Designer Diary: Thunderstone

by Mike Elliott
December 3, 2009

Editor’s note: If you haven’t played Thunderstone, you might consider reading the rules before tackling this article in order to understand all the game mechanisms to which Elliott refers.

Most people who know me know me as a trading card game designer. I have been doing trading card games for about 15 years now. I worked on Magic: The Gathering for about ten years back when I was with Wizards of the Coast, working on design and development for about 30 Magic sets. In addition to working on other designer’s games, I have had about a dozen of my trading card game designs published in the U.S. and Japan and a couple of them (Duel Masters and Battle Spirits) are still going quite strong in Japan.

The trading card game industry is a very interesting and complicated market. Between the U.S. and Japan, over 400 trading card games have been published since Richard Garfield created the category back in 1993. While I have not played every single one of these, I have probably read the rules and analyzed the cards on most of them. I also had a short stint with Wizkids/Topps where I got to study a lot of Jordan Weisman’s games and I worked on a number of miniatures games, including a couple of hybrid minis/trading card games. Most trading card games borrow mechanics from other games. I keep track of various mechanics and what trading card games they are used in and I am constantly updating my list. When I do a new trading card game, I often look for mechanics that are under-represented and try to use those as a starting point, adding new material on top. Sometimes I start with a mechanic and work to alter it slightly. Duel Masters, which is my most successful TCG from a market standpoint, started with the premise of “What if you took the prizes from Pokemon and made it so the losing player got them instead of the player scoring the knockout?” That got combined with “What if you attacked the player, but had the cards be your life instead of having a separate numeric life total?” These and various other questions and mechanics tweaks resulted in a new game system. This sort of “Chinese menu” approach happens all the time in trading card games and many of the games end up with very similar mechanics and play styles.

That brings us to November of 2008 when I first saw Dominion. I had met designer Donald X back in 1997, and he played a couple of his spec games with me that year. He did not have Dominion back then in 1997, but all of his games were above average and one of them was exceptionally good. (Sadly not published yet but that might change with his success.) I still have a copy of the stickers that Donald X made as part of a homemade Magic set, which included a number of cards that were fairly innovative for the time. The stickers are actually a collector’s item these days, and a number of the concepts in them are seen on real cards. He came back a few more times over the years, but I got to play with him only one more time in that period.

The shorter version of this story is that he is a brilliant designer who is finally now getting his deserved recognition. I watch for new card games and trading card games, and I was told about Dominion by one of my gaming friends who saw it at Essen. He gave me his copy and I was surprised to see Donald X’s name on the cover. I played a couple of games and was amazed at how elegant and streamlined the game was and how it captured the replay value of trading card games with a much more limited card set. I played a few more times in November and December. Having worked so long in design and development and having played so many trading card games, I notice quickly when games have a common path to victory. In over half the games, the strategy seemed to be to note the “engine cards” that generated coins and cards and race to start buying provinces.

Once you work for a while in the industry, it is hard to play games without going into “work” mode. The game kept going around and around in my head. The game had an amazingly elegant execution, and my normal game analysis process of stripping out bad pieces did not apply. There were really no extra bits thrown on that did not fit. The only things that seemed like potential areas of improvement were:

1. The victory condition: While the beginning layout had a large amount of variation, the convergence towards Province buying seemed to fight the infinite permutations concept of the game.

2. The theme: The cards have a medieval town naming pattern, but the game plays on a purely mechanics level. A large number of core hobby gamers couldn’t care less about flavor and theme for most games and just really want games with strong mechanics and high replay value, but there are also a large number of core hobby gamers who like the immersive aspect that many games have. Who hasn’t played Tichu and thought, “I really feel like I am the dog in this game”? (That was a joke. Game designers are very sarcastic individuals, which comes from having about 75 percent of your design work end up unpublished in your garage gathering dust.)

3. The limited interactivity: There are not many games where one player can stop playing in the middle of the game and the other players can easily go on without them.

