APR

16

edmund

Why am I so... hard?

Posted by Edmund on April 16th, 2010 in Super Meat Boy!

So I'm going to attempt to write about difficulty in game design then talk a bit about the Super Meat Boy design process, namely when it comes to how we approached dealing with difficulty. This post probably isn't for everyone but we get a lot of questions regarding this stuff so here goes.... (don't worry there are pictures!)

----------

Difficulty in a platformer is usually established by this very simple formula.

(% chance the player will die) X (Penalty for dying) = Difficulty?

Pretty basic stuff, the higher the chance the player will die and the bigger price they pay for dying the harder the game will appear to the player. This is a formula that's been around from the start, but the one thing that's changed drastically over the years is the "Penalty" aspect.

Penalty for dying in video games started in the arcades where the major penalty was adding a quarter.

Image

This worked very well when it came to getting money from kids, but once home consoles became the norm the player no longer had the ability to add credits with coins and the formula had to change. The goal of high score had been replaced with progression to completion and the major penalty became going back to start.

Image

This is also when "risk/reward" was heavily established. Risk reward was a way for the designer to give the player a way to gain more credits by taking bigger risks, in mario this risk was coin collecting and exploration to find 1ups.

The Mario formula was solid, but as video games tapped into a more mainstream market, penalty for losing had to become less frustrating and penalty = frustration. Companies wanted more people to be able to complete their games, and by the early 90s most platformers added the "continue" option.


Image

As time has passed, lives systems and penalty have almost vanished from most games due to the amount of frustration they caused and difficulty had become watered down to the point of it not really being a factor anymore.

By the early-mid '00s the independent video game scene started to use a more direct and simple formula.

Image

Removing lives all together let the designer base difficulty more on the actual level design and challenge and less around the penalty of losing lives and restarting, in doing so the formula for difficulty changed. The player no longer had to worry about dying, penalty for death basically turned into the amount of time you took to restart after death and the length of the current level.

----------

So how could we take this existing formula and refine it and apply it to Super Meat Boy?

How could we make a seemingly aggravatingly difficult game into something fun that the player could get lost in?

When starting the development of Super Meat Boy these were the big questions that needed answers right away, and this is what we came up with.


1. Keep the levels small

Image

First off it was very important that the levels in Super Meat Boy be bite sized, you could almost think of most of them as micro levels, thrown at the player in rapid succession much like the micro games in the Wario Ware series. If we keep the levels small enough for the player to see their goal, it lowers the stress of not knowing what's to come and the distance they will have to start over from if they die.

2. Keep the action constant

Image

It was imperative that the action never stopped, even when the player was killed. The time it takes for Meat Boy to die and respawn is almost instant, the player never waits to get back into the game, the pace never dropps and the player doesn't even have time to think about dying before they are right back where the left off. This same idea was applied to the level progression, the player never leaves the action till they want to, the levels keep coming as fast as the player can beat them, and all the complete screens, transitions and cut scenes are sped up to keep the fast pace of the game flowing.

3. Reward

Image

The player should always feel good about completing something hard, so what better a reward then a reminder of just how hard that level was? Early in development Tommy implemented the replay system, a replay mode that would start when a level was completed showing the player's past 40+ attempts all playing at once. This simple visual reward for taking a beating not only reminds the player of just how hard they tried but also shows a time line of how they learned and got better as they played.

So we had our basic outline, a hardcore platformer geared towards our horribly spoiled ADD generation, but how could we stay true to the extremely high difficulty par set by IWBTG, Jumper and N+ yet still be accessible enough for someone who was new to the genre to pick up and enjoy?

Was there a way to make something accessible and still hardcore?

This is where the Dark World system comes into play. The dark world is an expert mode set parallel to the main game. As the player completes levels they will unlock expert versions in the dark world if they complete the level under a set par.

Image

In a lot of ways dark worlds difficulty starts where the main games difficulty leaves off. You could say that level 1-5 in the dark world is almost as difficult as level 3-5 in main game.

Image

The goal here was to setup a system that allowed expert players to start the game experiencing high difficulty right off the bat, yet not require those levels to be beaten to complete the main game.

All in all the dark world system allows for Super Meat Boy to become 2 full games, 150+ main game levels for the average gamer and 150+ expert levels for the hardcore gamer, but set up in a way that an average gamer who completes the main game, can easily transition into the difficulty of the dark world levels, and the game will unfold even more.

