Thursday, February 05, 2015

Let's compare two theoretical and very simple games:

Game A: Roll a 20-sided dice. You win if you roll higher than a 5, you lose if you roll 5 or lower.

Game B: Roll a 20-sided dice. You win if you roll higher than a 15, you lose if you roll 15 or lower.

Question: Is game B more difficult than game A? Does your answer change if the method of determining a number between 1 and 20 changes from dice to let's say throwing a dart at a dart board?

Please only answer if you understand the question. I don't need a comment to tell me that all versions of games I describe here are extremely stupid, I know that. The question is about comparing a player performance that is a result of luck, skill, or a mix of both with an arbitrary win condition.

Well it depends on the type of game you want to enforce that game design.

Type A promotes determination and longevity, even though it also promotes frustration.

Whereas Type B takes the exact opposite approach.

If you are making an MMORpg longevity and determination is what a developer should look for IMHO whereas if you are making a fun relaxing and short game Type B is well suited.

Of course complexity is the same in both cases.

Difficult: no.
Less liklely to win: yes.

Does it change....: yes, dice rolling is 100% random, throwing darts is not. I get better at throwing darts with time, I don't get better at rolling dice (ok, this second point is debatable, since cheating at dice-rolling is indeed possible). In this case game B is more difficult (with normal assumptions about a dart board).

Games of pure luck has zero skill involved and thus there is no difference in difficulty.

A game of darts depends heavily on hand/eye coordination and you can get better as you play more, thus it is harder for some people than others. Obviously, if you had set a larger scoring area during game A and a smaller one in game B, B would be more difficult.

I'm curious to see where you're getting at...

I'm curious to see where you're getting at...

I was just reading in my news feed, which includes the Steam RSS feed, about some new "rogue-like" indie game. In the announcements of such rogue-like games, "high difficulty" is frequently mentioned as a feature.

In my experience with rogue-like games, there is a significant impact of luck, and a smaller impact of skill / experience with the game. Changing the "difficulty" of such a game usually involves making monsters have higher stats, and slowing down recovery. Result is that it becomes statistically more probable to "lose", while it takes an extreme streak of luck to "win". But isn't that just a very arbitrary win condition which has nothing to do with actual difficulty?

For something to be easy or hard you need to have some control over the outcome. Difficulty is proportional to the skill required to affect a change in the outcome.

When an outcome is entirely luck based difficulty is non-existent

The overall outcome after multiple luck based attempts may involve skill if there is an degree of adjusting strategy to meet the random events to maximise the probability of a favourable outcome at the end of the chain. (IE. Having the best hand in poker is luck but the outcome of who ends the night with the most money features a high component of skill).

For me, yes. Game B is more difficult (and more annoying) than Game A in both dice and dart version as there's a lower range of "win" available.

Using darts will improve your odds over time though as the player will get better at influencing where the dart will go.

The same can be applied with dice but people frown upon that as "cheating". ^_~

Dice games: no, because there is no way to improve in throwing dices. Both a total "n00b" and the best player in the World has 75% chance to win A and 25 to win B.

In the dart case: yes. Less skill is enough to win if the dartboard is bigger.

In a theoretical answer: "difficulty" of the game is defined by the distance of the results of the "noobs" and the "pros". The game is easy if these groups have similar results (snakes and ladders) and difficult if their results are very different (chess).

On a good rogue-like, harder difficulty might only involve enemy mobs having bigger numbers, but you should have a greater ability to modify the outcome by employing tactics and preparation.

Yes game B is harder. I define a game is hard if I have a low chance to win. In most game, winning is a combinaison of luck, time, skill, and sometimes money. For a given game, if my chance to win is 1/4, the game is hard for me, or at least harder than a similar game where I win 3/4 !

The dictionary I have at hand gives three definitions for difficult:
1. hard to do, make, or carry out: arduous
2a. hard to deal with, manage, or overcome
2b. hard to understand: puzzling.
Going by those, the dice game is not difficult in either version under definition 1 because rolling a dice requires only a minimal amount of physical ability. Game B is more difficult under definition 2a because the probability of winning goes down, making the goal harder to "overcome." Neither game is difficult under definition 2b because the rules are quite simple.

Neither game is harder. They have the same difficulty curve, just with different win conditions.

However, roguelikes are different.

What makes roguelikes compelling is the knowledge, on entering the game, that there is no guarantee of a win condition, but the rules are all the same. In XCom, the aliens have a specific ruleset, period. They have a behavior they will follow - they don't cheat to create artificial moments of tension, and they don't suddenly flip over to fighting with lesser AI just because things are going badly for you.

You can lose, but you'll lose fairly within the context of the game.