Thinking about the theme led me down the fantasy path since it is one of the most popular themes in gaming and Dungeons and Dragons had a pivotal role in creating the core hobby gaming industry of today. From there it was a quick jump to “What if you fought monsters instead of buying victory points?” and then, “What if the monsters you fought changed every game?”

I have a card creator template that I use for trading card games, and I quickly started to have a lot of fun with the theme. I have played a number of online fantasy games, and I always chuckle when I see Kingdom of Loathing, which shows you can have a decent game system without having high end textured graphics like all those MMORPGs I have been addicted to over the years.

Deck building games build in concept from the various draft formats for trading card games. The premise in deck building games is that you start off really bad and eventually end up with the “killer” deck that wins you the game. To highlight that, the original starting cards were Peasants (which became Militia), Broken Stick (Dagger), Candle (Torch), and Dead Rat (Iron Rations). Because I wanted to drive home how bad the player’s starting hand was, the version featured some hand-drawn custom stick figure art. I draw a very evocative stick figure version of Broken Stick.

I had the concept of what the monster deck was going to do early on. The monster deck concept was based on all the boss fights in various games like Diablo and WOW. The idea was that you would fight a bunch of monsters and then at some point “The Boss” would come out and wipe out your party, ending your night’s play. Seemed like a fitting end for all your hard work building up your party. The idea of shuffling in the end condition cards goes back to a number of older card deck games and it seemed like a good fit for the end condition. Three sets of monsters with between 20 and 30 appearing seemed like it would hit around the target game length that I wanted.

The light concept came out of the monster deck layout. I had settled on three ranks and wanted to have some way of making the monsters in the back harder to kill than the monsters in the front, so that as they moved up they became easier to fight. Light seemed like the most reasonable explanation. I was also thinking about design space at this point and wanted to have more than just one attack parameter. Having one that was unlimited (Attack) and one that was semi-limited and that you could overspend on (Light) seemed like a reasonable way to mix up the attack stats a little. After reading through a number of role-playing books again, I threw in Magic Attack also.

The heroes were inspired by all of the fantasy role playing games I have played over the years. I boiled down the basic classes to a few abilities. Clerics had spell casting and card drawing, and could cure diseases. Diseases actually came out of reading the books as well and served as a different catch-up feature than killing adventurers or the normal gumming up of your deck by killing monsters. Magic users could cast spells and do various manipulative things with deck order, monster order, etc. Thieves could deprive opponents of cards and were the only adventurers who produced gold, and fighters had the best attack value and could carry big weapons. The food concept to boost attack was sort of taken from the old Gauntlet video game (Warrior needs food …badly) and I ended up tying that in with the fighters as well with the “higher attack if you eat food” concept. The idea of having weapons led to the strength mechanic.

There were four initial types of cards: adventurers, spells, items, and weapons. Adventurers had good attack values and sometimes light, which I was envisioning along the lines of infra-vision, and other than Thieves, did not generate gold. Weapons had the best gold generation and good attack values, but were conditional on being attached to an adventurer and the better the attack values, the more restrictive it was in attachment requirement. Spells and village cards (which were essentially town spells) had situational effects and occasionally damage effects. Some were originally flavored as potions and had a gold value, since a potion was a physical object you could sell, which was the primary basis for a card generating gold.

I printed up the cards and goldfished it a couple of times. I then set it aside and worked up three other card buy games over the next couple of weeks, each starting with a different base premise and each having different secondary mechanics. And then they all sat on my shelf for a few months. You see, I am mainly a trading card game designer, and most of my success is in Japan, so here in the U.S. I am not particularly well known, don’t have a lot of fans, and at the beginning of the year did not have a lot of industry contacts among companies that publish this style of game. Since I had a lot of trading card game work, I was not overly motivated to try to push any of these games. I was at GAMA in April with Bandai, which brought over my Battle Spirits game from Japan to the U.S. and was introducing it to retailers and distributors at the show in Las Vegas. I ran into John Zinzer from Alderac Entertainment Group (AEG) and we chatted a bit about the game industry. He mentioned that AEG was starting to branch out into more box-style games like Tomb since new trading card games were a hard sell in the U.S. these days. He brought up Dominion, and I mentioned I had worked up a few games in the same style. He asked if I would be willing to show him them and I explained that they had not really been played and were pretty rough. I indicated the best one was probably the dungeon-themed one that I had codenamed Dungeons of Krell (a made-up town name). I told John if he was really interested I could probably polish it up, play it with a few of my play groups, and send him a copy in a couple of weeks.

