Problem Solving for Hocus's Design

We could discuss every problem in every version of the game, but we think it best if we focus on the problems we identified and how we fixed them, roughly chronologically, for Hocus Poker as it exists today. If you have follow up questions about a specific portion of this, comment below and we’ll be happy to answer. This post grew a tad long (we've done a lot of work!), so look forward to a part 2 (or even 3) in the future.

If you want to try Hocus Poker, you can get the PNP here and read the rules here.

Post by: Joshua Buergel and Grant Rodiek

Grant: The starting point of the game was “Texas Hold ‘Em plus Spells.” In the first iteration, the cost of the spell was to discard a defined quantity or set of cards. For example, “discard a pair from your hand to take this Action.” The problem was, the actions weren't very good and you needed a crazy good draw (basically a pocket pair) to do the Action. And, why would you get rid of a perfectly good hand to draw something random?

Next, I tried reducing the cost. Discard a single card, or a single card of a defined rank (ex: face card), but still it was a problem. I changed it to coins. Spend coins to activate a spell.

The problem then was that if a spell was really good, people would just activate the same one over and over again. Economics acted as the guide. With every use, I made it so the spell cost more. This naturally caused players to use different spells. It added a layer of resource management to the game and opened up for a lot more interesting decisions.

This is one of the game’s oldest and most important mechanics. Now, how we’ve gone about indicating the increase in cost is the topic of an entire post, so we’ll cover that in the future.

Josh: This all happened before I ever saw the game. In the first version that I ever saw, it had the basic structure of paying mana (then called coins) to buy spells, with the increasing cost in place. It’s just a huge mechanic, truly crucial to the game, but I didn’t have to worry about it. What makes it work is that it creates a market at the table. At every action of the players, each of their various actions has a different cost and benefit, and they weigh those and take what they regard as the best value. But then it changes! The next player has to pay more as things are bought, which is more or less micro-economics 101, and it works as well here as it does with idealized widget firms. I've never had a moment of doubt about that mechanic.

Grant: We wanted to eliminate player elimination. This is something I think is not fun about real poker. Really, I don’t think it’s fun in most games with more than 2 people.

Josh: I think there are specific cases where it can work well, but they’re rare, and certainly not appropriate with this style of game.

Grant: My first take was to add a way for players to regain currency. Players could fold in order to gain a payout of Mana. The problem with this was that the rich remained rich and the poor had to not play to get a little bit of money. It’s rare that telling someone to not play will result in a fun experience for them. That’s another problem with typical poker -- you’re supposed to fold most hands. Also, in this setup the chip leader could just squash the poor players.

It was also non-intuitive to receive money for folding. People didn't get it.

Josh: I first encountered the game with this mechanic in place. During my first playtest, there was still betting during the game, and folding (called “Cash Out”) pulled money back to players. It was an attempt to keep a lid on the problem of a big stack pushing the table around, providing a method for players to recover cash without winning a hand. The money in question came from the bank, meaning the economy slowly (or not so slowly!) grew over the course of the game. It diluted the big stacks, yes, but it was kind of confusing and also didn't really dilute the big stacks fast enough. It just kind of didn't work.

Added to that was the fact that the original rules didn't really cover what happened when someone made a bet that was too large for the table to cover. It was effectively limit Hold ‘Em (bets were capped at five Mana), but it was easily possible for that to still be too much. I was advocating for “you can always call” rules, but those are really, really hard to write from scratch (give it a try!). The economy was neither fish nor fowl at this point.

Grant: I did write them. They were bad and I think they took an entire page just to cover the concept. There was a brief phase where we toyed with having betting as a means to bring bluffing back. It was a lazy attempt. But, in doing so, I think it demonstrated to us that we needed to be more or less like poker. We chose less and it was a good, pivotal decision. You can’t keep everything. But, by freeing yourself, you get to do new stuff.

Josh: Stack management is a huge part of no-limit Hold ‘Em. It’s kind of what separates great players from so-so players. Understanding things like pot commitment, pot odds, all that junk - it’s huge, and it’s an important part of skill, and it’s really beyond the pale for a casual game like this. Having any of that stuff be part of the economy in a lightweight card game was a mismatch.

Coming away from that first playtest, this was my biggest concern with the game. The economy just had to change. There were really two games here fighting each other: on one hand, this light-hearted game of spellcasting, hand and card management, which encouraged wishcasting and chasing hands; on the other, a darker, deeper game that reflected its poker roots, of ruthlessly bleeding out your opponents in Hold ‘Em fashion.

Grant: This also brings up the similar symptoms of runaway leaders. It was often clear that after a win or two, the chip leader was going to win. They could even just park it, add no Mana to the economy to be won, and win after a number of hands were played, which was the end condition.

As a result of all of this, we had a breakthrough. What if the win condition wasn't Mana, as it is in poker, but points? Suddenly, having a lot of Mana meant you had a large amount of currency, but it wouldn't mean the win.