Yes, random events are part of that - but that's not hidden from you. You're given the knowledge of precisely what is random during your first playthrough. It is that randomness that makes the game unpredictable, that incorporates risk.

Roguelikes are tough to make and make them fair - but the recent XCom, FTL, and even Papers, Please! are great examples of the genre done right. The thing is: when you win, you never feel as though the game /gave/ you the win.

For boardgame equivalents, look at Pandemic, Forbidden Island and Desert, or the new Shadowrun: Crossfire deckbuilder. Same thing applies - you aren't always going to win, and that gives you the tension you need for the games to be compelling. However, you have all the tools - and it's up to you to use the tools well with a reasonable sense that you can win if you don't screw up too badly.

It depends on how you define difficulty. If you define it as success probability for a beginner, they are the same. If you define it as success probability for a skilled player, the second is easier. If you define it as the increase in success probability due to skill, the second is more difficult.

This type of "difficulty" in roguelikes is made trick players into not noticing that the strategy element is actually quite easy. For instance, my first time playing FTL ended in a rather predictable stomping at the hands of the end boss. I went online to look up strategy, and it turns out I had figured out 90% of it on my first play-through. The other 10% were fairly simple, I just hadn't thought of them yet. An hour into the game and I'm a strategy master. All that was left was to spam the game over and over and over until I was lucky enough to win, learning and changing nothing each time.

You could say the same thing about X-COM. How long did it really take you to figure out the strategy surrounding the cover mechanic? Are you really doing anything differently than after your first few hours?

But there is some strategy, combined with the low odds a given player can complete a given attempt. This makes players who do stick it out feel like brilliant, special flowers.

I think the key with roguelikes is that yes, the overall strategy is usually not too hard to figure out (if there are a lot of wildly different strategies, the game is likely to be unbalanced). But it's not trivial to implement, and you have to play the angles tactically to increase your chances of success.

Hmmm roguelikes may have a lot of luck, but my memories of playing nethack are that it's VERY important to learn all the tricks. Of course in the time of the internet this would amount to pure and simple googling, but in the original nethack days it was more like "interpreting" what the weirdest menu commands would do and trying it out. I still have memories about drawing in the ground with a wand or dropping a ring into a sink.....

Yes, the dice game B is more difficult than game A. If you throw a thousand random players at each game, a greater proportion will win game A than game B.

I would measure DIFFICULTY or HARDness of a game as how exclusive the winners are. Fewer winners per player = harder to win = difficult.

When you switch to the dartboard, you change win/loss ratios based on how hard it is to aim for the 'win' areas for the average player ability, so there might be more or fewer winners than the dice game, changing the difficulty.

But imagine it's set up just perfectly so that there are exactly as many winners in the dart games as the dice games. I would say it's the same difficulty level as the dice, but more FAIR. A more skilled player is more likely to be a winner than a less skilled player.

What I think some people mean by a hard or difficult game is what I would call a game that is more CHALLENGING. Imagine if you took your dart game and said you also had to make your throw while jumping up and down on only one foot. But the areas you have to hit to win are also bigger, so overall win rates are the same. The game doesn't get harder to win, it just becomes more challenging because your input tests more/different skills or makes you work harder to achieve that same win rate.

I think B is more difficult than A in both cases. The dart game is fairly clear cut because it requires more skill to hit 1/4 of the area than it requires to land in 3/4. The dice game does not require any skill but does require persistence if you want to win. In game B you will need greater persistence to ensure a win than game A. Therefore I think it is still correct to say that game B is more difficult to win.

This comment has been removed by the author.

Both games are trivial.

B is not harder, only harsher with fewer winners.

Well, most video games have elements of luck and skill.

I play a lot of FIFA, and some games you bang them off the post five or six times, sometimes you score three goals. The interaction of a ton of factors create luck, good and bad, and your skill helps inform whether you take advantage of the good luck and cope with the bad luck. Without luck, there is only one way to play the game correctly. Once you figure it out, it would become very boring.

Same thing in FPS games, or MMOs, or any game.

Dude goes on a crit streak, you get destroyed. Guy is hiding in a bush and gets a lucky headshot, you get killed. One of your crack units in a strategy game routs, you get a bad hand in Magic. Outside of the dumber casino games I can't think of games that don't mix skill and luck in a healthy mix.

To bring this back to MMOs, the main point of the game is stacking the deck, so to speak. A bottom 2nd standard deviation player decked out in the best PVP gear will beat a top 2nd standard deviation player decked out in quest greens, much less someone ten levels below them.

So, in your example, you kind of have two people, one who has to role 5 or lower and one who has gotten enough epics that he only needs 15 or lower. The game design in MMOS is to spend tons of time eliminating the need for skill, and increasing your structural odds of winning (aka luck).

Another reason MMOS kinda blow.