So I went back to Seattle and pulled Dungeons of Krell off the shelf. I rounded out the monster groups and tried to give each one a unique flavor that would make it play differently in games. I mentioned before that one of the issues with Dominion was the low level of interaction. I played around with a few ways to spice up the interaction, but several of them did not fit the dungeon crawl theme and most of them made the game less friendly for beginning players. In this type of game, most interaction cards are good for one player and bad for everyone else, and many of them lead to situations where the other players are forced to pass or have bad turns due to lack of cards or a gummed up deck. Because I was going for a dungeon crawl theme and I thought most of the easy interactions would feel like you were fighting your fellow players and not the dungeon, I decided to try to play up the resource scarcity fight and the race to get the monsters as the main interactive element.

The first version did not have the character leveling. jim pinto, a solid designer and developer from the AEG group, commented after testing with his group that it would be cool if the heroes somehow got better as the game went on. This led to basic and advanced heroes, then to level 1 through 3 heroes fairly quickly. It required adding another stat since there was not really an easy way to tie it to victory points (VPs). It also allowed another knob for monster costing since you could now have high XP low VP monsters and low XP high VP monsters and everything in between. The numbers of cards in each level got tweaked and eventually we settled on the 6-4-2 ratio. This actually became a very interesting resource race in the later playtest games. It did, however, have a negative effect on the five-player games because of the limited quantity of the level 3 heroes. In three- and four-player games, usually only one or two players go after a particular hero type, but in five-player games one or two players would often end up getting locked out from the level 3 heroes. To balance that out, the ratios for a five-player game would probably be better at 8-6-3. The five-player situation is further exacerbated by the fact that with all the players buying heroes, the level 2 and 3 heroes get uncovered more quickly. In various iterations, the level 2 and 3 heroes were not buyable, but required leveling up to acquire them. If you want a different play experience, it is an interesting variant, but I felt the tension between going to the dungeon and going to the village to level up a hero in your hand was preferable and led to more choices in the game, especially when there was a race to get the higher level heroes. A player might have a hand that could kill a middle range monster, but might be worried about missing out on the hero leveling race so might instead go to the village. It also let players that had gone a more economic route early have a shot of catching up by directly buying the higher-end heroes once the level 1 heroes were bought out. I felt the final solution of having the higher level heroes buyable but at a higher cost was the best option and led to more balanced game scores. In the straight XP leveling version it was often a rich-get-richer scenario, where the players killing the monsters early got the XP, leveled up and ran away with the game. The games often play a lot differently when Trainer or Pawnbroker is in the game since players can use both of these cards to get rid of militia and buy or level up other cards.

Spells went through an interesting evolution. Originally you had to have a Cleric or Wizard to cast spells. Unlike weapons, which key to a stat that every hero including the starting militia heroes at least possess, you can get random layouts with no Clerics or Wizards and enforcing a class-type restriction on the setup did not seem to be a good choice. So I changed the game to give you one spell for free and Clerics and Wizards let you play multiples. However, there were not many situations where you wanted to play two spells because with six cards, you rarely had two or more spells, and when you did it had the same issue that you felt cheated without a Cleric or Wizard. While it is important to have players work for combos and have some amount of misses where the combo does not fire, this was already present with the strength system, where you could often get a weapon and not have the correct hero who had the strength to wield it, or did not have the combination of weak heroes and items that boosted strength to wield the item. Too many misses leaves a bad taste in players’ mouths and they have a bad game experience, so I ended up stripping the spell requirement. This required modifying the flavor a little since it now meant that instead of just being a player controller, you were now a spell-casting leader that went along with your party. Since most RPG games involve single character activity, almost anything I did flavorwise was going to be a slight disconnect for traditional RPG fans, but I felt the solution preserved as much of the flavor as possible without having to either abandon the spell card group or make them unpalatable to play to preserve flavor.

