adventures in game design

Five simple steps to designing melee attacks

Designing melee attacks has been part of my weekly routine for a while now, and I think it's time I sit down and distill the process. Here are the steps, along with an example.

First, I sit down and identify the key points of the attack. What is the primary purpose? Is it a heavy or light attack? Is it part of a chain or series? Does the previous pose matter? How long should it be?

• So lets say that we're implementing an overhead katana attack. It's a one-off attack, but it needs to come from the base-combat pose. This attack should appear powerful, as if the character is slicing through a tameshigiri

Second, we get it to the animators! The important part of the above step is to know ahead of time what is the main thing that needs to be communicated with this attack. The animator will inevitably put his own creativity into this, and needs to know what to keep in mind. I try to act out a couple swings myself, but always try to end with "the main thing is ____". Props are great for this part.

• Makeshift weapons are great props. A nerf sword, piece of PVC pipe, or practice kendo sword works great for this attack. The most important part is to convey the notes we made above. So as I practice a couple swings in front of the animator, I repeat the words that matter most: Powerful, slicing, and from the base-combat pose. As soon as we get that part communicated, get that prop in the animator's hand! He's much better at imagining this stuff than you. Let him work it out. Some animators like to draw the poses out first too.

Third, we have the animators only work on three poses. This is called pose-to-pose animation. What we want to do here is not waste anyone's time. The animator's best use of time right now is getting 3 poses on screen for you and anyone else to critique. The Anticipation, Actualization, and Follow-Through. In fighting game terms, this is the Wind-up, Hit Frame, and Recovery. Eric Williams has a great blog post about this. This lets everyone agree quickly on the framework of the move.

• In our example, we happen to be designing Haohmaru's A+B attack from Samurai Shodown. Here are the poses we end up with:

• Wind-Up/Anticipation  
• Hit Frame/Actualization
• Recovery/Follow Through  
   

Fourth, We plug these poses into the game, hook up damage collision, the necessary input sequence, and the scale the timing for each pose so that we get a basic timeline for the motion. We then report the timeline to the animator in terms of something he can use when he animates the in-between frames. We might also report if anything isn't quite right, like if the sword is going over a short enemy's head. It's very important that the designer has the tools, or the knowledge and authority to tweak and tune the timing of attack animations. After all, you're responsible for making the game play correctly, right?

• So using our example, we'll set it up so that it does damage appropriately, causes the right hit react on enemies, and then we'll adjust the timing between the poses until it plays about right. Once we're set on the timing, we'll report back the animator with something like this: "Hit Frame should be at frame 14, and recovery should end at frame 34".

Fifth, the animator makes everything look awesome using the guidelines we've set. He plugs it into the game, and we play with it and scale the timing one last time.

• We might notice that once the animation is complete that our original timing doesn't work as well now that the animation is fully keyframed. We fine tune the timing this time so that it looks and plays as good as possible. And we're done!

 

These images were taken from Fighters Generation. It's a great resource when you need ideas.
 

Responsiveness, feedback, and iteration. Part two.

    Pro Evolution Soccer is a great game series, but it is absolutely horrible in regards to responsiveness. In PES, the designers went to great lengths to get the actual feel of a soccer game by perfecting the soccer ball physics and taking into account each player's particular strengths and weaknesses, but they sacrificed responsiveness along the way. When you press the X button to pass the ball in PES, the game stores that input as a request and then you have to wait until the kicker has caught up to the ball, gained control of it, and placed his foot before the actual kick animation begins. Sometimes this can be 3 seconds away if the player is far from the ball. However, the X button also happens to be the command for canceling a pass. You can see where this is going right? Many times a player will press the X button to pass, and the game quietly buffers the request and continues to simulate the game. Meanwhile, our player notices that his command didn't execute, so he pushes the X button again. Now the game cancels the kick request, and when the striker gains control of the ball and stops moving, the pass never happens.

    To the untrained player, this is simply a broken experience. The punishing situation above teaches the player that the X button does nothing. However, eventually that player might learn the nuanced use of the pass mechanic. But this still causes trouble for me even though I understand the mechanic. Often I will press the X button to buffer a pass, but then while I'm waiting for the action I wonder "did I press the button hard enough?" and I'll press it a second time. Other times I'll mash the X button in a tense moment, only to realize afterward that I have no idea how many times I pressed the X button so I don't know if the last press meant Pass or Cancel Pass.

 

   The decision Konami made with PES was a more realistic soccer game instead of a game that responded instantly to player input like World Cup Soccer for the old school Nintendo. The interesting thing about PES is that it is still a highly regarded game despite it's lack of responsiveness. I appreciate the attention to detail of kicking with the left or right foot, how the placement and spin of the ball have just as much effect as the weather and team's energy levels. I agree with the decision to sacrifice instant responsiveness for more realistic kicks and sprints and nuances of the real game reflected in the simulation. The problem with PES is not the lack of responsiveness, it is the lack of feedback to the player.

    Imagine being at a fancy new restaurant and ordering food through a computer at your table. You press the cheeseburger button, the onion rings button, and the Guinness button while the computer screen shows you a message that says, "Welcome to fancy burger! press the buttons below to order!". Well, imagine that the fancy restaurant doesn't confirm your order until a cashier behind a curtain actually accepts the order at another computer terminal. Since your order hasn't confirmed your button pushes immediately, you would logically think that the order wasn't going through. Thinking that the computer isn't working, you might end up pushing the cheeseburger button 5 times before any sort of confirmation comes up, at which point it will say "5 cheeseburgers, 3 fries, 3 Guinnesses - would you like a shake?". The system works, but it doesn't fulfill the communication requirement of letting the person making the requests that their request has been logged and is waiting for execution.

