9 Takeaways from Duelyst: From Tabletop to Digital Game

[Music] listen to us and you guys are here for the nine takeaways from duelists from tabletop to digital game and we want to share a lot of our experiences from prototyping it all the way to all the obstacles and challenges that we personally faced Eric and I on developing this game from tabletop prototype all the way to a digital game on Steam so just to kick things off have any of you guys played duelist okay Wow and how many people here are have a tabletop prototype that they're planning to convert or to start to pour it over to digital okay great so I'm gonna just start off with a little bit of background about ourselves Eric can start off hi I'm Eric the lead game designer of duelist and I'm primarily

from the tabletop space I've designed board games and card games for the last 15 to 17 years mostly I'm mostly known for living card games for Fantasy Flight I did a little work for Wizards of the coast on magic and Duel Masters most people know me for big box board games by this point I teams like blood rage and XCOM and chaos in the old world I'm a huge fan of both magic and the games like Fire Emblem and Final Fantasy Tactics which makes our meeting very serendipitous because it was really exciting to work on the game and I'm Keith Lee I'm the CEO and creative director of counter play games I was the former lead producer of Diablo 3 at Blizzard Entertainment and also the former lead game programmer for the Ratchet and Clank series at insomniac games but more importantly what's dualist so you guys already know a little bit about duelist but for those who haven't had heard about it or played

it it's essentially a collectable tactics game and it's a way for you to assemble a small army or deck of cards that are actually playable on a nine by five grid and it's online you play 1v1 head-to-head and has turn-based battles on there currently we're only on the PC Mac and it's playable also on the web and we have about 250,000 matches that are played per day the inspiration for this game is the fact that we love Final Fantasy Tactics we're huge fans of Fire Emblem but those are primarily single-player experiences where they gate you at a boss or a certain liked level cap before you can progress and our curiosity and the question that we wanted to ask was can we build a competitive tactic style game but still be able to make it in a very short to play a match in a very short reasonable time so our goal was to build a game

that lasted 10 minutes per session or less and to still keep the breath of being able to still make it relatively competitive to have a lot of Ark types and options from aggressive to control or mid-range and we felt that using a CCG sort of model for us or totally well because it's very hard to build a competitive game if every character that you have have sort of different power levels or things like that so there was a really great way for us to combine CCG elements that people are familiar with but to also have positioning and to introduce like a lot of new concepts around that and so this game itself is a 9 by 5 grid there's a bunch of minions that you can play on the board there's also artifacts and spells that create effects on your units and the object of the game is for you to defeat the enemy general and we kept the winning objective and the condition

really simple it doesn't require a lot of math or calculations and things like that and essentially you have a hand at the bottom which we call an action bar and you can also replace a unit every single turn and these are some of the cards that you can collect and we have 400 plus different minions that you can play on the battlefield each with their own animations and things like that the games about she been out also the games been out for a year in past open beta and we have we're working on our third expansion right now yeah it will release soon so how do we get started more importantly this was basically what duelist looked like when we first started and we we would be testing and prototyping our our units and we wanted to be able to iterate on visualization of the board state as well as how we can prototype as quickly as possible and we had a lot of iterations and one of the important things is that we tested not only what people who love tabletop games but people who only play digital games

and we also play it with folks that are in the industry and also people that are outside of it so in the bottom left these are some folks from Riot Games Paul SATA Santi and Steven Lin and these guys are guys from riot that play a lot of tabletop and then there's other folks that have never played tabletop and only played games on Steam now we wanted to also have like a very cheap DIY way to be able to prototype so for the game that we wanted to build it was relatively simple we just bought a chess X vinyl mat that you just get off of Amazon for probably like less than $30 and nice thing about those is they have a hex ID as well as a square tiles so depending on the type of game you want to make it's really really easy to prototype we just get a bunch of polyhedral dice and typically if you really know what type of attack ranges that you're gonna have or hitpoints or HP they're gonna have for your units we

did like a 28 d 20 dice for the hit points as a tracker and a counter and then we did like a d8 for a like attack or anything that you want to do for defense and things like that we then also started to print out all the effects and spells and cards that we're gonna have in the game and then we cut them up every day when we iterated and put them into card sleep protectors so that we can always put it on top of the square for the type of effects and buffs that are occurring for your units and then we we love to get like Pathfinder miniatures because they have a lot of unique heroes there's a you can actually add a little bit of flavor into the prototype if it's a flying unit or a knight and to kind of kind of create a little bit of a sense or if you have a unique hero then at least they would sort of stand out but then of course just to keep things relatively cheap we also got like a bag of like generic fantasy miniatures that are like a bunch of knights and things like that so there are viewed as like the pawns of like

chess and you can just get them really cheaply so there's a lot of repetition here so there's a really important design point also why I want a prototype like this for as simply and stripped down like this is that there's a bit of a misconception in tabletop and digital that just because you're designing a digital game because in the digital space you can make the game really really complicated have a lot of hidden information we found that isn't true we wanted the game we wanted to make sure that the game is comprehensive even on tabletop so that people can really really quickly crock what the board state is without having to rely on figuring out what is going on in the back end of the game because a lot of players assume that that's cheating yeah a little side tip too we like to do a 2 by 6 on the square tiles because they actually fit the cards that you want to place easily so if you're moving your characters and units it's easy for people to remember what a tile they're on for your cards and and it's really simple the other important point too is

that although we have a lot of different types of players you can actually have a pretty small sample size to be able to test your prototype and the key thing is just making sure you can observe what they're doing and not trying to bias the opinions for how they're playing and just getting a sense of what areas that they might be confused around or they might be doing a certain strategy that you've never considered and we usually just have a logbook so every single time we play we just track what units they've been using and how they how they won and we don't really spend too much time on asking that they had fun or not because if for them if they start to look a little distracted they're they're starting to lose interest we know that it's it's not capturing them anyway and that's that usually like our first sort of gut test for how things are going absolutely and also the because one of the keys to a digital game is how