Village cards went through a similar evolution. Because they were worthless in the dungeon, many players did not buy them even though they were good early because they gummed up your hand when you starting hitting the dungeon a lot. Even with better and better effects, the village cards were still not seeing play. The costing did not matter since usually you had sufficient gold to buy better cards and with only one purchase per turn, these often got cut out. The trick to making them playable ended up being the ability to ditch them midgame after a few uses. Now they no longer gummed up your hand when you went to the dungeon and you could get on with fighting monsters. In an ideal situation, you want to buy them early, then trash them whenever you feel you are pretty much done with the town. The tension is whether or not you can figure out when this point is and do it before your opponents start hitting the dungeon a lot.

The weapons actually borrow a page from creature enchantments in Magic, where you have to have a creature out to play them and they can often be dead cards if you have nothing to play them on. The talented staff at WOTC has to go to great lengths to make this type of card attractive and playable, and balancing the power of weapons versus the amount of times you could equip them was a tough balancing act. Weapons ended up being the most efficient gold sources as well. A common zero level early strategy if you don’t know anything about the monsters or don’t know what the strong cards in the layout are is to buy weapons early and use the gold to buy the top end cards.

The game got unusual playtest results. Between the light system and the strength system there is a lot of math going on, and a number of players had trouble calculating whether they could attack a monster and it led to frustration early in some games. Several of the developers at AEG repeatedly tried to convince me to get rid of strength and potentially replace it with a keyword system, and get rid of light and just have an attack penalty based on being in the back rank. Because the game does not have the engine-style card drawing of Dominion, I thought it was important to keep the interactive card combos, such as the Militia plus Iron Rations letting me wield a Short Sword. Everyone feels good when they pull off a combo like this and take down a monster. Sure you sometimes don’t get the combo, but that often makes it that much better when you do.

The types of gamers testing it had differing results as well. For players who had never played Dominion, learning the card buy elements on top of all of the various light, strength, and monster effects was daunting. If the players survived past about the 6th or 7th turn, they general loved the game and wanted to play again. However, several of them had a desire to stop after the first few turns. Among Dominion players, the really hardcore Dominion fans had a wide variety of reactions. Most loved the variable victory condition. Many disliked that the game was more complicated and took longer to play, and this comment tended to track with players who did not care about flavor closely. If you did not mind the very lightly themed, extremely streamlined presentation of Dominion, you tended to have a more negative reaction to Thunderstone. The more RPG leanings you had, the more you forgave the additional complexity and layers that came with the additional flavor that let you bash monsters instead of collecting gold to buy land. I fully expected comparison to Dominion and because of the elegant design was unlikely to come out favorably going head-to-head in the same gaming space, so I elected to go with a longer game, the dual resource track, and the variable and interactive victory condition which often fought back. Between these differences and the flavor, along with the character XP and leveling system, I felt the game fit into a different space than Dominion and might have a niche in the gaming community in a spot that Dominion was not occupying. The reactions are definitely mixed, ranging from players who think the game is innovative to players who think it tracks too closely to Dominion.

The game actually had the complexity reduced during development. At one point there were more interactive monster effects. These were called global effects and had an effect while the monster was in the hallway. The problem was that everyone had to pay very close attention to the monsters in the hallway, even when they were not considering going to the dungeon. As such, a lot of these effects got missed, and a couple of turns later someone would realize they had misplayed due to missing a global effect on a monster. They were also just one more thing that was confusing to beginning players, so all of the straight global effects got moved out of the base set. The only remnants are the breach effects, which were intended to represent the monsters coming out of the hallway and attacking the players and the village.

