Almost one year ago, I put a lot of time and energy into designing my own fan-made set. I wrote a post about the process and included all 135 card designs here, if you haven’t seen it yet. It was a hell of a process and I was very happy and proud of with how it was received by the community. Thanks to everyone for their kind words and encouragement after this went live!
Now that the Year of the Dragon has, presumably, given us all the new cards it has to give, I thought it would be a good time to look back.
One of my all-time favorite podcasts, Mark Rosewater’s “Drive to Work,” has a recurring segment called “lessons learned” in which Mark goes back to take a look at a Magic set he worked on and talk about all the lessons they learned from its successes and failures. That is, he highlights what worked, and what didn’t work as well as they wanted, and tries to talk about what Magic design as a whole has gained from the experience of making the set. It’s my favorite of his show’s recurring episode types because it give a lot of insight into the design process, both at the time the set was made and for Magic generally, going forward. It’s the type of candid and nitty-gritty design discussion that you don’t get many other places. I also like that it combines general design lessons with very specific examples and, as someone who played Magic for about half the time the game has been around, they are often examples that I have person experience with to boot.
When doing his episodes, Mark has the benefit of talking about cards that were eventually released to the public, which helps inform things like how well cards/mechanics were received and design and/or balance issues. He also has access to a sea of data, especially on the sets that have been released more recently, in the digital age. I, of course, don’t have any of that. But there’s another measure of how good my set was: how many things that I came up with have ended up in Hearthstone, and how those things were received by the community. At least in theory, there could be no better indicator of me being “on the right track” as a hopeful Hearthstone designer than for me to come up with something close to something that ends up in the final game. On the flip side, looking at stuff I missed/messed up on can be a good learning experience. So, without further ado, let’s dive right in and see how I did!
Some Things I Predicted/Got Right
First things first, I called the year’s theme! I made my post a few days before the year was announced and, I don’t know why, but it felt like it was time for Dragons. Turns out I was right on that front and, in an unexpected twist, this was the year in which the annual theme was more important than ever before. Sweet!
Short aside: This year, I’m feeling like it might be time for the Year of the Phoenix. This is definitely part wishful thinking and part prediction, but there have been some talks about potentially remaking Priest’s Classic set–and mention of big announcements this year–so why not make 2020 the year of rebirth for Hearthstone?! I’d love to see this be the year that we finally revisit the Evergreen sets and switch to a rotating core, but more on that another time.
That’s not all, I was actually even closer on upcoming themes than that. The theme of my set involved Toki messing with time and, in doing so, drawing the attention of the bronze and infinite dragonflights. Both of those dragonflights are in Descent of Dragons, and so is the underlying theme of time manipulation (a little)!
Another overall theme for the year that I also incorporated was the callbacks to prior Hearthstone content. In my set, there was a cycle in which one of each class’s Legendaries was a remake of one of the class’s most iconic cards. As this was part of my “alternative timeline” theme, several other cards in the set were also references and remakes of old cards. The entire Year of the Dragon has been a similar parade of callbacks to Hearthstone’s biggest villains (the League of E.V.I.L.), heroes (the League of Explorers), and other iconic characters (like the Wild event as a whole, the Classic dragons resurfacing in DoD, etc.).
In terms of specific cards and mechanics, I’m pretty happy with some of the specific stuff I foresaw there as well. Here are a few examples:
– I called the return of the Bomb mechanic (albeit, I had it as a one-of card, not as a major set mechanic and cycle). I wrote during Rise of Shadows about how the Hearthstone Team’s decision to make the mechanic into a vertical cycle allowed them to get the desired complexity I was going for, without being too wordy or losing clarity, like I struggled with due to the fact I was only making one card. I don’t think a vertical cycle worked for my set, but maybe that’s the type of thing where you wait until it is the right set to do the cycle instead of doing the one-of.
– I made a card with the same effect as The Amazing Reno’s Battlecry, and I wrote what I thought was a really interesting twitter thread about the design decisions in wording the effects. The Twitter thread didn’t get much traction at the time, but it’s like a mini- blawg post if you want to check it out.
Jumping ahead a little to the “things I could’ve done better” portion, I think mine would’ve worked well as having the single-target card’s rules text say “erase from time” or “remove from the timeline” or something like that (including a card name change, if needed), to add that “poof” factor, but more on the issue with that later.
– I also had a version of the Scheme mechanic in my set. Though mine was a little clunkier in wording, and slightly mechanically different, it’s pretty close to a card (Dr. Boom’s Scheme) that actually shipped. Now, being similar to one of the most panned-cards ever printed probably isn’t something to be overly proud of, but the card did get printed, so I think it’s still a (maybe minor) win.