smoothly it plays it one of the reasons we give ourselves limitations of this medium is of course we knew that we could we could fake a lot of the we can fake a lot of the user input in on the digital side we want to make sure that even if we could get this prototype playing fairly smoothly and fairly easily that that putting the extra polish on the digital side would be absolutely top-notch mm-hmm now before just even starting with tabletop and moving a digital we took a step back and started to think about what the strengths are of digital and I would say 20% of the time before we even started implementation because it's really easy to roll up your sleeves get excited to start to code a couple things and try a couple ideas on digital because it's still really fast to prototype but we took a step back and we said hey what are what are the strengths on the tabletop side and the weaknesses and what are the areas on digital that we know that we need to bolster and to continue to support that's very different from tabletop so the biggest

salient thing on the tabletop side of course is that you have a very very custom highly social experience with friends and it's very easy to have games where you have secondary expressions and ways that you can Bluff and to read signals because people are interacting you're face to face and of course the challenge and the weaknesses of table tabletop is the fact that it does take time to schedule usually need to have a champion or leader to kind of organize things generally as a larger time come in to make sure that things work out for that week now you don't have any of those problems on the strengths of digital you can play randos anytime anywhere players across the world so that's a huge strength from our perspective but it's also a subtle weakness as well because you lack that level of personal interaction with the players that you're the people that you're playing with and of course it hides a lot of number crunching on the digital side so there's for the type of game that we

built on tabletop we knew that it's going to work even better on digital because we have lots of interactions and secondary effects that all interact with all of the keywords that we have that actually we had now ported it back to tabletop that's right it would actually be very hard to have this exact do Louis game because there's so much number crunching that actually occurs on the board state especially introducing movement and things like that now on the digital side the area that we wanted to focus on and to bolster is we knew that there was going to be a lack of deeper person-to-person interaction and there's this sort of mindset that you're looking at a screen you have the web that you can get onto there's movies there's there's like more distractions for you as you are on digital so as a result anything that you can do where there the players ultra engaged never has this feeling that they have to wait is utmost critical important versus

being on a tabletop you know night situation right and also the key corollary that is that in digital you lose the social contract of the table right so I mean we've all played digital games we know how what the different the difference in politeness between a tabletop game and in digital so we had to make sure when we were designing the game to make in in a tabletop game when you're not going to rope out every turn you're not going to make people you're not going to great people for no reason you're going to play efficiently and there's a lot of unsaid communication between players you know one day and the turn etc and this we had to design the game around that to make sure or around the worst possible type of griefing player in every possible way to make the experience bad for their opponent we had to assume that the player would so I would say that this is a very common graph for anyone that's going to move anything from tabletop to digital and something just for for you to consider but then it opens up another question is a tabletop player similar and what's the

overlap to a digital persona do though does a player that love tabletop where they even play duelist if on digital so that's a hypothesis that was a big question we didn't know that so we had to just make a couple assumptions and to kind of test out our hypotheses so our initial first most important assumption is people don't want to play 40 minute games and then lose and feel bad on digital so we wanted to ensure that our play sessions were less than 10 minutes and then the other difference is that the velocity and the frequency in the volume of games that you're going to play on digital is astronomically exponentially higher than playing a tabletop game where you might just have one game that night and most of the variety may not even come from the board or the game itself but from the interactions that you have that night with the players that are playing with you so from that perspective we had

to make sure that we increase the priority around creating match variety in every single game so the the goal is that for every new match you might encounter a different general you might have a you might encounter a different deck or different set of minions when you play against other players so that things don't feel a samey and this is of course the number one challenge for building a game like ours is to avoid - how to make the people who grind through 50 games a day to still have them enjoy themselves in the for the folks who might only play a few games in the day and the hypothesis for us is that we had to make sure that you're playing with anonymous players and how we can alleviate any of the negatives around that so on the data side we also track Peter Drucker has always said that what's measured is improved and so we took that hypotheses and then we started to test it out on the web first so we

actually didn't push out our game on the PC at first we wrote our entire engine in JavaScript and then kind of put it out on on the web and the main reason why and why people always asked us this question is like why'd you guys do that it's because it's very easy yet for us to deploy a new build to change the balance of things to push it out and it's very easy for a user a new player to get into the game without going through a lot of Hoops and to login and do all these type of things and we started to track very early on what the average playtime was for the players what the actual match times were and so when we started to tune our cards when we started to tune all of our systems our number one goal is to make sure that the match time didn't increase over ten minutes and so originally it started at around 15 that's right yeah so once again we haven't even started talking about the takeaways yet but just we really wanted to set the design

pillars that we had started with at the prototype stage again to enforce a match time upper bound to ten minutes and that's already we feel at the highest for for a type of game that we have especially if you want to potentially move your game onto mobile as well and then number two we wanted to offer a wide variety of deck arc types and then third we wanted to find ways that we can build player confidence in the game and also to allow them to see that they're improving in terms of their skill level and then lastly we also started to predict and start to think about how we also wanted to build our ranked player distribution so early on we knew that it wasn't going to be a Gaussian normal distribution bell curve we instead knew that we wanted to skew the curve towards time served and the amount of time that you are playing and skewers slightly towards like the lower level so this is actually the results of where our ladder

