Dynamic Difficulty Adjustment in Final Fantasy, circa 1987
Dynamic Difficulty Adjustment (DDA) is something that has been popping up a decent amount in the video game world recently; games that automatically, or through player input, adjust the difficulty to make games easier or harder to play are appealing because they make titles usable by large ranges of the population. I realized this past weekend, while going nuts with the cgcmarathon, that it’s been around for a while and that the ORIGINAL Final Fantasy, a Dungeons and Dragons inspired 8-bit classic on the original NES, had a great self-DDA system that let the user choose how hard they wanted the game to be at the very beginning of the game! The only problem is that it isn’t clear at all that is what’ s happening, since it is the “party select” screen, and doesn’t give any hints about how hard it will make the game.
So how does Self-DDA work in Final Fantasy? Let’s examine, with this easy to reference formula.
How to determine game difficulty in Final Fantasy based on party make up
When choosing your party in Final Fantasy, there are six characters to choose from. I’ve assigned them numbers here based on their usefulness:
5 – Fighter:Can soak up damage, dish it out consistently throughout the game, and is cheaper than the mages.
3 – Black Belt: Cheap and damage dealing at higher levels, this character loses points for not being able to take a hit and for being relatively worthless in battle until around level 10 or so.
2 – Thief:A slightly less useful version of the Black Belt, except he never really gets good.
3.5 – Red Mage:Can deal damage, wear some armor, and knows both white and black magic spells to boot. A bit cheaper than his white and black variant counterparts.
1 – White Mage:Can’t take a hit, can’t deal damage, very expensive and his healing isn’t super helpful. This guy (girl?) sucks.
2 – Black Mage:Also can’t take a hit but at least can occasionally deal damage. Very expensive.
Modifiers:
-3 – No Magic: If you’ve got no mages in your party, subtract 3 from the total points. Magic is generally less useful than attacking but has a knack of pulling you out of a hard battle here and there.
-3 – No Class Change: Not getting the rat tail and doing a class change makes the end game harder as there will be weapons, armor, and magic that your characters can’t use. The end of the game is somewhat easier than the beginning though, so it’s not as damming as you might think.
Once you’ve made your party, add up the total number of points from your four characters and check against the list below. The difficulty level assumes that the player has a basic knowledge of what to do in the game but is by no means an expert who knows all the game’s secrets.
16+ points: Easy, grind free playing. Just walk through the game and kill anything in your path!
13 – 15 points: Medium difficulty, it shouldn’t be too much of a challenge.
9 – 12 points: Hard, and you’ll have to grind at points, but doable.
8 points or less: Grind-tastic. Enjoy your hard game full of pain, disappointment, and failure.
That’s it! Pick your party and you are ready to go, knowing that you have defined how difficult your game will be.
My “I want to take it easy and not have to grind at all” party? It’s these guys.
The party I played with during the CGC Marathon was probably the worst one possible. SCREW YOU, ALL WHITE MAGE PARTY!
March 5, 2010 | | Comments (0)
Category: Games, Numbers, Random
Tags: adjustment, belt, black, DDA, difficulty, dynamic, Enix, fantasy, fighter, final, mage, magic, NES, points, red, Square, thief, white
On the Relative Difficulty of the Mario Kart Titles
Abstract
The four console incarnations of Mario Kart (SNES, N64, GC, and Wii) are compared in terms of difficulty to complete 150cc races (the hardest setting) with a gold medal, that games’ highest honor. A team of game development ninjas were unleashed upon the game over a 36 hour period to determine comparative difficulty, playing single player races in every 150cc cup event for the four titles. Collected data indicate that the order of difficulty is
SNES SMK > GC MK:DD > N64 MK64 > Wii MKW
With the original Super Nintendo title being the most difficult, and the newest Wii title being the easiest. Readers are welcome to verify and corroborate our findings with their own experimentation.
Introduction
The Mario Kart franchise is one of the most beloved racing series in existence, boasting a title on every major Nintendo console since the Super Nintendo. While most users simple play the game in battle mode against inebriated comrades, the astute gamer will note that hidden away in the main menu there is also a “single player” mode in which users can race against the computer in a series of races. When played on the hardest “150cc” setting these races are infuriatingly challenging, and we therefore felt it was necessary to have an understanding of the relative difficulty of the four main console Mario Karts (the SNES original, Mario Kart 64, Mario Kart: Double Dash, and Mario Kart Wii).
