Lessons from Crypto Native Games

[Music] [Applause] [Music] um cool so thanks again to uh eath global for having me for this year's ethonline i'm gub sheep um i am one of the uh creators of dark forest and today i'm going to be presenting on lessons from fully on chain gaming um so first for a little bit of context uh about what my team works on we're building dark force which is a procedurally generated mmo real-time strategy built with zk snarks for the evm so currently dark forest runs on xdi previously it's run on the robston test network it's a game that can be launched onto any evm compatible blockchain and basically what you can how you can think of dark forest is it is a it is a game that is fully on chain so all of the data um all of the state all of the game logic is on chain can be deployed by anybody um and as such it is a quote unquote infinitely hackable strategy game uh where players can write scripts or interoper interoperable contracts or any other modifications that they might desire during the course of gameplay so yeah like we mentioned all the logic runs on the evm um formerly we ran on robson now we run on xti during dark forest rounds we generally see about we have a white list system and we generally see about 2 500 players or so during these rounds and uh over the course of the weeks that we run these monthly rounds we'll see these players consume more gas than all of ethereum mainnet combined actually um so you know another impetus for for scalability if we haven't seen enough arguments yet already today and a little bit about myself i'm guv sheep i work on dark forest more broadly i work on applied zero knowledge crypto research and development and i also work on a number of educational and eco dev projects okay so to dive into crypto native games we first have to answer the question of why build games on the blockchain at all and i'm sure many people here are probably familiar with this experience but when i talk about dapps in general or games on blockchains or whatever other whatever other applications people today are interested in building on blockchains the thing that i always want to be wary of and the thing that i see in a lot of projects is that basically a decentralized version of something is oftentimes just like a centralized version but worse and we really want to lean into the unique affordances of decentralization without just taking purely many of the obvious disadvantages that come with building on a blockchain you know blockchains are slow they're uh difficult to program on they are they have certain scalability limits um all of these different things so uh make an argument for why to build why we care about building games on the blockchain and this argument sort of has three parts so i think that there's three answers that we've seen over the last couple of years to why crypto games are interesting or important there's a broke answer there's a woke answer and there is finally a bespoke answer so the broke answer is that people are interested in building games on the blockchain because it is easier to build ponzi schemes or scams um so i think in a lot of the early days of ethereum we saw applications like you know if people are familiar with fomo 3d from 2017-2018 kind of the last bull run um fomo 3d was sort of a tongue-in-cheek pyramid scheme slash exit scam application where people would basically put in ether into a big pot and every time you put ether into this pot you know a timer would uh increase and the timer is ticking down and if the timer hits zero i.e like you know no one has put in any contribution in the last like five blocks or something like that then the whole pot would go to the last person who had put ether into the pot um and i think that these sorts of quote unquote games are what characterize a lot of the early experimentation around crypto games um and fomo 3d was in my opinion one of one of the better ones honestly but a lot of these things were just like straight up like pyramid schemes or scams or whatever else and i think it left a bad taste in a lot of people's mouths um and so you know i'm glad that the space has moved away from this sort of thing being like the dominant narrative in crypto gaming um but obviously there's a lot of lessons to be learned from this so the woke answer to why build crypto games which i think is uh more the dominant narrative today is that with crypto when you build applications you have apis into like the native concept of money or value so that means that uh it is really easy to make for example your game assets be worth real money and consequently it makes it much easier to build digital worlds where your actions have quote real or economic consequences so for example if you like uh lose a particular item or you you know manage to find a particular item or something like that then that has economic consequences for you as a player which you know kind of makes the game experience more immersive so a lot of uh collectibles projects i think fall into this category so you know i think everybody pretty much has is aware of slash has seen crypto punks um we've seen games like uh cryptokitties and you know these days axe infinity is drawing a lot of attention um and also we're seeing a lot of uh you know games like for example sky weaver or gods unchained where you've got these trading card games where gameplay happens off chain but the ownership layer of the game lives on chain so cards are erc721 tokens and ownership is recorded on chain and these can be transacted on marketplaces like openc so yeah like we mentioned these the fact that you have access to these money native apis so easily in crypto games means that you can build these games with real consequences or economic consequences um and that i think does two that's really exciting for two reasons this does two things that uh traditional games cannot uh really tap into the first is that because there's something at stake the game is more immersive um you know immersivity is a concept which you can measure along a bunch of different dynamics uh one is like you know how graphically immersive is a game um another is like uh you know does the game feel like something where you've really got something at stake um and i think crypto uniquely enables that dimension and the second thing that's really powerful is that the fact that these are economic universes means that it might be easier to incentive align the community members