stands in the last few months and you'll notice that the distribution of these five different divisions and just just to give you guys a little background guys haven't played ranked is that every month that resets and as you continue to play you will progress up the ladder if you win but at the lower levels you just have to play to be able to go up and we actually have a floor on every single division so what this does is alleviate sort of laddering stress because once you hit a certain division you can start to experiment a bit play around with different decks that you do try out without losing rank so this is the distribution we share this with our community because a lot of times people think hey I'm not doing well in this game because I'm not silver or gold or or even diamond or s-rank but but they'd be able to show this to the community also helps people realize that

it's not just about being in bronze because a lot of people spend literally hundreds of hours playing our game and playing like a hundred matches a day to get to s ranked so alright so we're gonna start with the first takeaway that we learned with dualist and the first one that is different from tabletop is we focused a lot on real time interactivity so what we mean by that is everything that we did in the game when it's your opponent's turn everything from mouse movements to what they hover over to what they're looking at to them dragging a card to display what they're doing is all done in real time so we have a real time system to track everything and you can actually see what your opponent is doing so that way you still have this feeling that you're playing against a real human opponent that they're looking at things and doing stuff rather than you know waiting for like a minute or 90 seconds and there's like nothing happening on the screen and you're just off to get a

coffee or something so we want people to still feel like they're engaged on the opposite ends turn and on a deeper level actually this is one of the I'm a hardcore CCG player and I one of the things I wanted to make sure that we did here was that in tabletop games you can you can read cues especially when you're bluffing any game with hidden information like we have you can read cues from players by from body language watching what they're looking at and you can also double bluff people by giving off false tells in this game if if you're real if you're a high-level and you're familiar with the cards you can actually watch what people you can watch what people are looking at you can watch how long it takes somebody to play a card at a particular time you can watch what they're what they're pointing out what they're not to see and start to formulate some hypotheses and it gives you its inches gives you a tiny little increment advantage but that's important in games like this where you're improving tiny tiny percentages at a time mm-hmm so the high-level idea for anything outside of even the real-time main activity the summary is just to think

about ways that you can engage your player when they're not actively doing something in the game and they're just like waiting or pausing so here's actually a video just to show you of the enemy's turn here so on the left side would be me let's say and your opponent now is actually hovering looking over some units looking at at the two units that you have and they're gonna deploy things so this is all done in real time or a real-time system in the backend to better track this and to track all your mouse movements and everything like names animated right now which means he's in the middle of he was clicking on the card in his hand and looking at and thinking about playing it yeah and of course this is an example also of besides just doing it Mouse interactions you can also do emotes and other like soft ways to communicate with each other so that you feel that there is another person that you're playing with but still in a kind of controlled way so

just to kind of dig a little deeper we also offer a restricted level of player two player interactions so as you seen emotes and the cool thing is that once you start playing a lot of duelist it becomes like an expressive language there's like a language in itself for how people start to communicate with each other in the game based on the dozens and dozens of emotes that we sort of have and that's like kind of a cool language that people start to learn we have a buddy list where if you add and mutually agree to add with each other you can now start to chat and we also have replay sharing within the buddy list so I can actually share a replay that I've just done to show off like a really cool lethal that I did and we also outside of like the game itself we have on our main menu a watch section so the idea here is that if you're not playing a game you can actually go into any division ladder and start to see some of the more interesting popular

games and we haven't algorithm for pushing out kind of unique sort of games that are happening you can watch on any side and then you can sort of see what people are doing so this is a fun way for players to just relax kind of watch what people are doing learn some new strategies or kind of see like what other people are doing you know in a very safe kind of environment especially if people are just starting to learn to ladder and jump into the game and of course we also have a spectate mode so you can hop in to any sort of at which feature person that we have and one of the things that we did we actually integrated twitch into our PC game so we highlight twitch streamers that are playing do a list and it's a benefit for them because they get players that are just hopping in and we also let them to jump right into twitch and watch it and of course you can actually spectate your friends and this is actually really important anyway because it's a requirement if you plan to do competitive tournaments and to be

able to have a spectating sort of system so that tournament organizers can start to display different sides and things like that so this is a real-time interactivity these are some of the emotes we spent a long time on these by the way I think we we must have spent at least a week in just design and to figure out because we didn't have infinite emotes and the key would I remember the key our discussion we had we want to make sure that the language was a hundred percent emergent so we wanted to pick some of the most common expressions that people would use that didn't necessarily point to a to an exact word or an exact adjective and allow people to use them individually or in strings in succession to communicate with each other and the meaning behind it like as you created a really cool secondary talking point across across the internet across our forums and reddit and stuff was pretty cool okay I'm still on takeaway number one then so the other thing too is that's really important is again people are

impatient people don't want to wait so anything you can do to shorten the matchmaking queue times ultimately if you can do it less than 20 seconds especially on the lower ranks is pretty critical so I played a lot of Warcraft three shows kind of like my age on this but tons of warcraft 3 and you have the roving eye and you have like a sort of a timer that displays like how long you've been waiting those all help because anything where it looks like the screen just froze and you don't have anything to show will concern people so we always have like a timer that displays and there's there's some level of interaction people really like you know that you're in a matchmaking queue and we really try to make sure that we can get below 20 seconds and sometimes that means that early on when you're first like pushing out your game on digital you may not have a lot of players jumping in right away and so we it's it's very like all your fans I'm gonna always ask to add like all these different modes that they want and we

all want to do that that's like the things that we love doing we want to have rainbow mode we want to have casual and all these other things but to have that you need to have literally 200-300 thousand matches happening a day before you can even get close to to hitting like a less than twenty second interval time so for us we started very simply with a rank ladder and then also we had a draft mode which we called the gauntlet so we didn't release casual mode at the start and I can talk a little bit more about casual later and only now for the next expansion that's coming in a few weeks are we adding this rainbow mode system where you can actually mix and match every single card from every faction for the first time so that's gonna be a lot of fun for people we hope those of you guys who play it's going to it's even more fun than you think it is okay then for so starting with take away number two you're only as so one thing we learned is that you're really only as good as your matchmaking

