Luck Vs Skill: The False Dichotomy (Repost)

Hey everybody – This week I had to grade research papers, so I did not have time to write a new post. The article that was scheduled for this week (about personality in games) will hopefully be up next Saturday. Until then, enjoy this post about the relationship between Luck and Skill in game design, which was originally posted on February 10th, 2018. Also, in a few weeks I am planning the next in my “Game Designer Spotlight” series, so if you have any designers that you think deserve more focus, let me know in the comments or on Twitter. See you next week!

---

When I started researching for this article, my original plan was to talk mainly about a category of games known as “Solved Games”, or games that have a mathematically proven optimal strategy. I was going to use that article to talk about what it means to have solved a game, different games that have been solved, things like that.

During my research, however, I uncovered what I believe to be a much more compelling issue. Specifically, I uncovered a debate about the role of randomness in games, and whether it makes them more or less skill based. This is a very heated argument among certain parts of the gaming community, so I figured I would use this article to give some of my thoughts on the topic.

“Ice Cream or Puppies?”

As far as I can tell, the main arguments on both sides break down like this. On the anti-randomness side, there are those that claim that randomness by it’s very nature reduces the amount of control that players have over the outcome of the game. They argue that randomness can lead to an inferior player defeating a superior player simply by luck, and therefore competitive games should try to minimize the effects of randomness as much as possible.

On the other hand, there are those that argue that randomness actually increases the amount of skill required in a game. Proponents of this viewpoint argue that in a game with no random elements there will always develop a supreme strategy which guarantees one player or the other a win, and random elements avoid this by forcing players to adapt and think on their feet. This promotes skill and strategy, as opposed to simple memorization.

Personally, I think that both of these arguments have a certain amount of merit. On the one hand, it’s easy to become frustrated when playing a game that you think you can win, only to be foiled by some sort of random element right at the end. An example of this is in the Pokemon video games – you may think that you have the game in the bag, but then your opponent ends up landing a critical hit and you lose. When situations like this occur, players can instinctively react by blaming the random elements of the game, and claim that the would have won “if it was only based on skill”.

The other side makes some great points as well. Games that are purely skill based often do end up becoming very rote, and memorization can become a huge part of it. Probably

the best example of this is chess. When first learning chess, players spend a lot of time learning about basic strategy – how the pieces move, pawn skeletons, promotion, things like that. After a while, however, players will end up spending much of their time memorizing opening moves or studying end-game positioning.

Unfortunately, neither side tells the whole story. I think that the biggest misconception here is that people use the word “randomness” to refer to a number of different things. In some cases, random elements in games do reduce the amount of skill required, mainly for the sake of creating exciting moments of variety. On the other hand, depending on how the randomness is implemented, it can test the player’s skill in a way that no perfect-information game ever could.

Some Random Thoughts

There are many different ways to implement randomness in games, and not all of them are equal. When implementing random features there are a number of different decisions to be made such as: when in the game does the randomness occur? Does the player have a change to influence or respond to the randomness? Does the random effect reward player skill, or punish it? Based on the answers to these questions, the random element could enhance the game or ruin it.

• When Does the Randomness Occur

The timing of random events in a game can have a huge effect on how this randomness is viewed. Generally, the earlier in the game that a random element occurs the more likely players are to accept it. To show this principle, let’s take a look at two possible variations of chess.

The first variation is what is known as Chess960. This variation was invented by Chess World Champion Bobby Fischer, and involves adjusting the starting position of the game. Each player still has a row of pawns, and the same number of pieces, but the order of pieces in their first row is changed. Aside from a few guidelines such as “players must have bishops on opposition colors” and “the king should be able to castle on either side”, the order of all major pieces is basically random.

The other variation is known as “Rando-Chess”, and was proposed by game designer Richard Garfield. In this version of chess, both players play the game normally until the very end. Then, just as one player is about to checkmate the other, a die is rolled. If the die rolls a 2-6, the checkmate occurs normally. If, however, the die rolls a 1 then the player who was about to checkmate loses instead.

Which of these variations has more randomness? Honestly, it’s hard to say. In the first, the entire setup of the game is different, while in the other there is only a single roll of the dice. Does rando-chess reduce the better player’s chances of winning more than Chess960? It’s hard to say. A significantly better player has about a 5/6th chance of winning the game, but it’s possible that Chess960 has similar or even worse odds. There are so many possible variations, and it’s entirely possible that a number of these possible setups give the opponent a significant advantage which would allow them to defeat a more skilled opponent.

