Smaller Units of Randomness

Slay the Spire gives the player many small decisions where Hearthstone’s dungeon run uses a few larger decisions. The big swings are more exciting but ultimately lead to more frustration, as losing one big roll ends your run while you can easily recover from a few bad small rolls.

For deckbuilding in either game, the decision you are making at any given time is about the same: pick one of three. The total amount you are deciding is roughly comparable, a couple dozen cards. Hearthstone groups those into big chunks, where Slay the Spire lets you pick cards individually, or skip if you don’t like those options, or collect gold to shop from larger selection pools, or remove or upgrade cards. You get more options and you feel more in control. I cannot tell you how many times I picked a Hearthstone artifact or card pool and then got almost nothing to go with it.

High stakes are exciting! Talent trees that give you 1% bonuses each level are boring! But small decisions are more impactful when the big decisions are mutable. Adding Arthas to my deck is more exciting than upgrading a card in Slay the Spire, but the choice of whether to pick Arthas is usually overdetermined. Maybe his card pool is clearly the best, or I already have a Cloak of Invisibility so picking a taunt pool is clearly bad… What I am getting at is that there is not much deciding going on, so much as recognizing what works well or badly given what was available/decided previously. You make exactly 11 deckbuilding decisions in a Hearthstone dungeon run, of which maybe 2 or 3 have any tension.

The stack of the deck is another source of randomness. In Hearthstone, you get 1 card per round, and you do not expect to see all your cards in a typical game. If you get your dungeon run artifacts in the first few turns, you win. if they are on the bottom of your deck, they don’t matter at all. You lose to A. F. Kay if you draw the wrong cards. In Slay the Spire, you go through your deck several times per fight. There are lower stakes to the randomization, and you can lower it even further by keeping your deck size small. That shifts importance back towards your decisions, not the luck of the draw.

These two build on each other. That small decision to upgrade a Spire card matters a lot when you play that card several times per fight, versus a few times at all in a Hearthstone dungeon run. Picking your Hearthstone artifacts is the big, exciting decision on getting powerful toys, but the second one happens so late in the dungeon that you use it at most twice and often never, due to the stack of the deck. I upgraded a Spire card just before walking into the final boss fight, and then I played it six times.

High variability with high stakes pushes towards “go big or go home.” But the randomness determines which of those happens.

: Zubon

Higher in the Spire

After many dungeon runs in Hearthstone, I decided that I liked this idea of roguelike deckbuilder games and picked up Slay the Spire. It was exactly what I wanted, and I immediately enjoyed a four-hour binge.

A great virtue of Slay the Spire is that it is designed to be what it is. The Hearthstone dungeon run is grafting the idea onto an existing game, which has been done quite well in Blizzard games (tower defense, DOTA) but ultimately bears its full flower in a dedicated game. Hearthstone’s dungeon run starts with much greater resources in terms of art, already developed cards and mechanics, and a minion-based combat mechanic. Hearthstone’s dungeon run is bigger and bolder, flashier, and more random and frustrating. Bringing along the infrastructure of Hearthstone brings along the baggage of Hearthstone.

Slay the Spire almost certainly must have configurations that are impossible. It is a roguelike still in early release. It seems to have less opportunity to stack perverse randomness on top of more perverse randomness. With practice, I have become rather good at the Hearthstone dungeon run, but I would be surprised if I could sustainably win more than a third of the time. I get the sense that Slay the Spire takes the same idea and gives it a much higher skill cap, along with more manageable randomness.

: Zubon

One Mana Cards

Low cost cards are one of the interesting balance decisions in Hearthstone. The real cost of a card is its mana cost plus any built-in penalties plus the fact that it is a card.

As an example of the first two, Squirming Tentacle is a 2/4 minion with taunt that costs 3 mana, and Vulgar Homunculus is a 2/4 minion with taunt that costs 2 mana and does 2 damage to you when you summon it. There are many ways to have costs, and there is synergy in building decks that avoid costs.

The cost of the card itself is a reason why 1-mana cards are more powerful than you would expect. The mana is less of a cost than the fact that you are using up a card. It is an opportunity cost; those cheap cards are only great in the first few turns or to burn extra mana around your stronger cards. If you have played all your cards by turn 5, drawing a 1-mana card when you have 6 mana available is suboptimal. You get one card per turn, and you have seven spaces for minions on the board.