----------

The point i'm trying to make here is, video games are an exercise in learning and growing. The designer acts as the teacher, giving the player problems that escalate in difficulty hoping their "course" will help them learn as they go, get better and feel good about what they achieve.

When you are trying to teach someone something, you don't punish them when they make a mistake, you let them learn from it and give them positive reinforcement when they do well.

Image

----------

Tune in next time when i'll be talking in depth about RISK REWARD!!! (Bandages, Warp Zones and Playable Characters)
 
COMMENTS

Skab says:

April 16th, 2010 2:04:00 am

very interesting analysis on the evolution of videogames's difficulty through ages !
The "no time for tears" paragraph is so true; i wish more "mainstream" games developpers had this concept in mind when creating their games.
Couldn't expect less from team meat :D !

Skab,
meat lover

Aceblues says:

April 16th, 2010 4:34:11 am

Great article. Totally agree with Skab about "no time for tears" in mainstream games. Really frustrating.

Max Vanlancker says:

April 16th, 2010 7:32:53 am

Beating children is fun, don't knock it till you try it. Also, good article.

Adrian says:

April 16th, 2010 7:42:11 am

I love this analysis! toying with game design myself, this was pretty informative but I felt like I should have already known some of it. I hadn't really thought about some of these ideas, so thanks for the words and I can't wait for the game! good work!

A Friend says:

April 16th, 2010 7:48:34 am

"The time it takes for Meat Boy to die and respawn is almost instant, the player never waits to get back into the game, the pace never dropps"

Should be drops?

Supermeatfanboy says:

April 16th, 2010 7:50:25 am

Very well said!! :)

roberto says:

April 16th, 2010 10:18:07 am

Yeah, I have seen that game design before and it doesn't appeal to me as I reached a treshold where it is too much effort and I stop playing because it is fun anymore...

roberto says:

April 16th, 2010 10:18:39 am

Yeah, I have seen that game design before and it doesn't appeal to me as I reached a treshold where it is too much effort and I stop playing because it is fun anymore...

Goldom says:

April 16th, 2010 12:24:58 pm

I wish all games had difficulty options like this, so people could just play what is fun rather than getting all preeny about playing OMG HARD games. More importantly, it lets people actually learn how to improve slowly, rather than just frustrating them by killing them over and over.

Dominic White says:

April 16th, 2010 12:38:39 pm

The idea that Hard mode is designed for players who have just beaten the normal difficulty is one I'd love to see more games adopt. The last major release I saw that used such a system was Bayonetta on the 360/PS3. Not surprising that it has a perfectly tuned difficulty curve (it's by the makers of God Hand and the original Devil May Cry), but it really does do it right in every way possible.

Once you beat Normal mode in Bayonetta, you unlock Hard, which is a lot like your Dark World. You get to take all your upgrades and weapons right back to the beginning of the game, but you'll be facing endgame enemies in increasingly tough fights right off the bat. It picks up exactly where Normal leaves off, effectively doubling the amount of content to play through.

Scott Rubin says:

April 16th, 2010 2:01:13 pm

I love indie games, nothing against them or those who make them, but infinite lives level restarting was not invented by indie developers in the early to mid '00s. Proper credit for that development goes to, of course, Nintendo with Warioland 2 for the Game Boy, one of the greatest platformers of all time.

In Wario 2, you didn't just have infinite lives, you were invincible. The levels were designed in such a clever way that failure was punished with backtracking and lost treasure. If you fell in a hole, you would always land somewhere. You just might have to walk back a room, climb back up a ladder, and try again.

Heed some advice I have been following lately.

Steal from the forgotten masters.

Pomfrod says:

April 16th, 2010 2:46:38 pm

Except Wario 2 was terrible exactly because it forced you to waste time trudging back to the beginning of a challenge rather than killing you and respawning you. I admire their emphasis on stat conditions rather than a health bar, but the way it was implemented made the game, and 3 as well, a chore to play.

Dominic White says:

April 16th, 2010 2:55:20 pm

Oh, aformentioned Bayonetta also goes the 'small snippets of gameplay, infinite lives' route, but dying takes a large chunk out of your end-of-level score. You're not punished if you're struggling, but it does let you know in no uncertain terms that you could be doing a lot better.

alastair john jack says:

April 16th, 2010 4:35:34 pm

A lot of the NES action games had infinite continues (unfortunately). But for longer games which have a lot of filler like Mario, continues/saves would be a good idea.

Ciro Continisio says:

April 16th, 2010 5:16:57 pm

Very cool, love the pictures.
I need to find a minute to read the text too.