– I also designed another card with functionally the same effect as an unplayable card that ended up getting made. But, in this case, I think it’s something to be more proud of because: 1) the card would, at least potentially, be less unplayable in the context of my full set with the Scout mechanic; and 2) the card is at least inherently interesting, unlike Boom’s Scheme which, even if it were good, probably wouldn’t be all that exciting.
– I named something “Dragon’s Bane,” but it’s otherwise completely different from the Dragonbane that was actually printed, so that’s more like a fun coincidence than anything else and, considering how generic the name is, it’s barely even that.
– I made the Chef Nomi condition (“[few or] no cards in deck”) one of the substhemes/ smaller mechanics of my set, with a handful of cards (mostly in Warlock) relating to it.
– I made a baby Amalgam, pretty close to the one that Curator has in Battlegrounds. (Long-time readers may recall that I also essentially designed the original Nightmare Amalgam before it was printed, too). If you take a look at my set overall, you will see that I like Tribal mechanics and I make them a major subtheme in my set.
Some Other Stuff I *Think* I Got Right/Might See in Hearthstone’s Future/Just Thought I Did a Good Job About
Stuff that has already been implemented, of course, isn’t the only measure of success. We hope that Hearthstone will be a long-running, successful game with many years ahead of it and, with that hope, we can also look at our designs and see areas where we think the game may go in the future. We can also just see designs we’re proud of for some other reason. Here are a few of those:
– I’m still in love with my Scout mechanic. Sure, maybe it could be implemented a bit better, but I think deck manipulation is a huge area of design space and interesting gameplay that Blizzard has only barely scratched the surface of. I’m confident we will get something like this mechanic at some point in the game–if for no other reason that one day I’ll finally wear them down and get them to hire me, move up the ranks, and at some point, like 20 years from now, finally have the power to put it in the game myself.
– I like the “Dormant” subtheme, and I like that I, generally, played with the concept of messing with board space. This is an area that Hearthstone has explored a little bit, but which I feel still has a lot more to offer.
– I think some of my class themes (especially Shaman’s Tribe-shifting and Warlock’s self-milling) are interesting themes and ways that we might see those classes expand in the future.
– A few of what I think are some of the cooler individual cards I designed:
(Murozond’s hero power is, 2 mana: Deal 3 damage and destroy the top card of your deck.) I really like that the first three here subvert expectations, but in a way that interestingly expands the way you can play the game, not just for the sake of doing it. I like the last one because it just seems so perfectly like a Hearthstone card that could actually exist in the game one day… and because it’s an interesting card in its own right. I think these individual cards are some of the strongest I made in the set–even though my buddy wasn’t as interested in Time to Think as I was. As long as I’m showing someone other than him, these are some of the cards I’d point to first when showing someone what I did with this set.
Some Things I Got Wrong/Lessons Learned
Alright, with all that self congratulation out of the way, lets get to the meat and potatoes of this post… the self flagellation.
In terms of things I could learn from my set, I reviewed my designs with the benefit of hindsight, looked at all the community contacts, and talked to some game design friends about where it could have been better. Of course, a lot of the comments were about power-level concerns, of which I admit there were a few, but those were never my highest priority because this was always more an exercise in initial design than in final design. I also had at least on card where I forgot to put a minion tag on an obvious dragon and that had vestigial text on it from a prior version of the card, which no longer made sense in the final version. Again, these are things that would most likely get caught with more time, more (any) in-game testing, and a few more eyes looking them over. Even if they weren’t, that doesn’t seem like the most grievous of errors, considering that real Hearthstone just made that mistake themselves. The things I’m more concerned about here are more fundamental design issues and similar places where I think I could improve.
And while there’s a lot that I’m proud of in the set–including just the fact that I put in all that work to complete it at all–there’s a lot to learn here as well.
– At the highest level, one issue was with my set theme: “time travel” is, paradoxically, simultaneously too broad and too narrow at the same time. It’s too broad in the sense that it can–and my set does–cover several different time periods and events (how else would you sell time travel?). The problem there is that it makes it hard for people who aren’t the most engaged to track how all the themes and other bits of flavor go together. It makes the set appear disjointed, as it looks like it’s doing a ton of different things, and it’s only when you remember that the theme is time travel that you start wondering how the Iron Horde are in the same set as some characters from the War of the Ancients, or why either of them are in a set with Dragons, etc. It would probably need a stronger, more overt structure–like maybe an adventure where each wing takes place in, and involves using/earning, one of Chromie’s epochs–to make a theme as broad and vague as “time” work.