Methodology
For the complete game completion marathon team Fire Hose decided to race each and every Grand Prix cup event and win the gold medal on the hardest 150cc setting, meaning 20 gold medals (4 SNES, 4 N64, 4 GC, and 8 Wii). Researchers/minions were locked in an airtight room at GAMBIT and not allowed to leave for food, water, or sanitary reasons until all 20 gold medals were earned. Difficulty was determined by listening for the amount of cursing emanating from the room, how long each medal took, and how often researchers swore they would “never play this fucking game again, for any reason”.
Results
Mario Kart Wii was first title tackled and the easiest of the bunch; an expert user can expect to win these races on a regular basis, and the various boost mechanics (popping wheelies, power sliding, midair tricks, and prolific booster pads) offer multiple opportunities to get ahead of the competition. Mario Kart 64 is only slightly more difficult, as the rubber banding mechanics aren’t so punishing as to prevent the first place racer from occasionally acquiring useful items. There is a large element of chance in this game though from environmental hazards, such as cars in Toad’s Turnpike and Bullet Bills in Rainbow Road.
Mario Kart: Double Dash is somewhat harder than the previously mentioned titles; rubber banding on items is more pronounced making it harder to keep a first place lead, blue shells are plentiful for racers towards the end of the pack, and the computerized opponents are adept at power sliding and stealing items from the unwary player. However, the hardest game of the bunch by far is the original Super Mario Kart. This game’s 150cc mode can be aptly described as “dickish”; computerized players have at-will abilities such as a feather jump and special attack that are severely damaging and often unavoidable, there is no rubber banding of items to help a player catch up from behind, computerized players that fall behind gain super human speed in order to catch up, and the coin mechanic (ditched in later games) adds an unnecessary level of complexity and another potential failure point.
Discussion
The original Super Mario Kart was likely so brutally difficult due to an inability to program especially smart AIs. Rather than spending lots of time trying to (most likely unsuccessfully) make smart AIs the programmers simply opted to create a set of advantages for computer players that just feel flat out unfair. Perhaps this helps to explain why the N64 version was easier, as maybe the developers wanted to create a more accessible sequel that would be less frustrating. With the advent of the Gamecube, a system designed to speak more to “hardcore” gamers, it seems that Nintendo made their greatest effort to create a truly balanced Mario Kart that would appeal to and work for a broader slice of audiences. Of course, with the Wii’s focus on a more casual audience it made sense to tone down the competitive nature of the game to make it easier for less frequent gamers to pick up and play (and play well).
Was this Nintendo’s actual thought process making these games? We don’t know, but after playing these games all weekend we’d like to think so.
References
Complete Game Completion Marathon Website (you can still donate!)
Fire Hose for Haiti’s games list
March 1, 2010 | | Comments (0)
Category: Games, Rant
Tags: 150cc, Abe Stein, abstract, discussion, frustrating, GAMBIT, Gamecube, GC, gold, Kart, marathon, Mario, medal, methodology, N64, nintendo, Rainbow Board, research, SNES, Wii
CGC Marathon, helping Haiti
Yeah, I just did a blog post about this last week, but it’s a good cause so here it is again. The Complete Game Completion Marathon is this weekend and they (we?) need your support! As of writing this we’re just above $2,200, well away from our rock star goal of $10K. Donate if you can afford it! And then check out the website and watch us make fools of ourselves for you and the rest of the intarwebs this weekend.
Not sure how to donate? Click this big link below! Want to see what we’re doing? Check out last week’s blog post.
February 24, 2010 | | Comments (1)
Category: Events
Tags: charity, complete, donate, GAMBIT, game, Haiti, marathon
What We’re Playing: Robot Unicorn Attack
Need a laugh? Have a burning desire to play more video games that look like they were based off the old Trapper Keeper binders from middle school? Well, you’re in luck! Robot Unicorn Attack is a new video game from the nuts over at Adult Swim Games and is definitely worth a play. The premise is pretty simple; you’re a robot unicorn chasing dreams and fairies with your three wishes. The aesthetic is what’s truly memorable (nauseating) as you’ll see when you start playing. I highly recommend playing with the sound on – the soundtrack is epic.
As a side note, I have to hand it to the Adult Swim Games guys. Rather than just churning out crap titles based on their well known IP they focus exclusively on new content that generally has a really catchy idea behind it, and their games are generally higher quality as a result. Well done!