of a game or the players and the devs so when everybody can share in the ownership of a protocol then everybody is incentive aligned to grow the community and and build out the unfortunately i think that the current state of crypto native gaming and a lot of the games that subscribe to this thesis is incomplete um i think that a lot of crypto games currently are more speculative asset classes really than they are games um in that either the gameplay is very limited or you know the the um game itself is just not very fun or most of the attention that's being driven to these games is surrounding the financial value of the assets rather than sort of the intrinsic value of owning these assets in a digital world um so i got i saw a great tweet where basically someone claimed crypto games will never be sustainable as long as the market around the game is more fun than the game itself um and you know i think this is super true i think that another uh way in which games that subscribe to this currently dominant thesis is incomplete is that these many of these games aren't fully crypto games so for example ownership is on chain but game logic or identity or you know account logic or permissioning is off chain so avsa tweeted this out uh earlier this summer um but basically he noted that this was a really popular game that had uh banned a large bot farm and abso was sort of calling into question like wait a second is it really a crypto game is it really a blockchain game if there is someone who can ban people or if there's someone who can prevent people from playing the game i think you know it comes into conflict with a lot of the ethos around permissionlessness and censorship resistance that's so important to blockchains not just for ideological reasons but also for practical reasons so the third part of the crypto gaming thesis that i want to talk about is the bespoke reason for building crypto games i think that the most exciting reason to build crypto games is that building a game on a blockchain enables something that we call permissionlessness or sorry permissionless interoperability rather than explaining what this means abstractly i want to show a few examples from dark forest of things that we have seen players do that are uniquely possible because dark forest is a quote-unquote crypto native game so one thing that emerged very early on was that we saw players were scripting very heavily in order to gain advantage in the game and one of the interesting consequences of building a game or honestly any application on the blockchain is that blockchains are blockchain applications are client agnostic in other words you can talk to the blockchain you can hit a function in a smart contract from a script that you write from a default or vanilla web client provided by the developers from a modified client that you have forked from the developer provided client you can do any of these things you can go into your chrome dev tools and you can just you know spin up a transaction um sign it with the private key that you're storing locally or or anywhere and send that off to the blockchain access to the game cannot be gated and therefore what that means is that players can actually build out things like automations or bots or even ais as first class features into the game so early on we released a plug-in system that allowed players to essentially take snippets of javascript code and inject those snippets into their client um and from this we saw players build like a ton of interesting utilities from data visualization tools to remote snarkers which allow you to prove um these zero knowledge proofs that are at the center of dark force on web servers that are you know the user maintained and offline um we saw players writing even like very complex ais to assist them in playing the game when they were offline or you know to allow them to play at a higher level and all of these things have kind of become first-class features of the game these snippets that are passed around between the community and that's added a whole other dimension to the gameplay experience we've even seen some folks fork or write entirely from scratch new clients so for example georgios and kobe did a lot of work to put together a dark force uh rust client that allows players to play dark forest in the terminal um and there's a lot of sort of performance boosts to to being able to do this especially because you're able to prove your snarks with a different tool stack that's actually a lot more efficient um and one thing that georgios hope you had to do was actually you know do some reconciliation work between these two different stark tool stacks the first that's meant to be used in browser and the second that's being meant to that's meant to be used you know inside like a rust web server um and this was super cool to see and something that you know is much harder to achieve in a traditional game with a canonical developer provided client um another central concept to dark force is the idea of the there's a cryptographic fog of war in the universe that you uncover via something that's pretty similar to proof-of-work mining but we saw some players come out with gpu miners for playing dark forest which was also really incredible to see um and i think the step function that we saw in the last couple of months that really changed the game and really opened the community's eyes to i think what's possible with these um permissions permissionlessly interoperable crypto games is keep building new features on top of dark forest as well in a way that fundamentally is not possible with traditional centralized games so as an example like because dark forest is a smart contract anybody can deploy a smart contract an auxiliary smart contract onto the blockchain that talks to the core smart contract and uh essentially you know these can these auxiliary contracts can look thing like things like escrow contracts or marketplaces uh or you know player corporations or anything else um in this case the very first example of this that we saw from a team called project sofon was project silphon built a broadcast marketplace with which would essentially allow you to trade information with other players via a third-party contract that talked to the core dark force contract um so you can kind of imagine someone uploading a simple escrow smart contract that says like you know anybody who deposits one each into this contract will cause the contract to