The Trickster's Hand

Somewhere between a Royal-Flush and an Ace High

Tag: game development

Among Rogues: The Persistence of Death

“You only live once…”

I chose that tagline for RogueLife for a few reasons. In many ways, I chose it because I disagree with it. Yes, put a bullet in the right parts of me, and I’ll run out of that one life, but I’ve also lead many lives. I’ve been a preacher, a stuntman, a corporate drone. Those were whole different lives from the one I lead now. That said, each of those lives carried something with them from the previous lives.

The same is true for the best RogueLikes.

Rogue, the original Rogue, was a game of absolutes and uncertainty. When you died, it would erase your whole save game. It has more in common with coin-operated arcade games than it does with modern, popular role playing games. This facet of the game – the permanence of death – is one of its most memorable features while also being its worst.

Rogue, while definitely a cult classic, is not a commercial success by any of today’s standards. Part of this is because of its unfriendliness. It treats its players like garbage, and it revels in this treatment. It’s unfair. Like, it’s really unfair. For example, potions are necessary to survive the dungeon, but which color of potion corresponds to healing is randomized each time you create a character. Aside from healing, potions can also be poison. Experimentation is both necessary and punished.

It’s real bad, and I hate it.

Rewards only feel good when they feel earned. The same is true from punishment. In Rogue, nothing felt earned. It’s hard to crow about victory when you were always one bad roll from losing.

These last two weeks, I’ve been playing Enter the Gungeon and Sunless Sea. They have very different approaches to the concept of death.

Enter the Gungeon is very much in that coin-op shooter territory. Lives can be measured in bullets, and – initially – it appears that there is no continuity at all. You get one life, and that’s it. Now select a character and go again.

However, you start accruing a currency when you defeat bosses. That currency seems useless until you rescue the shopkeepers and unlock the ability to buy new weapons to appear in later runs. Slowly, other characters and events show up which start to affect later runs. Each run – or every few runs – players feel like they have a little more of a chance to kill their past and defeat the ruler of the gungeon.

No, there is no “reloading” after a failed run. You can’t save-scum. It still feels like a RogueLike even though it violates the most basic tenet of the Berlin Interpretation.

(I’m not going to hop on that steaming pile of garbage right now. I’ve got something special in mind for good ol’ Berlin. You have to wait for that. Sorry.)

Sunless Sea takes a different approach. It tells you from the start that your captain will die. It tells you that is alright, and it even shows you ways to improve the inheritance you can leave for later captains. Each time your captain dies, you are asked what your next captain’s relationship to them is. Were they lovers? Rivals? Kinsfolk?

This direct and open continuity feels different from Enter the Gungeon, and it can be crueler in some ways. A life in Sunless Sea can last hours, while a life in the Gungeon will rarely go for even half an hour. The near-total loss of progress in the Gungeon feels acceptable for its session investment. Sunless Sea’s losses can be much larger, and much more cruel. In fact, one short life after a series of long ones can lead to a loss of progress in Sunless Sea that can’t be rivaled by even Rogue. Sunless Sea will happily put you right back at square one even after twenty hours of progress. Rogue would never do that. Rogue couldn’t. Rogue never had the ability to build an investment like that.

In the coming weeks, I’ll be exploring the interplay of “Persistent Death” and “Variable Success” in greater detail. Both are big parts of the game we’re building at The Crooked Thimble, and both are dear concepts to me as a craftsman.

Tomorrow, I’ll be setting out onto the Unterzee again. Captain Eliza is gone, but a woman of letters has taken her place. Perhaps Penny will fare better. I hope she does.

If you wish to see the tragedy that befell poor Eliza, take a look at our episode below. Also, please take a moment to support our Patreon:

Hidden Information: Poker and Oil

Back in the 1980s and 90s, Exxon and Imperial Oil developed a model for global climate change whose predictive power has met with almost unparalleled success. This was a major breakthrough. The model made many predictions including the relationship between various greenhouse gas emissions and the warming of the overall climate.

The model, being a predictive model, was built to advise its creators on wise investments. With emission levels unchanged, it could predict what land would be destroyed by rising sea levels and which land would become uninhabitable due to the heat. It also predicted what ice would melt and create new habitable land. It could also predict these things for reduced and increased levels of emissions.

The takeaway: Oil rights on the Canadian Beaufort Sea would become valuable indeed as long as emission rates did not decrease.


Hidden information games are among the most compelling games on the market. Most games incorporate hidden information in some form or fashion, otherwise they are a Perfect Information game. Perfect Information games include Chess and Checkers as well as any game where no element is hidden or left to an unknown chance.

Not just any game with hidden information is a Hidden Information game. To be a Hidden Information game, the hiding of the information between players needs to be an important aspect of the game itself. The classic example of this is Poker, and Texas Hold ‘em is a particularly well-crafted specimen.

Texas Hold ‘em contains all the necessary elements of a Hidden Information game: public information, private information, and unknown information.