On the flip-side, the theme was too narrow in that by just making it “time,” I limited myself on flavor as well. I definitely felt this while designing the set, as there were many instances of card names feeling a little clunky, or repetitive, because I couldn’t think up any more puns or aphorisms concerning the concept of time. It was pointed out to me that Hearthstone tends to do two flavor themes for each set (Karazhan, but disco; dragons having an air battle; trolls/loa and wrestling, etc.). Doing that in the future would: 1) make my fan sets feel more Hearthstoney; and 2) help concentrate the set’s theme on those points, which will simultaneously give me more to work with and also make my set feel more streamlined.
– As mentioned above in discussing Reno, and in my twitter thread on “poof,” I had an overall issue with plain text in my set. As discussed in the twitter thread, digital card games like Hearthstone have a powerful tool that allows them to sometimes use less precise/clear language: the game itself. If you want to find out what a card does, and you aren’t clear from the text of it, you can usually just play it and the game will show you! But, I leaned a bit too heavily on that tool in my set. Too much of this tool adds a barrier to entry and player frustration, so traditionally, Hearthstone only does it a few times per set. I had too much of it, and it’s something I’ll try to keep an eye on for future designs.
– Another big issue I had was with the role that some of my card designs played. To explain: it is generally the role of the evergreen sets to establish class themes, archetypes, etc., whereas it is the role of the expansions to iterate on those in new and interesting ways. Obviously, I knew this going into designing my set, but I didn’t quite realize how Hearthstone tends to make this work. What Hearthstone tends to do is to rely on the evergreen sets to establish the “core,” workhorse cards for most decks in a class, and then use the expansion cards for the more exciting stuff like build-arounds and finishers.
Who’s having the fun?
Rogue and Druid are good examples of this: both classes tend to have at least one archetype in most metas that uses 10-15 evergreen cards to build a backbone for the deck, and then which uses expansion cards (C’Thun, Jades, Pirates, Galakrond, etc.) to fill in the rest of the deck. The expansion cards tend to give the deck its name, and make it feel new, whereas the evergreen cards make sure the deck still feels like a deck from that class. This works best when the finishers are big flashy expansion cards, because the human brain remembers the big flashy thing that killed you more vividly than it remembers the 10 evergreen cards you played to set it up.
(Related tangent: the nerfs to the Druid and Rogue’s evergreen toolkits, and the current calls to maybe look at Rogue’s evergreen toolkit again are not, in my opinion, because this method is inherently wrong, but more because the balance is/was a bit off and because people get tired of seeing the same thing so much even if it’s a good thing–I still want a rotating core, but that’s a discussion for another day).
Back on topic: Rogue is actually a particularly good example because it’s the class in which I probably failed in this the most in that I made a ton of workhorse cards that, I thought, had interesting mechanics and takes on classic Rogue themes. But, none of them really popped.
Looking back at my Rogue designs, and going over them with a smart friend, I was reminded of a related design lesson that, it appears, I lost track of while designing my set. Legendary game designer Sid Meier has a famous design principle which is often phrased as a key question every designer should ask themselves when designing their game: “Who’s having the fun?”
The premise of the design principle is that it should be the player of a game who is having the fun, not the game designer, and certainly not the computer. I know, players having fun while playing your game, what a novel concept, but it’s really easy to fall into the trap of really clever designs that exist more because you wanted to do something tricky than to serve the ends of the game.
Let’s take a look at my Rogue set:
Timeshifting Recruiter is yet another card I designed that falls into the category of “probably just a worse version of Zephrys,” even if costed appropriately, so go ahead and ignore that one. If we look at the rest of my cards, though, there are a lot of clever things here: I play with the set’s theme and new keyword mechanic, I have cards that reduce in cost if you do rogue-y stuff, and I have cards that either essentially (Sleight of Hand, Added Insult) or actually (Time to Think) do nothing unless you have the right situation for them. All very tricky, and very on-brand for Rogue, I’d say. I also have a few cards that are just solid and if/when tuned appropriately, would probably see some play.
The problem is that none of these cards are the finishers, build-arounds, or otherwise flashy cards that get the average player excited! Even though established players might look at stuff like Time to Think, Added Insult, and Sleight of Hand and see nutty tools for Miracle Rogue, even they would probably really enjoy that for a few days and then start complaining about how “the same” Miracle Rogue deck is “always good.” There are some good tempo tools there, too, but they are also more the workhorse cards than the flashy cards people remember. And Rogue is just the best/worst example of my designing this way in the set.
My set had a lot of solid cards that–once/if balanced for it–would probably see some play and fill in some decklists, but which wouldn’t “pop” in the way we like expansion cards to. I’m not saying that every expansion card has to be flashy–that’s obviously not true–but I do think I felt short of the mark on how many of those marquee, pack-selling, can’t-believe-they-printed-that-card cards I included in my set.