systems so on a high level how we kind of build our algorithm was we thought of a bunch of different metrics that we wanted to track and we put a weighted parameter on each of these metrics and then later on once we did that we reduced it into a normalized range and created a rating score and then that way it was very easy for us to compare in real time how to matchmake different people so some of the metrics that we used or the most salient ones is rank radius so what we mean by that is we would start out matching people that are both ranked 24 let's say and if that didn't coincide then we would actually expand the radius on that so you'll start to match someone that's 23 and 24 and we were starting to expand that out and that would actually have its own that trick that would get normalized into a weighted system we also have a time serve so what we mean by that is that we look at every player's profile and we look at how much time they've actually invested in to do

a list and we also track as a proxy like it's a high correlation we also look at their wind counts and how well they've sort of performed internally and then we we give priority to people who have been waiting for a longer amount of time so then we start to actually expand that put a higher weight on this and that's the queue wait time so we feel like people have been waiting for more than 20 seconds will will actually give in and make sure that they can jump onto a game even if it's not the ideal match and the number one thing that we wanted to avoid at all costs and people have different opinions on this so other people are totally fine not putting this into the system but for us we felt that it never felt good to be able to fight and know to play the same player again on the next matchmaking when you when you're matched because the loser is gonna feel super frustrated because they're gonna ask like why am I getting matched up with the same guy that just beat me so I think that that we view as

a kind of a bad thing and then more recently we started to think about once you get to over 200 to 300 thousand matches a day you can start to do this thing called cohort based matchmaking so what we can now do is take a tranche of players that in a came in and a certain season as a cohort and then match them all together so what this helps us do is it actually is a way to alleviate it it actually solved all of these sort of problems because it's almost like everyone's in the same generation or wave or class and they're all starting to learn and kind of grow their decks and it solves a lot of the kind of issues for us so court based thing is something new that we started implement that has worked nicely for us it's also a really good problem also with the we were talking about with a deck sizing collection right one of the invisible issues with a lot of matchmaking games is especially with people with asymmetric power levels is the new player power doesn't necessarily

equal rarity in this game but it can feel like that to some player somebody doesn't want to come up against something at the same rank that has like 50 legendary cards in their deck so this helps almost by itself alleviate that now again Peter Drucker said what gets measured gets improved so specifically around matchmaking we also built a system where we actually graded ourselves at any moment in time as a snapshot throughout the day and we had a rating system based on those parameters a 1.0 perfect point seven five point five and zero would be for example the same player played twice in a row and the important thing with this metric is that we did a system for every single division so you need to have a granular enough matchmaking system to be able to track your performance across every division because in some cases your bronze matchmaking might be at close to

1.0 but then some happening on silver and gold and those match matches aren't really working so here's an example so we actually have a real-time graph that all of us have on the screen at our office to be able to track matchmaking quality so in any moment in time during the day we can see how things are performing if things are above at what division these numbers are hitting a good number so if it's over point five I think we're relatively happy and the nice thing with this too is that if we put in a balance change and we tune something we can already or we change like how our algorithm does stuff for the matchmaking this helps a lot this doesn't ever explain why your matchmaking is going down but at least is a good early indicator that something is wrong and that you can actually think about how to solve it so we don't have this it yours guessing at that point if matchmaking is good because hearing stuff from the community is always helpful but at the

same time you're gonna get so much noise coming in from the community people always complain that your matchmaking sucks so you need to have your own system to be able to at least have a way to measure that yourself and internally understand how you can improve that and sometimes people actually wonder when we're doing doing balance changes in tweaks like why why we'll address a particular issue that wasn't necessarily called out by the community it's usually because we have this is one of our starting points and if it does hurt matchmaking at some level we that is we've considered that as important as any other balance metric so the third takeaway that we we thought about was realizing how can we extract fun and progression for people that are still losing matches so no matter what the some it's a zero-sum game someone's going to lose and we offer the most fun for people that continue to lose matches so one of the differences with tabletop

is you can walk away finishing that night ending a session and even if you lose you still had some fun in some way because you're interacting with your friends you had some sort of social capital so what we did on the quest side is to think about how we can offer play quests that you can complete and get rewards for that and also a win based quests so there's a delicate balance because again if you only do play style quests to accomplish your quests in the game you're gonna have a lot of people starting to concede in your matches for PvP and you so you'll have to always develop a system for penalties and an algorithm for detecting when a player is conceding and that's one of the things that you have to twist a lot and change a lot because you can't ever let anyone know how you're detecting people are that people are conceding so one week we might track how long how many turns that

they're playing before they start to exit so we always like change how we detect the conceding algorithm and it's a lot of work but if you do want to avoid the exploit of players constantly conceding that's something that has to get built and a lot of people don't think about that and it's it's just one factor to think about as you as you develop your game for a digital that isn't something that's common on tabletop and then faction progression so the idea for us is for every faction we don't just give you a big starter bundle with like 30 to 50 cards and you get overwhelmed with like how to even start so we always have a couple marquee cards so that you continue to play your faction you start to get introduced to new cards and we make them really clear with a single ability very few words on the card text and typically it's very obvious to the player that this card has a wind condition that this is something I can use to start playing and to win so