programmatically trigger a release of some resource from my dark forest planet to your dark forest planet um and then you know what a third-party developer can do is they can write a plug-in that essentially exposes a ui for connecting to this escrow contract um and share that plug-in around or they can provide a third-party client that other people can use and what we've essentially what we essentially see here is that like players have come in and they've written a new feature for the game without any input whatsoever from the developers contrast this with a traditional centralized game where if you wanted to add something like a trade feature you would have to there's there's pretty much no way for this to be done except through the core developers um so if jagex wants to add peer-to-peer trading in runescape then they have to write a data transformation on their proprietary you know data stores that live in their web servers then they've got to write api routes that um you know allow uh that will trigger these data transformations and then in the canonical runescape client they've got to write like a ui component that's a button that's hooked into their um you know uh apis those api routes that are running on their web servers um there's no way for player like players can do so much more in a decentralized game um compared to in the case where jagex or blizzard or whatever else owns the stack from end to end um so we also saw for example a bunch of different marketplaces pop up in the last round of dark forest where again these were entirely player-made they were not built with any input from the core devs um and we think that this is a very early example of some of the richness or complexity that we might be able to see once players can actually impact the land that they're playing on um this was taken from a little like this is like a little mock-up um from a presentation that someone in the dark forest team gave internally recently um but you could imagine a world where someday like game objectives or even like totally new resources or even nested games are plug-ins that anybody can install in the game where logic is run in these smart contracts that are maintained by the community or by players of the game itself um you know whether it's like game modes and objectives or competitions uh or you know we've got here like oil and uranium resource plug-in or something like that those are things that can live in auxiliary smart contracts outside of the core um core smart contracts deployed by the core devs um and i think the last thing that i'll tease here uh is there is another game that we have been collaborating with um coming out from ludin's on twitter who's building a zk dungeon game uh and this is something that we're really excited about because we think it's going to take a lot of these zk mechanics and interoperability mechanics to the next level so keep an eye out for dk dungeon and the luden's account on twitter so now that we've covered those three reasons for building crypto native games it's finally time to answer the question like what is a crypto native game we've been tossing that term around a lot um so let's actually get to something that looks more like a definition um to us a crypto native game is a game that is maximally on chain and that can maximally leverage the unique affordances of the blockchain so this means that the source of truth for game data is on the blockchain the game logic and the game rules are implemented via smart contracts the game is developed in accordance with open ecosystem and open source principles um the game is client agnostic and the game embraces real world digital value so we think that this is a superset of what is currently going on with a lot of uh the efforts that we're seeing in crypto gaming right now um to also provide a few negative examples things that we don't consider to be crypto native games under this definition are things like a mobile shooter game with the crypto wallet integration or like a trading card game where c7 21 nfts um but all the gameplay uh and the battles happen off chain or um and this example is kind of funny to me um especially for just looking back at d5 summer defy apps that have like a playful skin or are very gamified or have a lot of emojis you know for all of these categories of applications we think that you know this isn't to say that only crypto native games can be successful commercially or artistically but we do think that crypto native games are the most interesting and important to look at to understand the full potential of blockchain for building digital worlds and to build a picture of where the space is going so that's the what um the why is something that we've sort of answered by example from some of the previous slides just to lay it out explicitly the reason that we think crypto native gaming is is exciting is because it enables permissionless interoperability and that allows that means that players can build the game experience not just the core devs beyond just being able to build on top of uh crypto native games it also means that multiple different crypto native games can have strong interoperability mechanics between them so other games can hook into your game to share assets identity systems reputation systems mechanics and more without needing to talk to like a support team or a bd team or anything like that so for example imagine if like you know you could take your sword from world of warcraft and take it with you into the the runescape universe or something like that currently that would require some sort of you know like integrations effort between jagex and blizzard but if zk dungeon wanted to hook into the dark forest artifacts contract that's as simple as just saying like oh you know anybody can take their dark force pyramid and deposit it into their zk dungeon dungeon heart and if they do that that like powers up all of their units in their zk dungeon and that means that these crypto native game assets become almost these like real physical tangible goods that you can pick up and carry around with you um throughout the ecosystem of these crypto native worlds so you know in short there's all sorts of things that you can build when a game is crypto native from marketplaces to dows to nested games on top of or inside of games to new resources and mechanics automations alternate clients interoperable mechanics shared reputation systems um and i think that you know