Rorkimaru says:

April 16th, 2010 5:30:18 pm

While I agree that the formula you have developed is great I must admit that I don't see it as a one size fits all kind of solution. I'm currently in early production of a survival horror game in the vein of the old Silent Hill and Resident Evil games. I will be using their methods of creating tension by the risk reward for a higher score for fewer saves but you are risking losing progress.

I know a lot of people find that having to repeat sections is too harsh a punishment but I feel that it is necessary for a Survival Horror game as without it there is no fear of death.

Not only that but the "You Have Died" screen bringing you to the title screen route after death will be taken. Again many people think gaming has gone past this but I disagree. Death at any time is a valid end to a story and if the wanabe hero dies before escaping then that is how the story ends. You can load up and try to get a different ending but I like that every death be treated like a true conclusion. In games where survival is key death has to be a big deal.

(Of course I'm yet to have a big game so I don't know a lot about commercial appeal, I'm just making the games I want to play)

Provider says:

April 16th, 2010 6:21:37 pm

You CAN continue in Super Mario Bros. (Hold A and press Start). You just have to start from World X-1.

MogKnight says:

April 17th, 2010 4:00:42 am

Dying in IWTBG was pretty harsh in that each little section of each stage is a little stage of its own. It was very nice in that you couldn't just endlessly jump through it without somehow getting lucky along all that as it was a chain of events that you had to perfect to advance.

That and IWTBG is the cause for many thrown PC controllers. :(

Barnaby says:

April 17th, 2010 1:29:09 pm

Excellent advice for devs about pitching difficulty by using a 'dark mode' and a clear summary of the history of gaming difficulty evolution, most importantly for the 'ADD' generation of today. The replay mode is also an excellent concept, I wonder how hard it is to implement (presumably by storing player input versus time somehow?) Awesome article, I really appreciated the clear, humorous, and concise explanation of these imporatant game design issues!! Long live Super Meat Boy :)

Barnaby says:

April 17th, 2010 1:29:50 pm

Excellent advice for devs about pitching difficulty by using a 'dark mode' and a clear summary of the history of gaming difficulty evolution, most importantly for the 'ADD' generation of today. The replay mode is also an excellent concept, I wonder how hard it is to implement (presumably by storing player input versus time somehow?) Awesome article, I really appreciated the clear, humorous, and concise explanation of these imporatant game design issues!! Long live Super Meat Boy :)

Solidplasma says:

April 17th, 2010 4:09:01 pm

This was a great, funny explanation. You guys have the perfect balance of challenge and reward. The beating children joke was just genius.

I'm actually making a game with short levels, but with lives. Thing about it is, the lives are part of the core gameplay mechanic, so when you run out, the level restarts.

As for the levels on ramped up difficulty mode, does anyone have any comments on the idea of adding additional difficulty factors/gameplay mechanics rather that modifying the level itself? I'm just wondering.

Edmund says:

April 17th, 2010 11:08:13 pm

@Solidplasma In a few weeks ill talk indepth about adding mechanics that add to dificulty in a post about fully exploring gameplay mechanics.

@Rorkimaru this formula only really fits the platformer genera of games, but i think you can take a lot from it and apply it to others.

@Provider I was waiting for someone to point out the obvious. Yes SMB had a secret continue option but it was hidden.. im not talking about hidden mechanics.

Cats says:

April 18th, 2010 1:19:33 pm

You are a smart guy. I've been wondering for some considerable time how I'm going to be able to make difficulty 'work' in an upcoming (probably-not-very-good) platformer I'm preoccupying myself with and your article really made a lot of sense.

The approach you take is similar the the VVVVVV approach - death has virtually no penalty, but the game itself is very hard. I personally think there's only so much you can squeeze out of this formula, it encourages players to be pretty reckless and it can lead to them winning just by 'bruteforcing' a level.

There's nothing really wrong with that, but for my platformer I decided upon big open levels with multiple pathways. While the game is very hard, there are many checkpoints and multiple exits to the level, so if the player is having trouble going down one path they can backtrack to the nearest alternate route and go there as well. There are also exploration\collection elements so if the player does head into an obstacle they can't overcome, their time isn't wasted. If you put regular checkpoints in and a generous number of lives (say 25) before you have to restart the level it means that when the player does restart the level they don't have to do the whole megaman-style 'play the whole level again in order to get to that one part you can't do' thing which I hate so much.