that's from from our perspective what we kind of built and this progression allows people to feel that they're they're learning the game that they can sort of win in the lower ranks that they can play well on practice and also have this feeling of getting more cool new cards and and getting excited about how they can change different cards and start to develop and master their deck or their faction and what we also do too is that if you start to lose a lot of games we we actually have achievements and things like that where at certain specific points of play behavior will give you orbs because we want to make sure that the people who aren't succeeding in the game can still have ways for them to find get new cards and and and still feel like they can progress even if they're losing a certain number of games so that's a that's a really good way to kind of bolster that format just to be clear for those who don't play orbs or the orbs are the public packs that contain new cards would I remember we tuned that a

couple times didn't we huh man how many games it would take to lose to get yeah because again people concede and try to get those orbs and and that's a natural thing but that's a that's a natural thing so it's a balance on what you're going to do so here are just some examples where we front-load a lot of the rewards for players that may know be winning or they're not grinding 50 games a day so we'll have win a practice game which is relatively easy because you're playing against an AI so that's a win condition game but then we have a welcome back quest where people feel like they can still play a few games and get some gold early on and kind of warm up if they haven't played duelist in a while we also have a first win of the day to motivate people to just keep playing until they can win their first game and to build that sort of confidence and still front load some gold and then we also still have like you know typical play quests first I'll talk about that for real quick like there was um the the question there's a real delicate balance between how we chose what what quests were win quests

and what quests were play quests for the most part we wanted people when you log into the game especially if you're new player you log in and the this is inviting this tells you all you just need to play X number of games the first one of the day was actually hidden we didn't show that we just showed it in anywhere on the HUD but of course the first time you win a game you're like oh I win first wind of the day which does motivate you to win we didn't only want players to just fit to make it feel like a Facebook game you just log in in play we wanted them to feel driven to win so you can you will get something even if you lose over and over again but the that the first time you when you'll feel you will get that you'll get that adrenaline rush and get your cool packs yeah and you can talk a little bit more about the in match stuff too right so the what we were talking about before the extracting fun from losses was in that was in the the meta game right in the game itself we want to make sure duelist is very it's a high skill game that's high higher skill than some of our competitors are but um which is in

which is daunting right sometimes people are going to lose by would not know why so we put a lot of subtle elements to the design which is taken from my trading card game background so in a match we put we built the game around milestones so even though there's one clear winner one clear loser I want to make sure that there were the were things that people could accomplish inside and match whether or not they won or lost so most of the game for those of you Poe of course many of the cards are very combo oriented they're simple but there's some clear synergies and cards and there's some subtle synergies and other cards if the game is highly consistent so if you can if you're playing a if you're playing the Vitruvian faction you want a bit you want to play a you want build the the or the deck that lets you build Watsa obelisks and generate tons of little guys you might lose but at least you got the cool feeling of having built that board state the learning moments of course the game is relatively high skill and highly interactive there are a lot the positioning on the board means that

cards have different value based on where they are in relation to your opponent's pieces so even if you you the takeaway you'll get from a game very often is the very clear like oh if I only had played if I only played by minion behind my general instead of in front of the general I actually would have won because he wouldn't been able to attack by X so that was none of these are explicitly out there but people do get the they will get the feeling they will learn quickly from game day game we also push really hard for being able to have replays very early on in our game cycle so you as a player you can just hop into your profile and look at an entire history of the last set of games that you've played you can start to watch and kind of see like what you potentially did and observe how you've performed and see if there's potentially like a few mistakes that you have made and how for example you might have lost lethal or things like that so I feel that it's a evergreen learning platform

to have replays in the game and although it took a lot of time for us to build because you want to fast forward and scrub and do a lot of different things like that it's worthwhile because people will go back and in a game and be able to learn from their mistakes which for a game like ours is challenging because sometimes you just don't know why you lost and that's that's the hardest part of trying to explain things in our game is you may not know because something you could have done like might have been even like two turns before so this is like a thing to kind of alleviate that too so takeaway number four is to we focus on a lot of systems - that can encourage positive behavior so we had this idea of player gifting so of course that's more standard with a lot of other games but we really wanted to make sure that the rewards that you're giving to your friend was was pretty significant and very high and you can still rate limit that like per so if that's an

issue you can watch rate limit it in some some way but you really feel like that the recipient is getting a really good reward from that and I think that's really important and also kind of a interesting system that we haven't seen in other games really where the winners can also tip the their opponent after the game and the idea for us of doing that was we felt that let's try to distribute gold and currency as far across the ecosystem and people actually really like tipping and kind of as a way to kind of help another person that may not have some of the cards or things like that and we also had a top 50 tipping leaderboard so that people had an alternative vector to progress in the game just by tipping some of their other opponents rather than trying to gruel through necessary like s-rank and things like that and then we would display that on Twitter or like our Facebook and things like that and we also send them

custom physical rewards that are unique from like in-game rewards so we actually mail them to some of the winners and things like that and as a cool little prize it was a it was besides us the tipping leaderboard being super quo is also important that we made sure that we'd never made any explicit reason for tipping it was entirely up to the players why you would tip so that they would they would make their decisions based on social norms rather than anything else rather than something they would try to exploit that's very important that actually dates back that's a tabletop style mechanic if any of you guys played old games workshop games back in the day Warhammer stuff but they had a sportsmanship award that you would that was awarded by the by the group by the players of the tournament to the player who made the game most fun for their opponent so from a design and systems perspective perspective when we're building do lists it's it's really easy to start to look at achievement systems and then faction

progression systems and the first-time user experience and think about okay we're starting to see an issue with how our players behavior is operating so let's try to add this quest in here to fix things and stuff like that and that's gonna be a problem because you're gonna start to have these kitchen sink sort of systems that don't really solve anything at all and they're trying to plug like little holes that are starting to appear and then the problem later too is that some of your later designers or people that started to join your team have no understanding like why you built this particular system and what problem it was even trying to solve in the first place so for all of the system that we develop we very clearly delineate what the goal of that particular system is trying to achieve on the game design side and if it if it does not fit that goal for anything we want to plug in we decide whether to remove a system or to develop another system that accomplishes that goal rather than trying to mix too much stuff into a single system so for