So why does one of these situations seem much more “fair” than the other? There are a couple of reasons, but one of the big ones is when the randomness happens. In Chess960, you may be presented with a board state that puts you at a pretty severe disadvantage, and yet you feel like you have an entire game to overcome this. On the other hand, Rando-chess waits until the entire game is over before implementing any randomness. In most cases this will not affect the outcome, but when it does it can ruin the entire experience.

• Can You Respond to Random Events?

Another thing that separates different types of randomness is the player’s ability to respond to it. While this principle applies to the chess variants above, I think an even better example is the game of Poker (specifically no-limit Texas Hold’em).  There is a common misconception that Texas Hold’em is nothing more than a game of pure chance – another way for casinos to separate unwary people from their money. However, because of the way randomness is implemented in this game this couldn’t be further from the truth.

I’m not going to argue that Poker doesn’t have random elements in it – it clearly does, and they make up a huge portion of the appeal of the game. However, this doesn’t mean that it is not a skill based game. Although players cannot control what hand they are dealt, they do have the power to choose how they respond to it. It is these responses that make up the bulk of the strategy of the game.

It’s true that, when playing Poker the better player is not going to win every hand. In fact, new players tend to play significantly more hands each game than more experienced players, and they are going to win a good number of those hands. However, in the long term the more experienced player will always win out.

The balance is easy to tip, however. Suppose that instead of giving players four different opportunities to adjust their bets each round, they were only given one. Players see their hands, make their bets, and then find out who won. This would be a much less skill based game, because players don’t have the opportunity to respond to events as they are happening.

• Does the Randomness add Skill to the Game?

In any competitive game there are going to be players who believe that personal skill in the game should be the only factor that determines the outcome of the game. I tend to agree – in any real competition, especially one with real stakes such as cash or prizes for the winner, the best player should come out on top. Some players take this argument a step further, and claim that this means randomness should be minimized from competition or removed entirely. Here, I tend to disagree.

Instead of saying that competitive games should have no randomness in them, I believe that randomness in competitive games should give players the opportunity to respond to it and show their skill. Take a game such as Magic: The Gathering. While Magic is not a perfect game, it is definitely a game of skill. That being said, it has a massive amount of randomness in it which mainly stems from the random shuffling of the deck at the beginning of each round.

While the randomness in Magic can occasionally result in some incredibly un-fun situations (*cough* mana screw *cough*), I think that the game would be much less fun if these random elements were removed. Shuffling the deck at the beginning of each game forces players to adapt their strategy every single game. If players were instead simply able to arrange their deck however they pleased, the game would not only be much less fun to play but also far less skillful.

This idea has to do with what Kieth Burgun refers to as “input and output randomness”. Basically, input randomness is the random elements of the game that the player must take into account when making his decision. This randomness can include things like the cards in your hand or previous actions taken by your opponent, which you use to inform your next move. Output randomness, on the other hand, is random events that respond from your actions.

In general, input randomness tends to enhance player skill in a game. Even though the player doesn’t have complete control over these elements, they have enough information to adapt to them and plan around them. Output randomness, on the other hand, tends to decrease the amount of skill in the game. After all, how can you make an optimal decision if you don’t even know what they outcome of your action will be? If a player has information about the probabilities of the random output they can still make relatively informed decisions, but even so this type of randomness tends to be looked down on by players.

In Conclusion

At the end of the day, randomness in games is a much more complicated issue than it may at first appear. It is entirely possible for random elements to detract from an otherwise great game, and make it uncompetitive. On the other hand, when used correctly random elements can actually make a game more skillful. It is not a question of whether competitive games should have randomness, but rather a question of how these elements should get integrated into the game. It is not a choice between luck and skill – every great game needs some measure of both.

Until Next Week

That’s all I have for this week! I’m sorry for switching the topic like this, and if you are still interested in reading an article about solved games, please let me know! If you liked this article, and would like to see more like it in the future, be sure to subscribe to the blog on Facebook, Twitter, or here on WordPress so you will always know when I post a new article.  If you didn’t like this article, let me know what I could be doing better in the comments below! And join me next time, where I will be talking all about Stones, Parchment and Shears!