A couple of elements were added to attempt to speed up the game because of the large randomness of the setups. The rest option, which is often missed especially by the Dominion players, allows you to take a turn off and get rid of a weak or negative card in your hand. In games without Clerics, you can often do this to get rid of disease cards, vastly improving your probability of getting successful draws on later passes through your deck. This is a tactic that players typically pick up after a few games since early on it is not clear whether buying a marginal card is better or worse than getting rid of a starting card. In some layouts, I will pass early to get rid of Daggers or Militia if there is not a reasonable strategy on the board to get rid of these weaker cards that you don’t want to draw late game.

The other added element was the ability to attack a monster to essentially get rid of it. Occasionally games will start with several large monsters out and the buildup phase will be longer. In some cases it can be advantageous to attack a particular monster even though you cannot defeat it. Some monsters destroy heroes when you attack them, so you can make the equivalent of a rest by attacking one and getting rid of one of your weak starting militia cards, while also burying the monster on the bottom of the deck. While in a four- or five-player game your benefit from this is less, in a two-player game, you will often use this tactic to deprive your opponent of a high point target if you cannot take it down. If you think your opponent is in position to take advantage of a high victory point monster, attacking it to get it off the table is a great strategy.

Monster ordering is another element that was considered and I have seen the concept work reasonably well in other games like Tikal. However, the playtest groups that played a lot of games liked the fact that some games started with a big monster jam and required everyone to build up a lot and other games had weaker monsters out early and players could kill them quickly. It also meant that you could never be sure whether any of the big point top monsters in each group would even come out in a game and overall just made the victory track play out differently, even when using the same monster sets from game to game. It is certainly possible to stage the monsters, picking out the weaker ones to shuffle into the first ten cards or so, but doing so in the rules would have required additional setup and would have somewhat reduced the variation that the random order produces. Granted, from a story perspective it makes sense to battle the weaker monsters first, and I will not be surprised if some players try out that variant as well, but the random order seemed to provide a higher replay value. While I am certain to be second guessed on this point, I think the random version leads to more diverse games, and since one of my goals was to not have a convergent victory condition, I was swayed to the full random path.

The game flavor got changed almost immediately. While it was cute for playtesting to have your Orc with a Broken Stick eat a Dead Rat to get enough attack to kill the Vampire, humor does not go over as well as a nicely textured fantasy world, and it would have been a crime to trivialize the great and quite serious art of Jason Engle by committing it to a comic setting. “The Boss” who came to wipe the party out got changed to the Thunderstone that you captured to save the village, and there was much rejoicing.

The Thunderstone was originally not worth any points since it was an unbeatable monster. It was originally just a tiebreak for the winner. With the flavor change to an object you were trying to get, a number of the playtesters thought it should be worth something, so I ended up making it worth around what a good low point monster was worth. I had a grand scheme where each set would come out with a new Thunderstone, and players could play a campaign-style game in which they played multiple consecutive games. In this format, the winner of the Thunderstone in the first game would start the next game with the Thunderstone in their deck, along with the normal 12 starting deck cards. Some of the Thunderstones might have different effects to allow for this sequential game play option. One might be worth 1 VP and give 1 light in your deck. One might be worth 1 VP and 1 gold. It basically starts the previous winner with a VP card that has a gumming effect on the player’s deck as well. It seemed interesting, but probably not a lot of players will play that way and it is just a crazy designer idea. Players will, however, probably play longer games with multiple Thunderstones and more than three monster sets.

The graphic design elements went through several iterations. Because the game had such amazing art, it was important to showcase it as much as possible. The feeling was that most core hobby gamers can pick up icons fairly quickly and there are several games on the market that depend on this fact. While it was not ideal for beginning players and is probably one of the more confusing elements for beginners next to light and strength, we ended up going with the smaller icons to allow for a larger art space. The one that probably gets missed the most is the light icon, which appears on several monsters, heroes, spells and weapons in addition to the light items.