This means that fast, aggressive (“aggro”) decks will always remain fairly effective, like classic “weenie” decks in Magic the Gathering. There are fairly inexpensive counters to large numbers of cheap cards, but cards that are card-inefficient are likely mana-efficient, or else they are just trash cards you would never use. If I can get a 2/4 minion with a special ability for 3 mana and 1 card, I better do better than a vanilla 1/2 minion for 1 mana and 1 card. The only way to keep cheap cards from being overpowered for their mana value is to make them underpowered for their card value.

The full cost of a card includes the card itself. This adds value back to cards that include a draw, although deck-thinning in that way can be a cost or a benefit depending on the context. In Dominion, cards that provide a card plus an action plus something else are popular because they have no opportunity cost that turn; the card refunds both the card and the action you used to play it. But the “plus something else” is usually pretty small unless the card is expensive, and there is still the opportunity cost that you used up the buy for a turn on that Village, when you could have got a Smithy.

: Zubon

The Hedgehog and the Fox

Yesterday‘s quote, “luck is where preparation meets opportunity,” is of disputed origin. There are several possible attributions, with an interesting one being a line that Seneca the Younger attributes to Demetrius the Cynic:

“The best wrestler is not he who has learned thoroughly all the tricks and twists of the art, which are seldom met with in actual wrestling, but he who has well and carefully trained himself in one or two of them, and watches keenly for an opportunity of practising them.”

I feel like I could mine that quote for hours.

For reference, consider the hedgehog and the fox. The fox knows many things, but the hedgehog knows one important thing. Continue reading The Hedgehog and the Fox

Luck Is Where Preparation Meets Opportunity

I am still indulging in a bit of Hearthstone. I am not interested enough to invest the money it would take to be competitive, so I have mostly played modes where your card collection matters little or not at all: Tavern Brawl and Dungeon Run. (I could also play Arena, but there is a gold cost there.)

In both modes, I must say that the skill ceiling is a little higher than advertised. I would not say a lot higher, but there is clearly better and worse, and you can see worse decisions being made. Tavern Brawl may tend towards very messy randomness, but understanding how to use that randomness can be the difference between winning and losing. I have won several games I should have lost because my opponent kept rolling the dice after reaching a decisive lead. As in, they could win simply by attacking, but instead they triggered a random effect that wiped the board. This is a skill important in many games (and outside them): recognizing that you have already won and taking your victory.

Continue reading Luck Is Where Preparation Meets Opportunity

Gain a Random Card

After completing my set of Hearthstone dungeon runs, I have kept on for more completions and many more attempts. The more I play, the more I see a common problem: overuse of the features that online card games allow.

For folks who did not click through for those old posts, let me summarize. When you take a card game online, you get exciting new options that are difficult or impossible in physical card games, like cloning cards, altering them in a variety of ways, and bringing random cards into the game. Hearthstone has many of this last, especially in the dungeon runs: many cards summon other random cards. And we find that once you can do something, you find lots of reasons and opportunities to do it. A narrow option can take a slippery slope to become a central feature.

In one sense, it is pretty cool to spin the wheel and get a random card. So very much of game monetization plays on our desire to open the box and see what’s inside. That is awesome, especially when it works out for you. And then it keeps happening and becomes the deciding factor of the game. To the extent that your decisions are not deciding the outcome of the game, you are watching it happen rather than playing.

This is bad game design. A bit of randomness adds spice, prevents the game from falling into a single solvable state, and rewards flexibility. A lot of randomness makes the player irrelevant.

Hearthstone does that a lot, and the dungeon run seems built around it. The NPC decks are chock full of cards with “summon a random X” effects. Blackseed upgrades minions randomly. Whompwhisker recruits random minions from each deck. King Togwaggle gets random treasures. Xol gets cards from a small random set of beams. Pathmaker Hamm damages random minions, then plays more minions that damage random characters, and kills himself roughly half the time I encounter him. Several bosses multiply the effects of randomness by multiplying cards: doubled battlecries, deathrattles, or minions in a mode with lots of ‘cries and ‘rattles that summon random minions.

Luck can dominate these fights, and you need only one bad run of luck to end your dungeon run. Sometimes you will crush Whompwhisker quickly because he will recruit your biggest minions or ones with interesting deathrattles; other times, he recruits his biggest minions along with your cheapest or your minions that are only worthwhile for their battlecries (recruiting ignores battlecries). Blackseed’s upgrades might create a minion that hurts him, or maybe something really powerful that is normally balanced by a costly battlecry.

