the beta. My biggest contemplation has been whether that had been the right decision. I wanted to take everyone with me on this journey. Being a developer, its a harsh reality of what software development is in general. Its not the fault of the players, how can they be expected to know etc. It can be a bit demotivating and public relations. And when something changes or needs to change, you dont have to be a complete open book. So there’s an understanding that things can change in that regard, so that’s one thing looking back I’d have thought about.
Regarding negative comments, I have a thick skin. But not everyone on my team has a thick skin. People take pride in their work, so it can be demotivating to watch a video where someone who has 0 clue about things is being negative about it. But I will say that I love the Ashes community, taking part in this living breathing project.
Q: Will the livestreams ever go back to talking while running around in the MMO world?
A: When the MMO world is ready to be run around in a way where the people looking at it for the first time will understand what they’re looking it, absolutely. We just need to be careful about it.
Q: What are the working conditions like at Intrepid? We’ve seen a lot about developers not doing so well.
A: There’s definitely truth behind the game industry having a systemic problem with. We’ve had times where we have had to crunch in the past like unexpected delays or unexpected opportunities. So we’ve had those in the past but one of the fundamental processes is having a production team that works on this to ensure from a expectation and production standpoint, it works right. We try to move forward avoiding things such as the crunch, if we do encounter it, we try to make it pleasant like working from home or providing dinners for the team and things to cushion those periods of times when they come up. So the culture at the studio is a great one for them to express what excites them. We’re a very collaborative studio with a desire to push creativity. Like when a milestone or feature is coming up, we deal with it with responsible scheduling to ease things out.
Q: Steven has spoken about objective based guild wars besides castle sieges
A: So objective based guild wars revolve around what types of actions that one of the guilds has participated in, so there’s default objectives and objectives relating to what a guild has had. So lets say a guild has recently obtained a guild hall on a freehold, it might be near that freehold, or it might be around the guild hall that might need a channeling time, if the guild has recently done a raid, there might be objectives to achieve one of the quest items that the guild has achieved and steal it from them or a guild member might have a bounty on him with special buffs to hunt them down and the player can call for help etc. We havent worked on it yet, we’ve worked on the castle siege stuff, but there’s a lot of ideas/variety of ways we can make it work.
Q: So lets assume they’re bringing back Archeage Classic with subscription
A: I used to speak to the AA producers and used to tell them “if you did a sub based no p2w server, I guarantee you it would make more money over the long run than the flash p2w fresh start server tactics” I was also talking to one of the investors and told them that they could give me the rights and we could see which performed better, the sub one or the P2W one. Archeage had a lot of great potential, I dont like faction based games. I like to decide friends/enemies on my own. I did like the risk that AA had with pack creation etc. but didnt like that there was no real feeling of loss and summoning ships out onto the ocean etc. But there was a lot more I enjoyed. If the new AA does it the right way, it’ll be a great thing to see, especially if the industry slowly moves back to the sub monetization model. It’d be cool if there was a resurgence of what makes mmo such a unique game, there’s a reason mmos are expensive to make, its because they retain players long term which is better for the sub model.
Giveaway Winner: Varkun56