for all the hype around uh the m word lately um if you're looking for um what might eventually seed the metaverse i think crypto native games are an interesting place to look um you know there's still a very very long way to go to realize the full potential of permissionless interoperability in games for one thing scalability is a huge bottleneck so right now darkforce is in a private whitelisted beta where we allow a couple thousand users at a time onto the game this is for scalability reasons because current blockchains and even you know sidechains plus rollups can't support more than thousands to tens of thousands of users um how do we get to a place where crypto native gaming is can be mainstream and can support hundreds of thousands or millions of users that's an open question developer experience is an open problem uh if we care about interoperability then we care about it not in the theoretical or abstract sense we care about it in the practical sense we want people to actually be able to build these marketplaces on top without it being a huge headache so that means that there's a lot of developer tools and infrastructure that needs to be built out all sorts of these public goods that will be required to sustain this ecosystem it's also going to be important to find more developers who are able to you know program comfortably with solidity or with the dk tool stacks that are that we think are going to be important for building out a lot of these next generation apps um and there's you know so so far to go in terms of uh understanding like application usability accessibility design patterns all of these things necessary to really um make crypto native games usable for a mainstream person who isn't a crypto enthusiast as a baseline the we think that all of those things are currently being worked on and a number of them are quote-unquote just engineering problems but one of the biggest um things that we one of the biggest open problems that we think will be necessary to unlock the full potential of crypto native games is uh zk technology and more advanced crypto so dark forest itself is made possible by zk snarks because zk snarks enable um the crypt uh cryptographic fog of war which is a really central mechanic in dark forest as a strategy game um but starks are in fact a very general purpose tool that can be used to unlock all sorts of new game mechanics that weren't current that aren't currently possible on chain today um so for example this is a slide from uh a presentation we gave um a couple days ago but one thing that zk snarks can be used for is to enable rich procedural generation algorithms essentially executed or verified on chain so if you wanted to build a decentralized minecraft uh then a decentralized minecraft analog so a fully on chain procedurally generated very rich um three-dimensional sandbox world what you might do is you might start by taking a procedural generation algorithm like purlin noise and putting that inside of a zk snark um and then putting that zk snark inside of your crypto game and we actually do have in both zk dungeon and dark forest uh very primitive examples of this of procedural generation algorithms that are now implemented in zk circuits another example of a place where zk snarks are very important is in allowing um things like card games or private in private asset inventories um in mmo games so one of the tricky things about card games in decentralized settings whether that be a blockchain or a state channel or whatever else is that a lot of the operations involved in playing a card game are things that naively seem impossible on a decentralized system so for example like randomly and privately drawing a card in a way that is uncheatable seems very hard on a system where all applications have totally transparent data layers but we've been working on some constructions and we have a few proof of concept implementations out of different card game mechanics where you can essentially use zk snarks to you know privately and randomly draw cards for example from a shuffled deck this is also interesting because uh zk snarks actually allow you to go one level further than commit reveal schemes in terms of both usability as well as what is fundamentally possible but yeah all of this basically points to you know we believe that games are part of a larger ecosystem and that games are actually where we're likely to find the frontier of application research and development for those of you who saw vitalik's talk at ecc a couple of months ago um you know one thing that vitalik mentioned was that uh today you know sort of the the major production use case of ethereum is finance um and a lot of times you know we're seeing that a lot of the application level work being done on ethereum or these decentralized systems uh is applications that are simply iterating on or permuting different um mechanics or primitives that we already know how to use so i think the the example that vitalik gave was something like you know people are building tokens on yields that allow you to you know tokenize the yield captured by yield tokens or whatever um and we think that like to really unlock the next generation of ethereum applications whether that be games or coordination mechanisms or dows or whatever else um we're going to need this this sort of new technology that's going to allow us to fundamentally move the needle on what is possible to do inside decentralized applications um so yeah that's pretty much it for this talk uh if you are interested in getting in touch um dark force is hiring you can check out a hiring announcement on the dark force blog one thing that we're going to announce soon as well which we'll just give a little preview of here is that we're forming a new research and development foundation to support application level research and development into the next generation of dapps and if you're more broadly curious about or interested in working with us on zk research and development you can email me at gubsheep xeroxpark.org my dms are also open i'm pretty active on twitter and also check out ludin's uh who is the developer behind who was one of the developers behind the zk dungeon game um ivan who is the lead developer of dark forest dark force twitter and eath university which is some of the educational initiatives that we support that's about it thanks so much for having me and um yeah i'll pass it to you karthik thank you so much for that