Whether there is space in gaming groups for a game like Thunderstone remains to be seen. Where Dominion is often a 20-30 minute game, Thunderstone is usually a 30-50 minute game and has a more complicated and longer setup. The hope was that with the variable victory condition, two track resource system, character leveling system, and the various other flavor changes that the game would hit a different segment of players much in the way that the four hundred or so trading card games that have come out attempt to capture different market segments and games in other categories like worker placement attempt to appeal to either different gamers or gamers who have played through a particular game and are looking for a different play experience that is still in one of their preferred styles of games. There are a lot of gamers that no matter how good your game is, they eventually burn through playing it and look to move on to the newer games in the market.

Thunderstone also has a solo play option in which you are racing against advancing monsters, trying to kill a greater number of points worth than have passed by when the Thunderstone comes out. Having come from a trading card game environment where many games are often not that innovative from their predecessors, these changes seemed like reasonable innovations, especially in a field where you often see games like Tichu and Werewolf, which despite the fact that they are derived heavily from Zheng Fen and Mafia, did not seem to inhibit them from having their own niche in the gaming world and from being the target of newer innovations. All games borrow material from other sources, and as I have said before: In the gaming world there is a fine line between what is perceived as “innovation” and what is perceived as an “unethical hack”. Hopefully players playing Thunderstone will recognize that it was not an attempt to hit the same game space but was instead an attempt to carve out a potentially unfilled niche using both similar and different parameters. Obviously I could have gone farther away and combined more static elements, but that would have taken it farther away from the trading card style game and more into the classic board game realm, and there are many, many strong classic board games, but very little in the area of card buy trading card games.

Lastly, the game has a lot of expansion potential. Usually trading card games put out about 400 to 500 cards a year. With these games, you get to pick just the best and most interesting cards to do, and you just don’t print all the filler cards that you would normally print for a trading card game. This is especially true for expansions, where you do not have to worry as much about complexity for new players just entering the game. It was really fun working on the base set, and it is even more fun working on the expansion, where we are forced to trim down on what we can put in a set from a ton of really good material. Almost anything from any fantasy role playing game can be adapted to fit into the game system, from traps to treasure to multiple monster encounters. The dungeon itself could even potentially change to a dimensional rift or a forest or mountain setting. If there is enough interest and popularity, you might even see the system adapted to an existing MMORPG or RPG world. These games really do have a lot of the positives of trading card games without the baggage of the collectability model and large card set sizes. At some point designers will look at various other models, but I think there is a lot of design space in just the trading card versions left to be explored, especially focusing more on direct player interaction.

---

Comments:

To comment, you must register with BGN. Registration is free, but donations are greatly appreciated!

Cool beans.  I look forward to trying the game out.  I am a big fan of Dominion and also like RPGs and games with theme.  (I think I like Dominion so much because of the “growth” of my deck over time… this game looks to have the same potential.) The complexity does take things up a notch, which makes it less likely for me to pull it out in all situations, but we’ll have to see how well it goes over with less experienced gamers. It’s always nice to see the thought and development put into a game like this and have it help dispel any notion people might have of it as a cheap Dominion knock-off. The one thing I currently worry about most in Dominion is turn order… I don’t see anything in the rules or development that address the first player’s slight advantage.  (However, in a longer game, I would assume we’re going more rounds and thus the first player advantage will be slightly less.) Posted by Matt J. Carlson on Dec 3, 2009 at 12:22 PM | #

Oh, and put me down for one copy of the don’t take yourself seriously version of the game… I want to fight (and lose) at the end of the game to the Big Bad Boss with my lvl 3 Little Orc and his broken stick.  (Perhaps a nice tie-in with Kingdom of Loathing or something… ?) Posted by Matt J. Carlson on Dec 3, 2009 at 01:28 PM | #

Very interesting history!  Do you have any idea of Donald X has seen the game yet?  Just curious if he had any impressions of it.  It sounds interesting for sure, and one I will very likely check out. Posted by vandemonium on Dec 3, 2009 at 10:00 PM | #