Josh: I think both of those are worth unpacking further. The first is points. There were a few ways to fiddle with the economy. One idea would be to reset each hand. Everybody goes back to the same X mana every turn, and then just tries to maximize their result for that hand. That has the advantage of simplicity and draining the big stacks of their power, but it has a downside: folding is meaningless.

Not only would folding be stupid in that setup, but paying big for effects is meaningless. The cost of things isn't very important, since you would just stay in every hand. It drains too much tension from the game. How do you fix that? Well, we need to have the pot be meaningful each hand, since we need people to weigh how much they’re paying in. But, we don’t want individual people to be able to bully the table just because they won previous hands. The answer suggests itself at that point: you need to record the results of each hand somehow while returning everybody to relatively even footing. That’s where the idea to record points came from.

Grant: According to you searching our email records, we tried four solutions, the last being the right one.

  1. If you win the hand, you earn 1 Point and reset Mana to all players. This has problems, as noted above.
  2. If you win the hand, you buy Points via a complicated rule. More on that in a second.
  3. If you win the hand, you buy Points at a fixed ratio by paying other players. Close!
  4. If you win the hand, you buy Points at the cost of 1 Mana paid to each player. Ding ding! We have a simple, scalable winner!

This solution fixed the economy by balancing it long term, giving weight to your mana expenditures (spending more = more points for the victor), removed the chip leader bullying issue, removed the runaway leader issue.

Josh: Can I just say, I still like the intellectual properties of solution #2?

Grant: You can say whatever you want. Free country!

Josh: The idea, basically, was that the chip leader could buy points with the cost determined by the size of the second biggest stack at the table. You thus could reward somebody who gets far out ahead of the table, but if you had a pursuer, your lead was less valuable. I think the dynamics of that would have been fun to watch. Clearly too complicated for a game like this one, and it was never actually tested, but I think it’s a cool, fun idea that I might re-use in another game.

Note that here is an area where having two designers paid off. Going from a Hold ‘Em  economy to our solution, which stands to this day, required only two games of Hocus Poker. Most of the intermediate steps were hashed out in discussion, saving us from burning out our playtesters too much.

Grant: Our solution did, however, introduce a new issue: hoarding. You could only buy Points if you won. However, a player could decline to purchase points and therefore hold all the Mana. Without Mana, other Wizards couldn't use Spells. This basically meant they couldn't win and it made the game really not fun.

To make a simple comparison, it’d be like a worker placement game where one player had the option to hold onto your workers.

My first solution was to ignore it. In the, put my hands over my ears and go “na na na na na” method.

Josh: I can remember you trying to talk yourself into this not being a problem. And it wasn't a huge one, honestly. Most players would just buy points as they had a chance, no problem. But some players would think one step deeper and decide that there was no point in buying points if they couldn't buy a winning set, and they’d sit on their Mana.

Bullying was back in the game, in a much more minor form. We needed a way to force Mana redistribution. It seemed obvious: just make them buy points. Heavy handed, but sometimes problems have easy solutions.

Grant: I tried to convince myself that I shouldn't have to design mechanics for awful people. And I still somewhat stand by that. However, awful people do exist. We debated whether to make it a minimum number of points, but that led to weird side issues, especially when players had to split the pot (which we later removed). Ultimately, the clean, global issue was to force players to spend Mana until they were equal to or less than the starting amount.

Josh: The line between “awful person” and “ruthless optimizer” is pretty thin. Speaking as someone that sometimes falls in the latter camp (I’ll leave the first for others to define), it was an issue. Maybe in 80% of games, it wasn't a problem. But 20% is still an issue.

Grant: It’s a good lesson, really. Just because I personally don’t think like that doesn't mean others don’t. Your design shouldn't have huge opportunities just screaming to be, exploited.

Josh: It can be easy to say that a particular audience won’t have an issue. But why not try and solve it for as many audiences as possible. I have a bunch of math majors in my regular gaming group. The game should work for them as well.

Grant: We then needed to solve our endgame, which was incredibly anti-climatic. When you played a set number of Hands, players in the lead would have an incentive to play cautiously and withhold potential winnings. For example, if someone had the most points and knew the game would end in 3 hands, they would slow play those 3 hands to choke the economy. If a player doesn't win any Mana, they can’t buy points.

Josh: In my first several games, there were players who were eliminated before the end of the game. There was literally no way that they could possibly buy enough points in the final hand to win the game, even if they won every mana chip at the table and bought points with them. That’s a bad thing.

Grant: People in that position leaving our game feeling awful. They might as well have spent 30 minutes being insulted.

We tried a race solution: first to 8 points wins. This felt lame, especially as in some rare cases, the game could end really quickly. We then tried first to 8 points, plus one final hand. In one out of about nine tests this final round mattered. In the remaining eight, it was a waste of time. A foregone conclusion.

Fundamentally, we had an issue where players in the lead remained in the lead. Players in the back felt terrible, especially if they lost without any points.