The private information is the two cards in each player’s hand. The public information, initially, is that there are five cards in the middle of the table, two cards in each private hand, and that there are a certain number of each card in the deck. The unknown information is the values of each of the cards face down in the center.

As the game advances, all unknown information becomes public, then the game ends. That’s the key thing for a hidden information game: the game is over when all information is made public.

Prior to that, though, is the most powerful part of Texas Hold ‘em: investment. Investment is another piece of public information that is entirely created by the players. This, as well as myriad physical and social tells, constitute the strategy. In poker, wagering is where the game is.


Exxon and Imperial Oil had a problem. Their findings were not unique. The information they worked on was public even though the model they developed was still private. They neared the end of the game, and they had not yet turned a profit. They needed to create more information, and they needed that information to be unknown.

Exxon and Imperial launched an unprecedented misinformation campaign. They found spokespeople willing to put their reputations on the line to make that public information back into unknown information. This was the only way to remain in the game. Once global climate change became widely accepted in the public, Exxon and Imperial would no longer be able to leverage their private information to make investments.

And invest they did.

With the debate over climate change fueled by company owned mouthpieces, Exxon and Imperial were free to purchase frozen-over land at rock bottom prices. Without the debate, the land would rise in price – as forecasts showed it growing increasingly desirable – but that didn’t happen. With the melting of the ice in question, Exxon and Imperial were free to swoop in and buy it up on the cheap.

Here’s a tip from one gambler to another, always watch where the other players choose to put their money.


Misinformation has no mechanical equivalent in Texas Hold ‘em. Well. Not in a normal game. For the actions above to have a clear parallel, one of the players would need to conspire with the dealer to stack the deck. More likely, the deck would need to be changed to another deck entirely. Maybe only one player would know of the existence of jokers. Or perhaps there were secret hands which always win. Aces and Eights and all that.

Then we’re not really playing Texas Hold ‘em. We’re breaking its rules. Let’s look at another game: Archipelago.

In Archipelago, each player has two key pieces of hidden information: their resources and their objective. Each player has an objective, and that objective performs two tasks. It establishes a hidden end condition for the game – such as building five markets – and establishes a scoring mechanic for all players at the end of the game – such as holding iron at the end of the game provides victory points.

The possible objectives are public, so the game cultivates a dynamic whereby players watch each other’s public actions and guess which secret objective the others might be trying to fulfill. Oh, is Bob hoarding iron? Then Bob must have the iron and markets objective!

The core mechanic for teasing that private information out comes from Crises. At the start of each round, there is a crisis, and the players must pool their resources or more of the indigenous people of the archipelago will rise up in rebellion. If rebels outnumber the population controlled by players, then the rebels rise up and kill the players in their sleep. Everybody dies. Everybody loses.

In a crisis, there is a key piece of public information. How many resources each player has is private, the remaining number of resources in the game’s supply – however – is known. Following the examples from earlier, people can be pretty sure that someone has the iron objective when the players fail to fulfill the demands of an iron crisis and yet the iron supply in the box is depleted.

When that happens, everyone knows that someone has the iron objective. That person is holding out, and they are willing to risk the entire game in order to maintain their advantage. They’d rather go extinct than miss out on a chance for profit.


Back in the really real world, Exxon and Imperial’s model was making some truly dire predictions in the far term. Apocalyptic warnings foretold the extinction of the human race. These warnings were as secret as they were remote. Action on emissions could be delayed with minimal cost to the company. Property and lives would be lost, but none of that property would be company property. The lives? Well, those didn’t belong to the company either.

In Archipelago, there is one objective which stands out from the rest. There is a chance that a player is holding an objective which states that they win if the archipelago rebels. Much like millennial sects of Christianity, these players will work to make certain the end times come sooner rather than later.

When a player like our iron-hoarder from earlier is at the table, other players are forced to come to terms with the possibility that the hoarder is not holding the iron objective at all. They may just be hoping to watch the world burn. Sometimes, this is the most fun and exciting moment of the game. The whole world – the game world in this case – is held hostage by a maniac. Or is it? Maybe it’s not a maniac at all. Maybe it’s just a man who would rather see everyone lose than anyone else win.

I’ll give you another tip from a tired gambler. When a player comes to your table and you can’t tell if they are motivated by profit or suicide, thank them and send them away. Maybe not the first time, they could just be trying something new. If it happens again and again over the course of years, just lead them away.

Your life will be better without them in it.

Play and Privilege

My friend and former classmate, Tanya Short, wrote this article about the “Unique Problems of Cooperative Game Design”. It recalled to me my former article about betrayal mechanics in games as well as Wil Wheaton’s series on good gamemastering. And, as we see with Shut Up & Sit Down’s excellent Tips for the Top 5 Problem Players, there is much more to cooperative gaming than making a mechanically sound and fair system.

Today, we’re going to explore the privilege, framing, and inequality.

Read More

Powered by WordPress & Theme by Anders Norén