I beat 20,000 on one wish. I challenge you to do better.
February 19, 2010 | | Comments (0)
Category: Games, What We're Playing
Tags: adult, always, attack, erasure, fairy, robot, swim, unicorn, wish
Fire Hose for Haiti
We’re taking part in the GAMBIT Complete Game-Completion Marathon in two weeks, it’s a fund raiser with all proceeds going to help the victims of the recent earthquake in Haiti. Over one weekend we’ll be spending tons of time playing some crazy games and trying to get people to donate to the cause. Here’s what we’ll be up to!
Games: It’s the Three Ring Circus!
1. Final Fantasy, All White Mage Run (NES): We will complete the entire Final Fantasy with the worst party possible, four white mages. No partial credit here; we only get points if we finish the game and defeat Chaos.
2. Mega Man 9 Challenge Run (XBLA/PSN/WiiWare): Mega Man 9 is a ridiculously hard game with 50 insane challenges that make the game that much harder. We will attempt to finish AT LEAST 40 of the 50 challenges, with extra credit for each achievement beyond 40.
3. Mario Kart Extravaganza (SNES, N64, GameCube, Wii): We will play through every 150cc cup in every console Mario Kart game. Our success will be based on the LOWEST trophy level we get; thus if we get 19 golds and 1 silver then our performance is rated silver. We will be trying for all golds!
Bonus: While we play we will blog, pontificate, and generally do interesting stuff in front of the webcam for people following online.
Extra New Bonus! World famous Fire Hose artist extraordinaire Jacques Pena will be making incredible video game art of the games we are playing and taking requests during the weekend. Contributors who donate a certain amount (probably $100) will be getting free one-of-a-kind art as a thank you! Details coming soon.
If you want to donate go to our page on Partners in Health, or just click on the big thermometer up top!
February 12, 2010 | | Comments (2)
Category: Events
Tags: charity, complete, donate, fantasy, final, fund, GAMBIT, Haiti, Jacques, marathon, mario kart, mega man, Pena, raiser, thermometer
GGJ 2010: Sprinkle
The title screen (and first level) of Sprinkle
Jason wasn’t the only Fire Hoser to participate in the Global Game Jam last weekend. I was also in the trenches slapping a game together, only I was on the other side of the river at the Northeastern University site.
The result is Sprinkle, a neat little puzzle game based on the idea of raindrops falling through a tree. The goal in each level is to move objects around to guide the falling water droplets to a flower. Once the flower has fully bloomed you’re taken to the next level. Go ahead and give it a shot — it only takes a few minutes to play through.
Looking at the game again after letting a week go by, I’m impressed with how fully developed the game seems for a two-day effort. It contains a coherent core mechanic, some interesting variations, and a fair number of well-thought-out (if simple) levels. This is all the more impressive considering that, in true game jam fashion, the first playable level came together at roughly 1:00 Sunday afternoon, only two hours before the deadline!
I was working on music and sound design for this game. The other team members were three programmers — Shakeib Afzal, Kirk Israel, and Heinz Pabst — and Zachary Fand, who created the nicely idyllic visuals.
Early on, I decided to try to put a bit of musical interactivity into the game, similar to that in a game like Elektroplankton. Specifically, the music I had in mind would go back and forth between two chords, and the sound effects would be designed to fit into these chords, so that the sound effect that any particular object makes would change depending on what part of the music loop is currently playing. The results are mixed; sometimes it comes off as intended, but in most levels the experience is a little awkward — either there are too few objects for anything interesting to happen, or there is a flood of objects creating an annoying and broken-sounding cacophony. Partly this is inevitable in a game jam, when exactly what the game is going to feel like is not necessarily determined until close to the end. But there were also some good takeaways for me.
First, as part of the sound design process, it pays to think carefully about what kinds of events will be triggering sounds, how often they will occur, and how likely each event is to repeat. If I were to revisit the sound design after seeing all of the levels that made it in, I would definitely reconsider the scheme for assigning sound effects to collisions, trying things like having collisions turn on or off a loop, or having each object emit a sequence of notes as it is hit multiple times rather than repeating a single note. Second, it’s important to pay attention to the technical details early on — can we get the music to loop smoothly? How are we going to handle repeating the same sound a dozen times in close succession? I’m happy with the overall sound of the game, but I would be much happier still if we had sorted this stuff out early! Having the technical aspects be a little bit off makes the whole experience much less polished and enjoyable.
