You might not think that from my Euro leanings, but I love randomness in games. Or, well, I love well-used randomness. I love dice drafting, I love card draw, I love modular setup, I love randomly seeded markets. Randomness can add a dynamic element to games which would otherwise get repetitive. It makes it so, even in a game I’ve played dozens of times, I still get surprised.

However, some games go too far with the randomness. Like an oversalted dish, the randomness takes over and breaks the nuances and subtleties of gameplay. Mostly, it robs players of their agency: where in a dice drafting game the dice give you a new puzzle to adapt to and enrich the decision making, in a “roll-to-succeed” game, it often supersedes your decision making entirely. In some games, magically trading in your ability to choose your actions to gain the ability to choose die results would increase your odds of winning. That’s pretty low agency, if you ask me.

Why follow that paragraph with a picture of Risk? Oh, no reason.

“Yeah Amy, isn’t that just input vs output randomness? It’s 2023, this is nothing new!”

Except no, it’s not just input vs output. King of Tokyo has an incredibly random output randomness: you don’t choose what you do, you roll and do what the dice command! That’s all of this roll-and-move stuff people complain about! However, you also get the ability to reroll any amount of dice twice. It is still random, but because you choose which dice to reroll, you still feel like you have agency. You decide whether or not to you reroll attacks because you don’t want to end in Tokyo; you choose whether or not to reroll Hearts because you feel threatened; you choose whether or not to try to get that third 3 for a risky but rewarding strategy. While the final result is still random, you have taken decisions throughout.

“Oh, so it’s about mitigating randomness?”

Yes, but not only that. Mitigating randomness is an important dial to consider when you include randomness in a game, but too much mitigation takes away the whole emotional impact, the whole tension of the randomness: if you have infinite rerolls, there’s no point in rolling the dice, you should just choose your results.

I think it is less about mitigating than about how you present the random result, and how you allow the luck to be mitigated.

Recently I had a discussion with a designer friend about a Push-Your-Luck mechanism in a card game she was working on. Players had 3 cards in hand, and every hand, following the players’ actions, 3 cards were added to the middle of the table. Amongst other things, each card is worth between 0 and 2 stars, and rather than straight up points, she was working on a push-your-luck mechanism for those stars.

The original mechanism was that if there were an even number of stars in the middle, each star in hand gave you a point; if there was an odd amount, stars in hand were negative points. It’s a 50-50 odd (with a bit of player control over which card went to the middle), and each player can decide how many stars to hold on to, whether they go big and risky, or small and safe. 

During the test, I didn’t like that mechanism. It’s technically a push-your-luck system with agency: I can decide whether to take a big swing by keeping high cards, or keep it safe, and I could even have a bit of impact on the final result. When mixed with the rest of the game system, it made for some really tough decisions. Yet, if I kept a big hand and busted, it was frustrating; if I scored it, it felt unearned. Truly, it was the worst of both worlds.

Then, they tried a Blackjack-like system: as long as you don’t have more stars than the center row, they’re worth points. However, if you do have more, you lose that many points. Suddenly, every 2 in the middle is permission to keep a 2. If you start with three high cards, you can try to replace one of them and add it to the center. If you start with only low cards, you can use them to make your opponents bust. Whatever happened was a strong moment: busting on a big hand was dramatic instead of just being hit in the face; busting with a small hand at least felt good because others lost more; and winning the points by staying close to the margin felt amazing. All in all, the reveal of the last card became a nailbiter.

Those emotional responses are why we mitigate randomness in game. Not because mitigating randomness makes the best people more likely to win, but because having a say over the random result gives you agency, it makes you feel involved, and when you’re involved… well that’s when a simple die roll can get you to jump out of your seat.