– There were a few mechanics in my set that didn’t quite work the way I wanted them to with how Hearthstone normally works. Obviously, I know Hearthstone better than most, and I tried to be very mindful of how cards would actually work in the game. If you go back to my original article, my game design document has several notes on specific cards that I anticipated raising questions in development, and tried to work them out myself. But, that said, there were some that I missed. A good example of this is in some of my Hunter cards:
The interesting gameplay that I was going after was forcing the player to either take a gamble or build their deck in a certain way (with lots of spells, or beasts, or Scout cards, as the case may be, to maximize the chances of getting the bonuses). The problem with that is that, in Hearthstone, we get a glowing yellow reminder aura whenever conditional effects are active. Oops! The mystery is gone and the interesting gameplay is ruined. It still kind of works in a time set, because now it’s like these cards are telling me the future, but that’s not the flavor I was going for at all with these cards, and it’s a mechanic that I think would probably work better in another class. Really, I’d prefer to noodle with my original concept. Luckily, there is a solution as to these specific cards: Make the effects “Scout 1 [card from your deck] and then [do effect] if the top card is [requirement].”
Another example, which would take a bit more time and energy to fix, is the set’s theme/mechanic (mostly in Shaman) of changing and/or augmenting minion Types. My friends pointed out that this may be something that the game client simply is not equipped to do at this point, since we don’t currently have anything that augments the tribal tag of minions–an area of the card that is explicitly not considered the rules text (for instance, you don’t Silence a minion’s Type). That probably doesn’t mean that the designs can’t happen at all, but it definitely means there needs to be a cost-benefit analysis to determine if it’s worth the time and energy it’ll take to get the system there. That’s always something to consider when adding new mechanics to the game.
Yet another example: Lynessa, my Paladin Legendary, doesn’t work the way I wanted it to because of something I expect I’m not the first person to fail to account for–localization! The “Blessings” are all unrelated cards from various different sets spanning all the way from Basic to Saviors of Uldum and, in some languages, they don’t all have the same term in their names! We can’t reference cards by name if the names aren’t uniform, at least not without re-naming all the outliers in other languages, which is a big, embarrassing undertaking for just one card.
There are also probably some other issues with the card, like the name (how much cringe is too much cringe?), the balance, and the variance between hitting a Blessing of Kings and a Blessing of Wisdom (also, does the Druid card Blessing of Ancients count?), but the localization hurdle is the biggest and most immediate. The solution would be make all the cards she can cast in this set, either as collectible cards or not, and make it clear that they are to be under the same umbrella, so the localization teams treat them as such and the card makes a lot more sense to everyone, no matter the language. This would also help with the balance and variance issues, since we could specifically design all the results to be more in line with one another. Anyway, always remember: think globally.
– If we get into the smallest minutiae of specific cards, there were definitely plenty that my buddy and I spitballed a bit and came up with better designs for. I’m not going to get into each and every one, but it just goes to show how helpful it is to work with a team of smart, like-minded people. I hear a lot that game design–and perhaps especially card game design–is about iterating on good ideas, and these conversations have shown me first-hand how important that could be. Here are a few more quick examples of things we iterated on:
- After messing with a few cards and finding them to be too wordy–or finding my excuse for why I didn’t make a change to be card space–my friend suggested that we make the default for Scout be from your own deck and to only use “from __ deck” when you either- must or get the opportunity to Scout from your opponent’s deck (similar to how Discover defaults to your own class, but can sometimes Discover from different pools). This sacrifices a little bit of clarity and consistency in the text, but it does free up a lot of space throughout the set, and I don’t think very many people would be confused by that change. It also works because Scout is more common on your own cards anyway.
- I had a subtheme of Dormant minions, and awakening them. To go with that, I had a Druid card that read, “Awaken all Dormant minions. Give them +2 Attack and Rush.” However, it’s not fun to have cards that too often do nothing, so this card would lead to less overall feelsbad moments, and therefore be a better overall card, if the two effects were not contingent on one another. So the card just gives all minions +2 Attack and Rush, regardless of whether they were just awakened. It still fits the theme and the mechanics still fit well together, we’re just exchanging a little bit of flavor for a lot of gameplay.
- The Hunter cards discussed above were another bunch of cards that I had an idea for but had to iterate on with my buddies to get them to actually work how I wanted in the game.
Alright, so that was a lot of words, and now seems like a good time to stop. I don’t really have a great way to wrap it all together, so, uh, I guess I’ll just end this how Maro ends his: “I’m pulling into work, so you know what that means, this is the end of my drive to work. So instead of making– wait, no, so instead of talking Magic, it’s time for me to be making Magic. I’ll see you guys next time. Buh-bye.”