Josh: Not only that, but there was even a kingmaker problem. In one playtest game I played at this time, a three-player game, I was pretty far in the lead going into the final round. My friend Jarrett was toast, with no hope of winning, and Marc had at least an outside shot if the pot got rich enough. I folded right away, so that Marc’s pot was going to be limited. The only way that Marc could win would be if Jarrett stayed in and paid liberally into the pot. But Jarrett had no way to win. It was a classic kingmaker situation: an eliminated player could help determine who wins with no chance themselves.

Both problems stemmed from the end condition. It comes down to race mechanics. You can only gain X number of points in any given hand, defined by how much Mana is available at the table. X tends to increase as hands go on, but it’s still a limit. Fall more than X behind, and you need more than one hand to catch up. Adding a final hand to the list just moves that threshold to 2X for futility. Yes, it’s better, but it’s still possible to be eliminated before the end of the game. This isn't a wargame, that’s really not OK. You can keep layering on extra rounds, but that does two things: one, it dilutes the impact of the “finish line”, which is crappy and makes the first part of the game feel futile; two, it doesn't actually solve the problem, just kind of buries it under the rug.

So, time to bust out the thinking hats. What properties do we want for the game? We returned to first principles: one of the guiding principles of the game was to avoid player elimination. At every moment, people need to have a chance. How do you capture that? Well, you can solve that if, at any given time, a player could go on a run and still win.

That’s when things came together: we don’t want a fixed number of rounds, due to kingmaker/elimination issues. We don’t want a fixed number of winning hands, because we want the size of each pot to be significant. A first-past-the-post victory condition still leaves people possibly eliminated. So we combined a few options. A points threshold added to a requirement to win one more hand. That gives a tangible goal for people to hit with the addition of someone needing one final win. That keeps everybody in it towards the end. If you can keep winning hands, you can still pull things out.

Grant: I love when a game ends with everyone in sudden death. It’s delightful knowing this is the last hand. Whoever wins takes it all.

Josh: I've come back from zero points to win the game by just running the table. It’s a great, great feeling. This endgame mechanic is similar to one I used before, in Foresight, and the hybrid of conditions works really well.

Grant: Oh man, we should talk about Antes! I forgot about them. For a phase there, they were so important. We had a problem where some spells were strictly better. You would always use them first. Keep in mind, at this point, all spells had a base activation of 1. Summon, which was then “Draw 1” was simply too good. You would always do it.

Josh: This was a problem I noted in the first playtest. There was just no reason not to use Summon right away. Sometimes several actions in a row. It was just too obvious.

Grant: As a side note, fixing Summon is how we came about the Show mechanic. Summon became Draw 1, Show 1, which meant you had to play a card from your hand. This revealed information, which is important, potentially exposed the card to some spells, but meant you couldn’t Show that card again. Show has since become one of our most important mechanics.

When you get to a point like this, you have two paths you can follow.

  1. Perfect Balance: Make every action balanced, which means it’s fine for them to share a cost. I think this path is very boring.
  2. Economic Balance: Very the cost of every action. This means you can have wildly disproportionate powers which are reigned in with an economic tool. I vastly prefer this and love finding solutions for these in my games.

We chose the economic solution -- the Ante. Better spells would have an ante. You would add 1 Mana from the Bank onto the spell, which would increase its starting cost by 1. This would also inflate the economy, which at the time seemed like a good thing. Smaller pots in the beginning, larger pots towards the end.

Inflation had a few problems. Plus, it was a clunky way to include the idea that “some spells should cost more from the start.” There was a simpler solution.

Josh: An ante was suggested after my first playtest to just make the differing spell power levels work. I think it’s more interesting to have variety in spells without having to keep everything leveled out. And the Ante mechanic did work pretty well - we have more or less that same mechanic in the game today. But there was a hidden problem: it pumped mana into the game. Each hand then became more significant as the game went on. In addition, the costs of spells became less significant as the game went on. Both effects weren’t huge, but they were there.

There was also a more subtle problem, which was the cost of production. Having the economy slowly increase meant that the game had to include enough mana to get the players through the game. With 5 players, you needed the start mana, plus roughly 13 hands of inflation (26 spells, two per hand). I sat down to compute out the worst case, and it worked out to something like 84 mana in the absolute worst case. Not insurmountable, but that was a fair bit to produce.

So, we again had an issue. Honestly, we lived with this one for a fair bit. It wasn’t a huge problem, it was a subtle one. The game basically worked under these conditions. But it nagged at us.

I think we both arrived at the same solution. There was no need to pull mana from the bank. It could just be the starting cost. The rest of the economy worked fine, you now had a fixed pool, off you go. It meant that each hand would be of roughly the same importance, encouraging players to go hard right out of the gate. Production was simpler (we got it down to a single sheet of tokens). There were essentially no down sides to it.

Thanks for reading. Hopefully some of these insights and development paths are useful to you in your own design. Potentially, they may aid in your enjoyment of Hocus Poker. If you want to try Hocus Poker, you can get the PNP here and read the rules here