Balatro's 'Cursed' Design Problem

Balatro is the current game design darling. But even the designer is aware of a ‘fundamental design flaw’. What is it, can it be fixed, and what does it teach us about making games for people?

=== Sources and Resources ===

- Sources

[1] [AMA] I am localthunk, developer and artist for Balatro. Ask me anything! | Reddit

[2] GD Column 17: Water Finds a Crack | Designer Notes

[3] 136: Going in Blind with Balatro | Eggplant

[4] Postmortem: McMillen and Himsl's The Binding of Isaac | Game Developer

[5] Isaac vs Mewgenics | Steam

[6] External Item Descriptions | Steam Workshop

=== Credits ===

=== Subtitles ===

[alexanderbrady]

This is also a famous problem in board games, typically referred to as "hidden but trackable information." The classic example is Catan, where player's hands are considered secret, even though every player knows exactly what is gained by each player and what cards each player spends, so a player with a pad of paper or a good memory can remember all of it.

Despite massive player push back, the designer of Catan has held firm on this rule. He said that the game just gets too slow and dull without it, as always knowing exactly what every other player has in hand means players can math out the game too precisely.

Interestingly, instead of allowing players to get around this rule, it is actually enforced in competitive Catan play (and players are not allowed to use paper to track). In effect, it adds a memory or deduction element to the game. According to top Catan players, logical deduction typically gets them close enough that very few players actually bother with trying to track every card draw and discard.

Just an interesting note that despite massive push back from the part of the player base that wants to "optimize" this rule away, it has remained. And that even the competitive community has ended up embracing it.

[Zhon]

Amusingly, mods have already done what you've described: there's a mod out there that adds a bunch of new jokers, and one of them is Scouter - and all it does is show you how many points you're going to score. Making it a Joker is also really interesting because it means you're not using that Joker slot for something that actually provides a mechanical benefit, so there's an inherent trade-off to having that information available.

[TheLogondo]

The thing with Binding of Isaac is, there's literally over 300 different items you can collect. And some of their effects can be pretty niche. I've been using the "item description mod" for as long as I can remember. Makes playing the game so much more fun when you can actually understand what you're getting. Maybe when BoI first came out, not having items described made sense. But now? After like 4 expansions and a ton of free updates? There's too much for any player to possibly remember. And it's just not that fun.

ESPECIALLY because so many power-ups can actually be detrimental to your playthrough. I have had a few runs end because I picked up a power-up that really fucked my guy over.

Imagine in Balatro if your Joker cards didn't tell you what they did until you used them. It would make the game unplayable.

[RBAlejandro]

To me, not being able to preview the score is one of the game's greatest strengths, but the ability to calculate it yourself as a last resort is exactly what makes Balatro an all timer. Sure, you can mostly rely on rough estimation without thinking too hard about everything, but there's also an undeniable magic in the adrenaline of having only 1 hand to cover a ton of points and having to desperately pull out your calculator to figure out what combination of jokers, card order or hand type can get you through.

Plus, even when I do so, there's been so many times when, in the heat of the moment, I fail to take card interactions, debuffs, effects, or boss abilities into consideration in my calculations and end up failing miserably when I was so confident, or crushing the blind when I thought all hope was lost, and it is extremely fun no matter the outcome.

To me, the only thing that would suck the fun out of it is using the dedicated hand calculator, but even then, if someone chooses to play like that and still gets a fun and rewarding experience out of it, then it's still a net positive.

[SystemsInPlayPod]

Cool mechanics and fun gameplay aside, can we talk about the art for the joker cards for a minute?
Look at the subtle smirk on their faces, the tasteful sheen of their foil. Oh my god, they even follow your cursor when you mouse over them.

[reneethefox]

I think Isaac's "figure it out" approach worked at the start because there were less 200 of them and most of them had a a pretty immediate effect. Nowadays there's over 700 of them, and some of them have really weird effects and/or only activate under specific circumstances. Especially with trinkets, I have over 1000 hours in the game and I really just do not remember what half of them do. Without Item Descriptions, a lot of trinkets just become "thing I pick up and then replace with something that I actually know about without ever seeing their effect activate".

I like the method used in The Void Rains Upon Her Heart. Item pick ups are initially kept hidden to the player, but after picking one up, you can unlock their description. Something like that in Isaac would be a nice compromise.

[paraalso]

One possible solution in a game like Binding of Isaac would be to implement an in-game knowledge database you could fill up as you learn mechanics. The game would reflect this in the HUD by only providing you the info that you've already learned.

