Addressing Richard Carrillo's RPS Question

At GDC 2017, Richard Carrillo of Ubisoft Toronto gave a talk on the game design interviewing process. It's a good video, and I'll link it here.

The first 'good' interview question Carrillo poses is the following:

What goes through your mind when asked to remove one of the three options in Rock Paper Scissors, giving the player only two to choose from?

I wrote this post as a draft one month ago, and I don't remember the follow-up he gave for this question, so I'm going to answer it in this post as if I was just asked it off-the-cuff during an interview having never seen it before.

My first thought is, does this break the game?

Fundamentally, yes. It not only changes any given play's odds of winning, it makes it so that you can never lose if you play the advantaged option. Rock Paper Scissors has always been a game of chance with a little smattering of psychological analysis/warfare in the background, and while it would stay that way, there's functionally no way to 'outplay' someone who plays the advantaged option. Here's an example of the original game of Rock Paper Scissors, assuming you're a child on a playground and it's best-of-five unless you insist on "just one more":

Original RPS Variant:

1. You play rock. They play paper. You lose.

2. You play rock thinking they'll switch it up and you either tie or win. They play paper. You lose.

3. You play rock a third time, thinking surely this time will be different. They play paper. You lose, and they laugh this time. You'll play another round.

4. You play scissors, switch it up a little, catch them off guard. They play scissors. You both stare at each other in a tie.

5. You play paper, unsure of what'll happen. They play rock. You finally win but they remind you it was best-of-five and they already have three wins. You admit defeat, defeatedly.

What if you removed paper, your adversary's favorite move?

The new RS Variant:

1. You don't have many options this time around, but you stick with rock. They play rock. You tie.

2. Now it's a game of chicken. Who decides that rock is boring, and plays scissors? You play rock. They play rock. You tie.

3. You're frustrated now since there's nothing you can do. You play rock. They play rock. Tie.

4. Rock looks a bit like a fist. No, it is a fist. You play fist. They play fist. Now you're fighting. No one wins.

If you play the game out in your head after removing an option, it should be obvious what the consequences of said removal would be. But how can you apply this to complex games? I'll look at Destiny 2 PvP for this, and explain why a wide-ranging change to damage that lowers time-to-kill may seem simple in concept but fundamentally ruins sandbox balance and breaks the game.

Take the most popular primary weapon type in the game: Handcannons. Now take one that dips in and out of popularity: Auto Rifles. If you increase the damage on both, what happens? Increase it by a small amount and Auto Rifles may kill significantly faster, or more consistently, while Handcannons will only see a slight bump to consistency (range falloff or 1 crit 2 body with 140 RPMs). Increase it by a large amount and suddenly Handcannons have their consistency dramatically improved and optimal time-to-kill halved (2 crit). You might think, Auto Rifles would also kill faster and have more consistency with a large buff, right? And you'd be correct. However! The nature of these two guns is different. Auto Rifles require sustained fire, and Handcannons can do burst damage in-between peeking out from behind cover. In a 1v1 between two players of similar skill, Handcannons would almost always win given playing around cover. Destiny 2 maps have a lot of cover, and Handcannons have a longer effective range as well. They would become extremely oppressive if they suddenly killed in two shots. You can see this in some ways when you play Momentum Control — high impact Scout Rifles like Jade Rabbit dominate, as they kill in two shots from across the map.

A more direct comparison: what if you removed a class from World of Warcraft? Sure, unlike RPS there are other classes that can fulfill the same roles. Remove Priest and you'll still have caster damage classes like Elementalist Shaman, or healer classes like Restoration Druid. But you take away an option that plenty of your playerbase enjoyed, and you remove the specific aspects that it brought to the game. Buffs, debuffs, utility. All of those unique things that Priest had are now gone, and while there may be things that provide a similar function, it isn't the same.

The best, and most niche example I have of this that actually happened was in Final Fantasy XIV: Endwalker. I was a Samurai main, and a major part of my rotation included an ability called "Hissatsu: Kaiten". This is what it did:

In Endwalker it was removed from the game. To balance the loss of damage in Samurai's rotation, the team buffed Samurai's other abilities and standardized some crit procs to make damage more consistent across the board while reducing "button bloat". The end result was that I no longer got to hit a cool big damage button that did a unique katana-sheath animation before the flashy attack that is Midare Setsugekka, and my rotation suddenly felt much duller even if I was doing more damage and doing it more consistently with less effort.

Rest in peace Kaiten, you will be missed.

This is all to say that in any game, no matter how simple, and no matter how simple the change: it will affect players, and it will have consequences. You may think a change fixes an issue, and it might — but chances are it'll introduce a new issue, and that issue might not be as easy a fix as the change you initially implemented was. Everything needs a wholistic analysis as to how it fits into the greater game, and how it will affect a player's minute-to-minute gameplay. Sometimes people will be unhappy, and obviously you can't please everyone. But decisions should be made such that they minimize short-term blowback where possible and maximize long-term health of the game and player base.

And, of course, the end goal should always be fun. That's what games are for, after all.