example achievements were very different from other systems like quests because the purpose of achievements was to give you a ding and like an achievement at a very specific point that we're tracking to reinforce a particular play pattern so for example if you're losing a lot in a row we actually give you a particular reward the first time you ever disenchant a card you get an achievement and also it gives you enough spirit to be able to then craft a new card so we're reinforcing that you can now actually build something from something you disenchanted rather than just getting all the spirit and not knowing what to do with it the first time user experienced the FTU e we made sure that we didn't really focus too much on the meta systems and how those work but the goal of the first time user experience was just to focus on the in match gameplay and the sort of dynamics of how you're playing day to day like practice an infection progression again is to solve the issue that we want to do

a non-starter bundles and packs and how we can slowly teach the player like new concepts in a way that we felt like was the best possible path for them to learn a faction as quickly as possible because again like an LC G model other things that do work on tabletop isn't and this it's not a necessity for doing something on digital and I think a lot of our community folks say hey why don't you do a LCG thing well we don't really have to do that I think that's actually a almost a result of like the practicality of having to print stuff on tabletop from our perspective giving someone a starter bundle is challenging because we don't know how to value that the players don't know how to value a start a bundle yet they're super overwhelmed so being able to actually show new cards and then in a nice sort of progressive way has worked better for us even though I think most people do want to do starter bundles I think that's a very common decision so

just something to consider I think every game is different but that's the sort of philosophy and the thinking that we took absolutely when we were doing design and tech worked very closely together and we were when we were working on systems design one of the the litmus tests we would always give ourselves is we would run a simulation of first-time user experience like virtually write it down on paper or figure it out on the graph and then plug that system in and make sure that it is as seamless as possible for the first-time user it's one of the biggest traps of does feature base design like this is it is really easy to lose sight of what the first-time user experience is once all of us have our own accounts that's it is time consuming to keep starting new accounts and run through that experience again so but we have to keep cognizant of it because these games live or die by acquisition ok I'm gonna zoom through a little bit faster so even besides doing some of the stuff around systems like our progression vector even around like our goals for like how we design cards had

the same idea had to fit like one goal so everything from our goal for our text was avoid hyper literal text so we started at as the this and we said hey we just got a lower word card count because our stuff skating like way too wordy and we try to aim for like a budget of 30 characters per line a hard max of like four lines single card abilities especially early on and just kind of move discoverable Xin to a different area so people kind of start to discover cards so here's like an example of a very typical day to day card for the original idea for the Cara which is a general blood-borne spell and we started with whenever you're whenever you summon a minion from anywhere this turn give it plus one plus one it's exactly technical it's very accurate it also represents how tokens work in the game but now you have to spend all this time to read it and most people won't really know the difference and the bottom version is cut almost in half and it's just generally more clear and accomplishes 80% of the objective of

what we wanted to say I know some players do complain that are that their text isn't is precise enough but aggregate complexity is a real thing the Wii you can never judge a card in a vacuum you have to judge it as part of a six card hand so every time you add a word onto a card you have to multiply that by six theoretically and the the aggregate complexity can turn off new players so fast that like again kills raka we still challenged with this all the time oh yes a very tough one so for takeaway number six again not everyone can catch up to hardcore players they may spend more time playing a game some people just are better with positioning they might have a better collection through time and originally we offered a discrete number of wins and you get certain like why gold but again this makes people the stronger richer and the and it makes you can create like essentially a death spiral where the bottom 10% of your community just quits just because they don't think they're good at the game or they're losing a lot

and then now the lat the next 10th 10th percentile will leave and then all you get are these super hardcore high skill super elite players that can play like 50 games a day so what we did was think a lot about how to do catch-up mechanics on digital so we had Welcome Back quests which we talked about and this one is a very interesting example the first original idea for doing a catch-up mechanic was hey we're gonna offer you 10 gold for every four ranks place so and we have a cap of 50 gold but we started see people progressively starting to concede 20 games on ladder this is like what they did they login just lose 20 games and they figure it out like how long they had to play and just estimate and optimize their their goal time so what we ended up doing is then switching it to five gold per two wins and what that means is that per completion you get +10 extra gold for a completion so what that means that in

the first 14 wins per day you're essentially getting 15 gold per two wins and it tapers off 14 so the nice thing is that it allows players to start to mentally think that I'm kind of losing extra gold if I don't play more but in a way that reinforces them to play and to win and it's it's it's a it's a positive it's almost like the original like wow rest XP system where and certainly punishing you just said hey you got some extra XP that you can get just by like playing a bit more or killing a few more mobs that day so that's that's actually a very similar psychological system than it is all positive reinforcement but then what this also solved to is that for people who wanted to grind and continue to play and play a hundred games a day they can still continue to earn gold because in the previous version we capped it at like a max of 50 so this pleases both sides of the playerbase absolutely so in the catch-up mechanics

that's what for outer game systems inside the game we use again a lot of standard CCG designs here so that mana ramps up every turn so the amount of stuff you can do the power level of your plays increase every turn so people who get an early advantage are could be defeated by somebody with who has weights on more powerful stuff that's pretty standard the stakes elevate from turn to turn so we have in the game we have manna Springs that allow people to get temporary mana for running into the middle of the board and interacting with their opponent they could even somebody who gets it that first will burn through more cards would the other player may decide to hold on and get card advantage by killing more of their stuff and we have something called strong board clears players who games like this often ramped boil down to how much tempo can I get how many threats can I put on the board as quickly as possible that they might cannot react to we decided to increase the strength of the cards that can wipe out multiple multiple threats at once so