awesome uh list of uh things to do and not to or how to prepare for this um we'll we'll just do one question um i i think i may have asked you this a variation of this in the past but given the nature of how these things work under the hood um how do you kind of think about or balance programmability to game versus kind of making it fun for the person like what would you see bots competing with bots and sort of drive out all the people who are doing uh playing a game for fun or do you think they'll just actually knock into a separate sort of subcommittee or group or universe and play those games um somewhere else yeah this is this is a great question definitely still an open question um in our head there are a couple of different approaches here um one approach is basically that we encourage all the players to and we try to bake it into the culture of the community that uh knowledge and tools is meant to be shared so we make it so this means a couple of things one is lowering the barrier to entry to building these automations and tools and providing players with uh you know building blocks which they can construct bots or other things out of um and this means you know clean apis or whatever else the other is encouraging players who are very active and trying to reward players who are active um in open sourcing their work and making it accessible for new players um and creating those touch points like an easy to use plug-in system um so that there can be sort of a level playing field this also has the dual effect of um making dark force something of an educational program for its players where people can steadily onboard you know depending on how deeply they want to go into the game they can go deeper and deeper into playing with the programmability of the game so i think another possible outcome here is that you know you if you're building a strategy game you want to build a game with a fair amount of strategic depth to the point where you know like chess was not boughtable for many many millennia um it's a game where you've really got limited resources and it's it takes uh strategy and intelligence in order to be a good chess player um similarly here like with rts games there is some degree to which bots can't help you but ultimately like naive strategies are still going to be beaten out by manual by players who are playing manually and have a deeper understanding of the strategic landscape of the game so in fact a lot of the top players of dark force are still primarily manual players or not developers um and i think that what we're evolving towards is sort of like a human ai hybrid uh competitive gameplay style um towards the top ends of dark forest now that that makes a lot of sense i think uh and also i guess to uh maybe just sort of think about how things are evolving in nfts overall um people are really going into derivatives of kind of games and nfts overall and sort of branching up into their own self communities too so as as things kind of get either more heated or complex or complicated i do think people will just self-select into zoning into their own world exactly yeah and i think that we're already starting to see um it's becoming difficult to place highly in dark force without being a part of a player dao whether that's a group of people sharing a multi-sig or sharing a burner wallet private key or who have some more advanced smart contract infrastructure and then you can start seeing specialization like this person is responsible for building out the automation and bot tooling this person's responsible for piloting the account during this time zone um this person is responsible for negotiations and diplomacy so that's been really really exciting to watch as well and maybe just one last question before we end from what you just said is that in line with what you wanted it to be or is that just uh you were just kind of talking about what's happened so far yeah i think um for the dark forest team we've always speculated that some of these directions might be possible but really it's up to and we want to make this true as much as possible that it's up to the players in the community to see um what uh third-party tools or what additions to the game dictate the gameplay experience like i would say obviously it's hard to measure this um quantitatively but you know i would guess that more than 50 percent of a player's top players like dark forest experience actually now lives in uh actions that are going through third party code rather than the dark force core client or contracts themselves so for us it's really about creating that room for experimentation and trusting that a community of players who is you know currently at least like culturally aligned and maybe even someday like economically aligned will have better ideas than just us and we can try to provide an unopinionated um base layer with just enough creative direction and just enough to play with that players can run wild with their imaginations that's amazing well with that i want to thank you so much for doing this amazing talk and uh i think one day we'll definitely get you to play this game live on one of these yeah maybe a hackathon or something like around cats don't do it justice so awesome well thanks karthik thank you so much guppy um all right with that we are ready to close off today's nft and creator summit i want to thank everybody who stuck around and asked questions and engage with us on the chat we had so many amazing talks and speakers and questions coming in from the audience and uh i also want to give a special shout out to everybody who reached out to us last week about giving feedback on the new eve global.tv design we got a lot of bugs still to fix but we're slowly working on them one of the harder pieces here is that we can only test these things on production when real time summit is going on because a lot of these are latency and race condition related bugs that only happen when a few hundred users are watching in real time but we're slowly working on this thing so thanks again for everybody uh with that feedback and hope you are enjoying some of your new profile photos so with that um we're gonna sign off and i will see you next thursday for the of a grand style summit and then next friday for our dao and governance summit um in the meantime check us out on twitter and we'll see everybody who is participating at the hackathon on discord take care everybody bye [Music] [Applause] [Music] you