I had a really fun time putting the game together with the team and was impressed by the level of energy, enthusiasm and creativity from everyone at the jam. If you’ve never tried doing a game jam before I highly recommend it!
February 8, 2010 | | Comments (1)
Category: Uncategorized
Tags:
What We’re Playing: RunRunRunJump
The Global Game Jam was last weekend, and a lot of amazing games were made (for a recount of how one went down check out Jason’s awesome dev blog). A full run down of the more interesting Boston offerings is coming soon, don’t worry!
The current front runner for our favorite GGJ game? That would have to be RunRunRunJump, a delightful and short game by Owen Macindoe, Scott Macmillan, Daniel Perry, Alex Schwartz, and Stephie Wu. The game is a simple platformer, but what is interesting is that the environment literally tells you what to do – if the ground says run, then run! If it says jump, then jump! The soundtrack is especially amazing as it is self referential as well – listen for a few seconds and you’ll know what I mean.
Of course, the theme of the game jam was deception so you should expect a couple of twists to be thrown at you as the game goes on. Keep playing though – the victory screen at the end of the game makes it all worthwhile.
One important thing to note is that the game requires the Unity web player to work, download it here. And then go play RunRunRunJump!
February 3, 2010 | | Comments Off
Category: Games, What We're Playing
Tags: game jam, global, jump, run, runrunrunjump, unity
GGJ 2010: Post Mortem-ing a Weekend
The Last Bullfight: Our beauty render
Global Game Jam 2010 Post #3:
The weekend is over and our game is playable (!).
You can find it here: http://www.globalgamejam.org/2010/last-bullfight.
We recommend the web/flash version (may ask you to download/install the small Unity/flash player).
If you are stuck in any way, please email me: Jason@FireHoseGames.com
What Went Awesome:
* Treating Each Other Well: listening to each other’s ideas, encouraging each other to take water and brain breaks, keeping good humor, appreciating each other’s contributions.
* Making Good Decisions as a team at regular intervals on what features to keep and what to toss in order to get the project done: late Saturday afternoon, late Saturday night and first thing Sunday morning.
* Planning Our Work Flow: in our design doc we structured the weekend into 3-hour block “sprints.” This helped to guide our productivity and made a good balance between being able to dive deeply into our work and regularly checking-in with each other. We also got each other’s contact info right at the start, chose a safe source system, and established an “it is OK to recuse yourself early from meetings in order to be productive” policy.
* Allowing Pablo Picasso to do all our concept art for us.
* Research: we spent two hours on Friday night watching and discussing bullfights on Youtube, and came away with a clear vision for our subject matter that supported our thinking all weekend.
* Manageable Art Scope: making a game with only one location and where the main animated characters–the bull and the matador–are frequently invisible made it much easier as the sole artist to produce any kind of quality rather than just quantity.
* Our wacky deception mechanic, to our surprise and delight, actually seems to make sense to people.
What Could Have Went Awesomer:
* The free version of Unity did not play nice with our choice of version safe software, SVN, and by my estimate our (amazing) two-person code team lost 20% of the weekend dealing with issues around synching their work until downloading a professional trial late Saturday night.
* Possibly a better balance between tweaking and implementing features: considering we had so little time to get our game made before the 3pm Sunday deadline, I wonder if we could have spent less time tweaking details of the game on Saturday and more time getting features in as quick and dirty as possible. As team artist, I was aware of how motivating it was for my team to see semi-polished art early on, but also that having rougher animations available earlier than the Saturday evening sprint would have helped the team to get gameplay feedback features in sooner, allowing us to test the game before the bleeding deadline edge.
* Defining/addressing Work Needs Better: we allowed one of our teammates, the research/QA lead, to work essentially on his lap all weekend when we should have gotten him a table, and another needed more quiet space than we found for him early on. A better discussion of what each person needs to be productive should have been part of our first meeting.
* More documentation: it is hard to document everything when creating a game at 90 miles an hour, but the few times we did document our work made such a huge difference in our thinking and productivity that I suspect any bit more would have helped.