again to give an advantage to somebody who is lying behind her or allow people to play defensively in a control style wait for your point to overcommit and just stabilize don't die and then finish them up play board clear and then win the game with your more powerful and be aware of course every design decision has trade-offs so we philosophically we decided to do that and we stuck with that decision there's also obvious negatives for having strong board clears and having a lot of options for removal and that's an issue that a community will will talk about as well but there's something that we stuck to on a top-down level to kind of maintain like fidelity of the game and the board now number seven what's different from tabletop compared to digital is also consistency in being one of release patches and having sort of like this military unit basically on your team that can continue to push up so as mentioned before we originally launched on the web and one of the huge benefits we're doing stuff on the web is that we can deploy things very quickly without people having to get updates and

very quickly just started tuning things behind the scenes especially like early on then once we knew that we had a pretty consistent game we had to like the amount of time that we wanted then we can like push out onto PC but our cadence is pretty much close to two two patches a month and we did 55 patches in one year so we're we're very focused on having a plan for what we want to push out and we think about the type of content that we want to do in terms of live events is split into three areas daily as well as weekly and then also for monthly so the monthly stuff would be like rank ladder and we have an inverse reward system so for example we front-load the best rewards actually at a lower rank and as you continue to go up the rewards are actually a little bit less awesome than the ones at the bottom so the bulk of people will get the coolest cards like that month and rather than the richer getting stronger by being more skilled and getting better stuff we have boss battles that we introduced

that you can play against the AI every week with completely different types of decks we have daily challenges and you know our game is like primarily PvP but we spend a lot of time also thinking about training challenges and single-player challenges that people can start to experiment with so that they can build confidence before you you jump into into ladder and so these are some of the cards that we release every month that are different and kind of changed things up in terms of meta now takeaway number 8 is creating a social community atmosphere which on tabletop everyone does too but especially on digital there's a lot of different cool ways that we did things we split it into like three distinct areas we use discord for like real-time interaction that is like short-lived ephemeral and then reddit was something that people might check daily and the content there are the folks that would check once a day and things like that and there's like some good content you can kind of see there that was happening and then we also have

we run our own forums and those are designed for long forum where people who spend time to write long articles and essays and like deck types and you know all that sort of thing we have a custom report support system where we actually have a 24 hour grading system so we look at how fast and how quickly and what the quality is for all the different requests that we get and it's tickets that we get and we our goal is to always get like the zero tickets like every day and that's one of the key things that we strive to do and so just to give you kind of a benchmark our Twitter number of Twitter followers about 18 K Facebook is like 35,000 likes and our our view right now at this point in time is that Twitter is still a verb is a very powerful mechanism to communicate like patches and ideas and new cards and things like that because the challenge right now on Facebook is that 1% of your facebook posts reach your audience even less than that even if they like your product or like your page

and even you posts up there just doesn't get to anyone anymore okay and again we integrated featured streamers directly in game so that people can also start to if they're if they're not sure how to play the game they can click immediately jump on to watch a game that's on Twitch and also to find someone that's there and also perhaps even interact with the player and kind of learn some things and we also run a lot of dev talks where different parts of our community can talk about an upcoming boss or like how we sort of design it or things that you can kind of think about number nine is unfortunately you just can't make everyone happy in digital and it's it's really tough and a lot of the decisions that we make isn't a majority sort of thing where all the players like 80% will make a decision generally it will polarize very evenly on a certain decision so if you query a large group of players it would always polarize until like two different areas for like

what they believe is a good design and they're both all valid I think that a lot of these decisions are feedback that we get so we just stick true to our original philosophy and we try to have some pillars that we do in terms of our planning for the next expansions and our goal is just to bolster the strengths that we have rather than trying to plug all the weaknesses necessarily in our game we just move forward try to build a the things that we know that we're strong at like having a game that lasts for less than 10 minutes building like a really cool set of art types and exciting wind conditions and things like that to do it and there's just no way you can build cohn cool like a perfect game and give you a couple examples is that we had originally like two card draw where you can draw like two cards and we change it to a single draw that that really polarized them in some ways pissed off the community but there were a lot of reasons why we needed to do that a lot of times we might have the

keyword random in our card and a lot of that stuff is to fix like variety and also to smooth out parts of the skill curve but again the folks that are playing a lot and they love deterministic like highly skilled like are gonna get frustrated so there's a balance and we we just decided hey we're not going to be we're gonna be have a little bit more aren't we have a more orangey than chess or go but we're gonna have less orangey than another typical like CCG that's out there or any of our competitors and we just kind of stuck with that with that position so yeah so this is again the summary so there's the bottom right quadrant is the area to start thinking about in terms of how to improve those areas I think this will always benefit for high frequency for the game and that's that's pretty much it from so thanks so much anything I think we have

a couple of minutes for questions if yeah well let's answer any questions you guys have hi thanks for a great talk I've been wondering recently you know with with board games being so huge we've seen some successful ports of board games into digital but it seems like aside from CCGs there haven't been a lot of like nated natively digital games that Faro board maybe for game mechanics or or feel like like board games you know we haven't seen a lot of digital deck builders or work replacement games for instance is that something you guys think there is potential for or is it just like nature the medium that it doesn't work well if you mean why we think there aren't any more natively digital board game s-type games is because of the strengths of the medium right once you once you build a digital game natively digital than you are we're clearly going to be playing to the strengths of the medium and the more there's go always going to be a point