Hearthstone seems to recognize this problem and attempt to mitigate it sometimes with discover and choice effects. Those combine randomness with player choice: pick one of three options. They may all be great or lousy or not fit with your deck, but you get a choice. (That is the current tavern brawl: every card in your deck is “discover a spell or minion.”) The same applies to the deckbuilding aspect of dungeon runs: you pick one of three options, like a permanent “discover” effect. You might pick up a quest and then never get options that complete it, or you might start with doubled battlecries and get jade golem cards every time you pick.

When Hearthstone launched, it quickly developed a reputation for low skill gameplay. There were few interesting choices to make and not a lot to decide on any given turn. Expansions have added more options, but also more randomness. “More” is both more options to introduce randomness and a broader range to draw from when rolling that die. For example, a card that summons a 1-cost minion had a limited number of options when the game launched, and now there are pages of 1-cost minions. Randomness stacked on randomness.

These are not bad features! They can be used well! Half a decade ago, when CCGs were going online and just starting to experiment with these kinds of features, it was a wonderful use of a new medium. But the dose makes the poison, and something that is medicine in small amounts can kill you if you have too much.

: Zubon

The Draw of Randomness

I think much of the appeal to randomness comes from shallow understanding/efficacy and this rubber duckie analysis. “Interesting things are in the sweet spot where they make enough sense you can form expectations and not so much sense that your expectations are wholly sufficient and the follow-through completely predictable.” The more you understand and can influence a process, the less desirable randomness is unless it is your only chance. The less understanding you have, the more random results fit your understanding while remaining unpredictable. The less control you have, the more random results are likely to be to your benefit (versus where others might push results).

When I was 12 and exploring D&D books, long tables of random values were awesome. It was a draw from a fixed pool of values, so I could have reasonable expectations but never be sure, with wild swings to make unlikely expectations occasionally viable. I had not really grasped that I could take the setting in my hands, as opposed to following what was laid out in the books, so adding randomness was closer to putting things in my control. It at least took it out of someone else’s control. Plus, young men seem bred to throw their fate to chance and hope for the best.

When I play games now, I like known conditions. I understand, plan, master. Rarely can randomness help me, at most foiling reasonable expectations. It is reasonable to have some range of expectation of how much damage your fireball might do; it is not reasonable that the fireball might cause the target to turn into a pink sheep, or whatever else a wild mage or wand of wonder might do.

Chaos is great for people with no long term plans. They have no intentions to foil, they can always insert more chaos if the results are bad, and at worst they can absolve themselves because they had little control over the outcome. Alternately, there is Littlefinger, who starts behind and inserts chaos to create new opportunities to get ahead. The young are revolutionaries who want to overthrow the current order; the old are invested in the current order, know how to get ahead under it, and have something to lose.

Go to a casino and observe who plays what. Some people study odds and think of themselves as masters of their own fate. They play card games with lots of decisions, preferably ones where you play against other people rather than the house. Some people trust to luck, the big win, the quick score. They are at slot machines or the roulette wheel, where the main decision is how quickly to lose your money.

Randomness prevents knowledge, foils plans, counteracts skill. That is a draw to some.

: Zubon

I remember having a work bowling outing, where every third frame had a different requirement like bowling with your off-hand, between your legs, or with your eyes closed. That was the great equalizer between league players and people who hadn’t bowled in 20 years.

Random Draws

I should probably stop playing Hearthstone, but it is a compelling trainwreck of randomness.

I have completed the dungeon run on five classes. It is frequently unwinnable because you were not offered any cards that work together, or you meet a boss that is overpowered or a hard counter to what you have built, or just random draws of the cards. You can even lose the first fight if Bink the Burglar gets the best possible draw and you get the worst (on some classes). But the dungeon run also offers powerful upgrades, fun combinations, and easy access to cards beyond the reach of new players. When it comes together, you get to do amazing and awful things like using Boots of Haste, doubled battlecries, and Coldlight Oracle to play multiple late game creatures on your first turn. I even had both sides come together: Thaddock the Thief got a perfect draw, completed her quest on turn 2, and cast Crystal Core on turn 3; I still won with an even more overpowered combo.

As a new player, it is hard to play the normal mode with the basic cards you are given, after seeing the dungeon mode. I suppose it could work as an advertisement for people who might drop money on card packs to get those cards they saw.

