Precursor is coming…
After a many years hiatus, some time in the service industry, doing some contract serious games development and now halfway through my computer science degree, and of course, staying up to date on the latest game tech, it’s become abundantly clear the iron has never been more hot to strike…
Browser render and physics stress test
It’s going to be awesome…
So we’re making a universally released (first targeting Browser, Windows, Mac and Linux) top down, instanced, mutiplayer-roguelite. Current server size is to be 20 players at 4km x 4km, servers initial conditions will be procedural (some of which behind the scenes will dictate what potential end states are possible). Like the player characters the servers die too and core gameplay loop will consist of experimenting with what potential raw resources the island has for smelting, cooking, and herbalism. As well as the type of creatures for skinning or NPC races to interact and, ahem, “wrestle” with, to unlock their genes as a new starting race for your next character.
Gameplay View Example
It took a lot of planning.
Sounds insane? It’s taken many more years of research than execution at this point, and will probably indefinitely stay that way. As it stands we’re done prototyping the rendering and input end and have our Amazon Web Servers booked for starting internal mechanic testing on the network. This is the overwhelming majority of work on this project now with all the client side rendering optimizations done. From 17 dismember-able body parts and 24 gear components to 3 skinned mesh renders at runtime, never mind the years long nightmare of deciding if working with the Unity cloth was worth unlocking it’s mysterious ways.
Early Cloth and Character Test
The Model: Mechanically and Beyond
The system will use an authoritative model that has built in tertiary animation that gives instant local feedback (like character leaning or pre-server state, minor local animations) while also keeping all the mechanics on the 20 tick server. If Amazon’s new gamelift is as amazing as being presented, we should have no problem integrating skill-shot based twinstick combat that does not use any predictive modeling (potentially one frame local max, 0.05s), and only introduces noticeable input delay or lag spikes on the users end, preventing any laggers compensation on the server end and in so far as we keep everything skill based, any sort of client based predictive combat hacking would be detrimental against any low-mid skilled player and I imagine it’s quite an undertaking to hack AWS on the back end for any game breaking hacks.
If all of this sounds particularly ridiculous to you, than I highly recommend you go check out our current (near finished rendering end) web build with our fully body layered cloth physics, or if you have a hard time getting your dGPU working in browser, or just want to double+ your fps, give the small client a download and enjoy!
Screenshot of current performance test in browser:
Monetization – Moneh, Monai, Money?
And last but not least, the monetization, this is so important to me I don’t think it’s legally safe for me to describe what I said I’d rather do than sell out with this game, so without predicting the future, I will fully explain our perspective and concerns, monetarily and logistically with this project.
Firstly, there will be no gameplay impacting purchases, there will be no gameplay aesthetic purchases. What is left to sell and what is the latter? Since every material in the game is explicitly sourced from the world, the type of wood impacts the end color and stats of armor, the raw ore quantity and ratios, the types of creatures to skin and flowers to create potions from… needless to say the visuals are explicitly married, like many things else in this game, to the core mechanics. So what ‘aesthetics’ do we plan on selling that aren’t at all gameplay impacting. The boats! That the player has at the main menu and floats into the island on, we were planning on adding achievement based boat cosmetics anyway, and since you spawn on the least inhabited portion of the island, there is a degree of gamifying showing off your boat that is the furthest thing from force-fed to free to play players.
Runtime Vertex Color Adjustment
So then how can we even cover server costs and is there anything of objective value that does not ruin the game players can purchase? Yes! Although you can at any point share a server ID with your friend for them to come join if the server is not full, if you for example wanted to get 20 people together on a ‘private’ server that in no way provided a direct mechanical advantage but to explicitly attempt PvE gameplay end states with a lot of friends. In this case, an as cheap as possible subscriber model, ~$5USD/m will serve perfectly (with some server creation limits for balance and cost set by the central server).
Free to Play?!?
And lastly what makes the free to play model potentially viable is somewhere between subscribers covering cost and offsetting the cost with… advertising. It is very important to me this not bastardize the artistic vision or atmosphere at all. To this end, I think the only option is scale ad time to gameplay time, and only serve ads when a player dies and cannot rejoin the same server (end game case), as such, when loading a new server to start a character on a world to which they had no prior impact or care, they would be served a minute ad to each hour of gameplay (or whatever scales best to keep the model alive).
Per vertex + Toon Masking Vertex Emission
As we move forward with prototyping with the AWS gamelift servers, we will find out what works best given the model and what is economically viable for us to create a genuinely valuable game as a service/pseudo MMO while NEVER whale hunting or altering the core mechanics and visuals for monetization. Worst case scenario we drop the free to play model for a cheaper base subscription, we will see though, and I fully intend to be completely transparent with the monetary design decisions, as with all, moving forward.
Be sure to follow us for developments from here on out, our PR guy, Mango, will be doing some of the updates, be nice! He will be primarily focusing on audience driven engagement while I will be focusing more on developer diary’s explaining some of the things we’ve done and how we done did them.