We want to thank Rik Eberhardt and Phillip Tan and everyone at the Singapore-MIT Gambit Game Lab for hosting this fantastic event and taking such good care of us all weekend, from the great organization to tech support to arranging meals. We recommend you try out all the games made at MIT this weekend. Some we are particularly fond of include:
* Run Run Run Jump: More joy than any single game should be capable of producing in one weekend. These guys shared a space with us, so yes, their sound track is permanently scarred into our brains: http://www.globalgamejam.org/2010/runrunrunjump
* Press X to Not Die: Amazingly clever and deliciously mean game, a send-up of Quicktime events with an exceptional amount of beautiful 2D animation: http://www.globalgamejam.org/2010/press-x-not-die
* Quest for Stick: Lovely Braid-like look and feel, with a very cool mechanic for interacting with the world: http://www.globalgamejam.org/2010/quest-stick
February 1, 2010 | | Comments (1)
Category: Uncategorized
Tags: bull, bullfighting, game development story, Global Game Jam 2010, matador, singapore-mit gambit game lab, The Last Bullfight
GGJ 2010: The Most Heartbreaking Laughter
My Team, EL MATADORS: Cole, me, Mike, Nazer, and Matt. You can't tell, but we are each holding little Mariachi band figures that decorate our workspace this weekend.
Global Game Jam 2010 Post #2:
It is 12:30am again. Day two of the 2010 Global Game Jam is in the bag. I planned to post an update in the afternoon, but there was not a second to spare all day. We sprinted through 15 hours of development today, and found ourselves in all three of what I am told are the classic Day-Two moments: a serious talk in the afternoon about what features can be kept and what needs to be discarded for the sake of completing the project, a major technical crisis where the tools broke in the face of our ambition, and reaching a place where the most frustrating and incomprehensible bugs brought laughter instead of tears, where the relief of having gotten this far, and the pleasure in doing this together, won over whether or not the thing we want to call a game ever becomes what we hoped.
And we were not the only ones. I overheard other teams talking about what they could cut and heard heart-rending stories of programmers losing their entire work so far due to a computer error of some sort (even in a 3-day project–or especially in a 3-day project!– a version-safe system is critical). And slap-happy giggles filled the corridors between the MIT Gambit labs in the last hours, developers enjoying after a long day the many interesting ways their games managed to break.
So how is “The Last Bullfight” coming along, you ask? Thanks to the art direction of our good friend Pablo P, in-game art and animation are pretty much complete (at least, sufficiently to show all of the features we planned). The sound work is in good shape, including a powerful song composed by our designer. The code features that have been completed are successfully producing an atmosphere, a sense of perspective and place, and we are happy about that so far. While the back-end code of the basic kill-or-be-killed game play is in, there is no real feedback yet to the player on which is occurring, which we understand is the difference between having a game and having a weird little interactive movie. We hope to have a game tomorrow.
One day left. See you on the other side,
Jason
January 31, 2010 | | Comments Off
Category: Development Blog, Events, Games
Tags: game development story, Global Game Jam 2010, Jason Wiener, MIT Gambit, The Last Bullfight
Global Game Jam 2010: Artist Jason Reports From The Front
Global Game Jam 2010 Post #1:
It’s Friday night, 12:30am. We have just been kicked out of MIT’s Gambit labs to get a few hours of sleep before a full day of development tomorrow. The last eight hours are a blur; meeting dozens of developers, hearing the theme and constraints of the 2010 jam (International theme: deception. Local theme: abstraction. Constraints to choose among: rain, plain, Spain), taking 90 minutes to discuss ideas, with proposals flying so fast and furious my head felt full ten minutes into it, pitching our ideas, forming teams, and then getting as much pre-production nailed before they sent us packing.
I pitched a game (and found a willing team) where you are a bull in a Spanish bullfight. The goal is to kill the matador before he kills you. Being in the bull’s perspective, the environment we are planning is distorted; black and white Pablo Picasso-style architecture and characters, where the matador is practically invisible unless you are a few feet from him, but his flowing red cape tantalizes you at every turn (yes, we know Bulls can’t actually see the color red. Game is more fun this way. We think.).
We started by making sure everyone had everyone else’s contact info, wrote up a design doc and rough asset list, discussed how we plan to communicate and treat each other, decided on tools and how we will handle version control, and looked up reference material. While beginning a VERY rough prototype, we watched about two hours of bullfighting video, and found ourselves, already in the mindset of the bull, booing the humans.
Can we get this done by Sunday? We have two programmers (one an industry pro, one currently at MIT) using the Unity engine, one sound designer from Berklee, a QA/research lead from Sloan, and me making the art. I have no idea where we will be on Sunday, but I am excited about the team.
More tomorrow,
Jason
Who would kill this cute little bull?