I played a round of Casual Play over lunch. It advertises, “Find an opponent of equal skill, and play for fun!” I was paired against someone with a golden hero, meaning they had won at least 500 ranked games on just that hero. I know Blizzard’s matchmaking algorithms have a wide margin for “close enough” on player rankings balanced against time to search for a closer match, but it paired “less than a week newbie” with “played at least 1000 games.” Look, maybe it is making a good assessment on “equal skill,” but it is not like my small box of imaginary cards has any chance against the deck they can field.

In this way, Hearthstone lets you feel good about your wins but wave away your losses. If you lost, it must have been an unfair match or a bad luck of the draw. While I might otherwise mock that like FPS players who always die due to lag and hackers, that is entirely valid in Hearthstone! The many stacks of randomness in the game mean you can easily get extreme swings with no difference in player skill or choice. But you can mostly ignore that when you win, because the things you were trying to do came together, and you reap the rewards. You did make some good choices along the way, and no absurd fortune happened to keep you from winning this time. And maybe you can observe some absurd fortune happening to your opponent, but it’s not like you can see their cards. If you had bad luck, it was pretty visible and you are not to blame; if your opponent had bad luck, that information was mostly hidden from you except the results, and it is easy to credit yourself for having won a fair-looking match.

: Zubon

Hearthstone: Deeper in the Dungeon

hearthstone screenshot showing 2 complete dungeon runs in 2 tries I decided to give Hearthstone a little more time by trying out the dungeon run with each class. It seemed like a good way to unlock all the basic cards. And, well, see the screenshot. Either the Druid has an easier time, Hearthstone times rogue-like is as random as suggested, or I am just that good.

: Zubon

I doubt I am just that good.

Hearthstone: Kobolds & Catacombs

The game of the weekend has been Hearthstone. I had never played before, but I was lured by the advertisement of the latest update and the surety that I had a bunch of free bonuses from various Blizzard promos. The new content is Hearthstone’s take on rogue-likes. It is a series of eight battles against increasingly powerful NPC decks, where the player accumulates cards and upgrades as he progresses. Rogue-like elements include facing a variable cast of foes from a set list, variable foe stats based on which “level” you find them, picks from randomized loot bags, and picks from themed, randomly selected pools of cards.

The good is that the design addresses many of the issues seen in rogue-likes. Each upgrade is a pick from three upgrades, and while you may not have the best upgrades on the list, you do get a chance to customize your deck around your preferences or a common theme. The theming of the content is good, a mix of kobolds, adventurers, and monsters.

The bad is that it combines the low-skill play Hearthstone is known for, minus most of the deckbuilding aspect, plus the randomness and degree of fairness you have come to expect from rogue-likes. Some of the NPC decks are vastly harder than others, or vastly harder for some classes or decks. Draw one of those and you will probably lose. Sometimes you will get several upgrade options in a row that build on one another, or maybe none of your three picks fits well with what you have. You may get the perfect options to go with your passive upgrade or few to none that work with it. And then there is the usual trust in the Heart of the Cards that comes with CCGs. The new mode claims massive replayability, which is to say there are many random variations (of varying degrees of difficulty) when you have several layers of randomization.

This leads to unproductive forum discussions where some people got through on their first try and others are hitting walls, or someone tried twice in a row and lost early then got really far without any real change. When the game stacks random enemies that can appear at different levels against several classes that get random items and seven picks from random card sets, plus randomization in the cards and some cards with random effects, the results can be explained at best statistically. All that randomization can even out to the expected experience or something radically off the rails. It needs to work out eight times in a row for the player to clear the dungeon run; the first two or three are free, the last two or three are a bit of a roll of the dice even under perfect play, just from the way some NPC decks are hard counters to player decks.

The dungeon runs are not dependent on which cards you own. The cards are provided along the way (and not kept). This makes it a good way for new players to see the game and be on even footing with veterans, and to unlock the basic cards for each class. To the extent that Hearthstone can be skill-based, this is more skill-based than standard CCG play, minus most of the deck-building aspects. As someone new to the game, it seems more compelling than entertaining. It is very slick and nicely done, with the sort of predictably mediocre gameplay we have come to expect from MMO PvE. It seems like a fine game for decompressing after work or to play absently on a bus. I feel like I have already experienced most of what Hearthstone has to offer after a mix of dungeon runs and normal games over a weekend. I do not suspect it is meant to have deep gameplay for me to discover.

: Zubon