Of course, I have yet to see if this formaula works because the game is in the very early stages. But it's always nice to try new things.

Difficulty curves can be hard to nail, as can pleasing everyone, so your solution with the dark world levels is quite elegant indeed, reminiscent of what Super Mario Galaxy did. Best of luck to ya!

Doctor Professor says:

April 18th, 2010 3:03:28 pm

It's fantastic seeing a game developer talk this way. :) These ideas - at least the parts that aren't specific to Super Meat Boy - are very similar to ones I've discussed in my own blog, most notably in this article: http://www.pixelpoppers.com/2009/11/t est-skills-not-patience-challenge.htm l

The decrease in player punishment is one of the most important and strongest trends in the evolution of game design, and it's good to see it embraced, and the reasons for that embracing phrased so clearly.

My own writeup doesn't have the adorable pictures, but you may find it interesting nonetheless. :)

FlashByNight says:

April 18th, 2010 8:21:58 pm

Great article. Balancing levels and difficulty is a real challenge.

Namo says:

April 19th, 2010 12:59:02 am

The biggest issue I had with the original Meat Boy was not knowing what levels had bandages and what levels I missed bandages in, making for a lot of back tracking.

crackattackspider says:

April 19th, 2010 4:33:24 am

supose no1 has played torchlight a gd isometric dungoen crawler with 4 dificultys easy, normal, hard and super hard with a tick box saying "hardcore" if you tick that box ur death is permenent so that 10 hours you spent getting to that point is wasted now that is rough

Flying hippo says:

June 28th, 2010 1:25:16 pm

This is brilliant, This is the first time I'm more excited about a downloadable game than any retail game.

Chad says:

August 17th, 2010 1:55:47 pm

I was just linked in from the AVClub game preview and Super Meat Boy looks fantastic. After reading this, I'd be really intrigued to hear your thoughts on 'Limbo.' The creators of that game should take a crash course in the above. While it looks interesting and there are some great 'wow' moments, I think it fails because not only are there no penalties for death you dies so often that it is impossible to have any kind of connection with the game or the character. I'm glad that you guys are thinking about it in a more developed sense.

Chad says:

August 17th, 2010 2:21:03 pm

I was just linked in from the AVClub game preview and Super Meat Boy looks fantastic. After reading this, I'd be really intrigued to hear your thoughts on 'Limbo.' The creators of that game should take a crash course in the above. While it looks interesting and there are some great 'wow' moments, I think it fails because not only are there no penalties for death you dies so often that it is impossible to have any kind of connection with the game or the character. I'm glad that you guys are thinking about it in a more developed sense.

Jason Lee says:

August 28th, 2010 2:09:37 pm

Thanks for this! This is exactly the kind of thing I'll be thinking about in the next year (I'm doing a thesis on this very topic, which reminds me: I'm making games in Grad School and getting away with it. Huzzah higher education!) and I love it when designers give something very basic like this some real, hard, thought. I like your thoughts on rewards, which is something that I haven't been too deeply concerned with. In my experience with IWBTG, the mere act of navigating to the next screen was its own reward, but perhaps that reward comes from the metroid-y level design also, as each screen serves as a massive obstacle that hinders your exploration of the next screen: the tantalizing desire to explore serves in an indirect way as the reward in that game, which is something you can't do in a bite-sized level design as well.

Something I remember distinctly from Super Metroid and Chrono Trigger was also the reward of mastery, of encountering a past obstacle that was at first difficult but then rendered easy because of the powers you've unlocked between now and then. Charging through a hallway in Super Metroid with the speed upgrade when before you had to carefully navigate the nest of vicious monsters makes the player feel a very satisfying sense of power, and one thing that IWBTG (probably purposefully) didn't do was have you be able to revisit the same challenges again, but rendered trivial due to some screw attack or space jump or what have you.

says:

TESTING




Post Comment!
Loading

"A hardcore superstar is born. the weirdest game to hit these pages in Xbox 360's history." -Official Xbox Magazine

@li5up6 that build is very old and broken :) you cant get past the start screen  Today
Congrats @terrycavanagh, @gaijingames and the Limbo team on making it into Indiecade! looks like an awesome year! kinda regret not entering.  Today
http://bit.ly/c3ijP9 Dtiod posts mysterious images from "teh INternets"  Today
Edmund McMillen Edmund McMillen

Edmund draws stuff and designs things.. whatever

Tommy Refenes Tommy Refenes

Tommy programs and macs on the ladies.