[MichaelFromTheAttic]

For Isaac, I always thought the easiest thing to do was to add full item description after you tried said item at least once. That way, I feel like players would be more willing to experiment.

[tterhead]

Wait, people calculate their score??? I just cross my fingers and get disappointed if I fucked up and didnt play a good enough hand against the needle.

[gromar]

Something very very similar happened with Rimworld. The game creator, Tynan Silvester, created the game as a story simulator with drama, loss and struggle. But that's not how most people play it. In fact it's most often played as a base building strategy game in which loss is seen as something to be avoided. The game designer decided to push back hard against that kind of playstyle. People made kill boxes for raiders so he nerfed turrets, people responded with traps so he added tunneling enemies which ignored walls so people made the wall into a killbox, so he made enemies which drop down from the sky right into your base and.... you get the idea. The constant arms race has, in my opinion, made the game worse for both types of players so I think it's best to just give the players the freedom to choose how to play is by far the best, possibly with a message explaining the pros and cons like you suggested.

[yeetman]

Fun fact: Balatro is actually an old word for jester or joker, which was used frequently in Ancient Rome

[headphonez]

I love that Balatro doesn't have the score preview for all the reasons you/localthunk have stated, I think an additional advantage is that it allows you to get a different feeling/intuition about each run, instead of having to do a calculation most of the time you can draw on your past experience in the run and just make an approximation of if the hand you're about to play is good enough based on your recent history.

[CNightfire]

On the subject of issac, I think one of the biggest issues with not giving info was the fact you have a fair amount of items with tradeoffs, negative effects, and anti-synergies. So times looking stuff up felt less like "which of these options is optimal" and "will taking this item ruin my run." Ironically, Issac Pills feel like a better implementation of "hidden information" - each run they're randomized, so you can't just wiki it. (and I think once you checked a pill, you'd know what it'd do for the rest of a run)

[wiremesh]

I wonder if McMillen only rented Zelda as a kid? The game came with an instruction booklet that explained all the items.

This does bring up something I think a lot of young people playing retro games don't always understand, which is that the instruction booklet was part of the experience back then. Thanks to memory and technical limitations, there were rarely in-game tutorials, so you were expected to reference the book. Of course, many games were also intentionally obtuse beyond that.

Every time I got a new game, I was always super excited to read the story section of the booklet. I kinda miss that feeling.

[View]

If people knew they would win before gambling, then it would be significantly less popular.

The adrenaline rush of not knowing a result until the ball drops is what keeps people playing. Seeing that before you even hit "Go" would definitely make Balatro far less popular than it currently is.

[curtmack]

When Tunic nerfed an overpowered item combination (increasing its mana cost from 1 to 4), they also added a code that reverts the nerf for 80 seconds, specifically because the speedrun depended on the cheaper mana cost. I can't remember if this code is alluded to in-game.

[Antoine]

The game does not give you your score, but it gives you enough info that you can reliably estimate pretty quick. The scores in Balatro are so bonkers anyway that you rarely need to calculate them precisely, but with a little experience you can ballpark them in a bracket narrow enough that you don't need to worry. And anyways this is not the kind of game where you have a lot of different ways to play the cards in your hand, so even if a hand is not enough there's probably nothing you can do better with what you have. To me the most fun about Balatro is to set this little formula in your head and tweak it to maximise your score, it's a really fun game for people that like mental mathematics and a score display would take that away.

[Mniacks]

In magic the gathering, the information in between "hidden" and "public" is "derived information".

"Public information" is everything a player is required to share to his opponent about his card on demand, things like amount of cards in hand, life total and others.

"Hidden" is every information unkown to one or both player, like what cards are in your hands, what is the order of cards in your library.

"Derived" Is everything you can deduce from public information, if a player returns a card from play to their hand, they are in no obligation to reveal that this card is now in their hand because it's a "hidden" zone, but a skilled opponent will remember this "derived" information.

I think it's a pretty good naming convention for this "in between" for information technically public but not easily available.

[dreamcanvas]

I kinda disagree with your conclusion here. I think the solution Balatro arrived at is awesome, because it's easy to "guesstimate" your score, which is a major strategy factor...but it's hard to precisely calculate it. Often not quite "impossible" but still...Thus, for me at least, the score preview provides guidelines, and I usually find it more interesting and efficient to experiment with different hands at various points in order to test or confirm what actually works best for my current deck and joker loadout.

[Zenith]

i think a score counter is always a bad idea. as someone who follows the “high level” community, balatro is an easy game, especially for optimization. adding a score counter doesn’t help high level players at all, it’s more a crutch for newer players than anything