there's always a where you have to choose we have to choose is this a digital game with all the with the strengths that Keith was talking about or are we going to try to make it a little bit more real time kitchen table style game so I mean there will at best they'll probably be more hybrid style games or games that are adapted from board games in order to teach board games a little bit easier thank you uh given the process that you guys are given the amount of time you spent and invested in you and going from the prototype to digital and now you guys are know are like I thinking about mobile or at least tablet or phone I don't know how is that how are you kind of looking at those challenges so going from like desktop to mobile what are the biggest challenges from desktop to mobile yeah the biggest challenge from desktop to mobile is screen space screen size and text size so text size was the hardest thing for us to adjust for and that's that's really actually the

biggest one was text size and also the UI and how to frame some things so if you guys already tried out our like beta for the mobile it's going to be there's a certain things that we had to remove out of the game on mobile so that we can accommodate that sort of space hi I wanted to actually ask two questions if I could so the first one is just for your matchmaking as a player progresses and starts playing and they get better they also start playing better players do you feel that they're still getting a rewarding experience even though their win rate might not be increasing but they're getting better well they're I mean every player is going to have every player is going to cap out at one particular division based on their on their ranking the one that's one of the main reasons why we have floors in in our ranking system what we generally

find the behavior is once a player hits their scale cap and starts going down to the floor of their rank that's where because you can't go down generally speaking more players tend to experiment more with decks and they play more fun decks or more experimental stuff and the chances of that player winning against those decks goes up quite substantially so they will end up with a bit of a rollercoaster but the our seasons being one month and resetting Jen's tends to solve a lot of that okay um can I actually ask one more quick question related to that same issue how much do you value skill starting out skill quality of Dex in matchmaking or does that just over time become a component of their skill if they keep playing the same deck long enough if you mean do you mean quality of Dex by how good gather how we we don't track for example what number of legendaries or like the rarity of the cards when you hop into a match where we I think what the cohort based

matchmaking does is it helps us solve a lot of those sort of issues okay thank you thank you so you went through lessons learned I'm wondering if you actually had any of these like design pillars in the beginning or are there lessons learned and how how late in the process did you learn them I'm sorry I didn't so the design pillars were pretty early on so when we were already developing the prototype and thinking about transitioning a digital those were the first thoughts making the game less than 10 minutes that was definitely at the start how we built the first set of 60 cards to test out so the first 60 cards helped to accomplish that the matches would last less than 10 minutes although at that time of course you had very clear hard counters and things like that because of the limited card size and then the other goal is how we can accommodate players that we're losing I think those were all things that we planned pretty early on in the game one

other thing that's like a little bit more on the technical side is that we knew that we might move to the mobile or to have the opportunity to expand there and that's why we did pixel art why we did like 2d style art because we knew that potentially we would have limited bandwidth texture space things like that we wanted to have an app that you can download have a small binary and things like that so those are more technical things that we planned out for early on because a lot of people ask why aren't your game why isn't your game like all in 3d but the reality is having four hundred units all 3d animated on a board is it's pretty challenging and it could actually be pretty high effort for doing that and they look so cool on your iPhone it does look pretty good that pixel right now so thank you thank you you talked a little bit about the challenge of making sure that people have a really good experience with matchmaking online and I'm curious about doing that when you have a premium game and you're trying to

build out the community how you might approach something like that making sure you have enough players there when you first launch so that all your new players aren't coming on and then finding they can't play online because no one's there premium gives a paint a paid game yeah like a paid online multiplayer game that's a tough one a lot of the digital board games are paid online multiplayer it's true they I mean you're I mean with a premium game you're going you have a you've got a hard gate already of how many players you're going to get our our game was because it's a free-to-play we knew we were going to have a huge cohort of players to start with I mean compared to a tabletop game so we made our ranking system our ranked ladder very granular the the best way I could think of them to start with you would have to make a very non granular ranking letter and just allow for a lot for a little bit more a little more a little more give in actual like in the fidelity of matchmaking for sake of time because I think time the time between match make

is that is your number one consideration because that's what makes your game feel active and alive and also one thing is that casual queue seems like just another thing to do but it's in our opinion really hard to build a casual matchmaking system because the mindset is oh it's casual but everyone still wants to win in that queue so as a result even if you lose it feels bad and then also you have people with insane skill level that can jump in there and then you also have players that one experiment decks it's really hard to find a quality match for everyone because everyone's goals are so different on casual so although people will say hey I want to practice on that and not have like ladder stress actually it they don't feel that great going on a casual unless you spend a lot of time building that system so if you're thinking of maybe building a lot of modes I still think that having a very simple ranked progression system that can reset that's PvP have a scene

q is probably better because casual is like very hard to build you think absolutely Hey my question is primarily for Eric how do you go about balancing your responsibilities on dualist with whatever responsibilities you may have on your more traditional tabletop games you're working on simultaneously oh well I mean I've got an amazing team so as I'm not like the only designer we have a team we have a full team of five designers that work as a group so my role at doulas is the I mean it was a design of the core game and the and the full alpha beta set but during that time we've been training our team of really really talented people - like - in the dualist design style and they all have the they've all got separate talents and perspective as well so my job is more of as at this point is that as a vision keeper so I'm the one who will lay out the foundation for that why this head is fun what what cool mechanics are we

going to explore but this is a point where it's so collaborative that there's four of us in any one given set so even as my exact participation rises and falls in the exact card design you probably won't even notice a difference I don't think and there be some cards in our previous expansion where designers just learn can learn best from the mistakes so there's some cards that some of our other designers have pushed out that and work out very well and then also we have a very specific top-level s-rank competitive balancer that that's really top ranked and it's always like ranked zero like every every month really understands what's happening in the game yeah exactly I'm like the worst player on our team but so hey thanks guys really appreciate it thank you so much for getting a time you