    Like the above example, in games we omit the essential part of communication from our system. If our responsiveness is not instant, we must let the player know "action confirmed, just a moment". We've taken great strides in making games more realistic. Jumping, running, throwing, and virtually any action we can think of an avatar doing (besides maybe pulling a trigger) has some time cost now, and often we pad additional time in to get the transitions to look smooth.

    All PES needs is a UI element on the screen to let me know if the pass action has been buffered. Draw the X button image when true, and remove the image when canceled. I haven't even said the word yet, so here it is: Feedback. It's what makes games games and not interactive movies.

bazooie's Top 10 Favorite Games!

M.E. has a fun post about her top 10 favorite games (defined on a purely subjective level). I really identify and agree with her observation of how her favorite games correspond to personal connections. It didn't take me long at all to come up with my top 10. So here are my top 10 favorite games in order of hours played (shortest to longest)

Maniac Mansion 2 - Day of the Tentacle: Frozen Hamsters! Time Travel! George Washington's Wooden Teeth! I fully credit this game with my knowledge of computers. My desire to play this game gave me the motivation to learn how to bend DOS to my will in order to get this game to play correctly on our 386 with 1 MB RAM. Oh how I bent it.

Space Quest 2: Vohaul's Revenge: I was the primary player, but it took my whole family to solve this game. I still remember my Mom, sisters and I spending weeks trying to figure out how to cross a bog with a man-eating monster.

We <3 Katamari: Boldly, definitely foolishly, I took this on my honeymoon with a PS2 as an attempt to get my new wife into games. Now Lindsay and I have fond memories of the music, and learning how to steer a katamari together (after much arguing on which direction to go). This game enhanced our communication skills!

Toe Jam & Earl: I loved playing this as a lad with my friend Jason Marlis, and even now I play with Lindsay. It used to be so hard to resist the urge to open every unknown present! I would always end up scrambling the gifts and dying early. The randomly generated worlds, and focus on escape rather than combat make this game still fun.

Halo: I hated this game at first because it was so SLOW compared to the PC shooters that I played. But Halo magically broke down the 'nerd' wall, and brought my roommates and friends into the hardcore gaming world. We had late night house meetings discussing strategies on Blood Gulch, and went through hoops and hacks to play online. Joyous celebrations, and loud arguments all stemmed from our experiences with this game.

Street Fighter 2: This game changed my life. I had a competitive nature already, but when SF2 came out, it gave me a community and a vehicle to feed my competitive side (turned me into a monster!). Despite being 10 and 11 years old, Jason and I could hold our own. That fact alone broke down a lot of self-esteem issues I probably would have had, and changed my paradigm of what age meant.

ATV2 Ice Hockey: This remains the only game that I ever worked on that I continue to play. The Ice Hockey mini-game in ATV2 has the perfect proportions of skill and randomness to make for one of the best competitive multiplayer games of all time. Patrick Connor and I have spent easily over a thousand hours playing this game head to head, and it doesn't get old. He just left to work in Maryland. I miss you, Pat.

Diablo 2: This was my first RPG ever, and it sucked me and my life away. This game taught me how to be efficient with my time. Seriously. I was the biggest slacker before Diablo 2. In fact, I stopped playing it because in the grand scheme of things, it really was a waste of my time since I didn't play with friends. Eventually, Pat and I had bots play Diablo 2 for us, and sold items on Ebay.

Counter-Strike: When I started as a tester at Crystal Dynamics, I was invited to play CS with the dev team during lunches. They were good. I sucked. As co-workers laughed and sent me screenshots of my avatar getting knifed and blown to bits (I'm looking at you Benny and Billy), they had no idea how much my internal competitive monster was stirring. I played religiously until I was top tier at work, and I was hooked from then on. Useless Trivia: I am one of the guys getting yelled at in the legendary song by Kurt Harland, The Terrible Mr Grimshaw.
 
Tekken Tag Tournament: This game truly is a part of me. It's competitive, it's technical, and plain fun. Without Tekken, I wouldn't be in the industry, or wouldn't have half the friends that I know and respect now. I've easily spent thousands of dollars and hours on this game, and I love it despite it's many flaws. It has become a method of communication between myself and anyone I play with: all over the US, in Korea, and in Japan. I started playing because it looked cool, and I continue playing because of the community.

 

Show me, don't tell me

Yes, it sounds like something a righteously angry significant other would say when you're trying to smooth things over. There's probably a nicer way to say this in the workplace that doesn't conjure up feelings of defensiveness and guilt...

I just wanted to document the merits of prototyping something as opposed to discussing at length what something might look or play like. Typically, designers like to postulate, theorize, discuss, argue, blah blah - we talk a LOT. Including my own desire to be always accurate, this can lead to some really long conversations about how to implement a particular system. Inevitably we end up talking so much that we could have developed the ideas on screen in the time we spend talking about how to develop our ideas.

Back when I was an aspiring graphic design student, I recall discussing next steps with my professor after a critique. I had a great idea on how to improve a particular composition and was describing how it related to my most recent work. About two sentences in, he said "It all sounds great - but I really can't judge until I see it. I need to see it." It was one of those mini-life lessons that 10 years later means even more to me when I think about it.

In the past few weeks I've been working in prototype pie in the sky anything goes land, and also in deadline driven get it done land. In both situations, I've noticed this tendency to debate being destructive to creative development. The only thing I've noticed that it improves is persuasive communication skills. In an effort to combat this tendency, many of us have committed to simply proving out our ideas in tandem; sometimes sacrificing our own biases to help someone else get their idea working. The end result has been absolutely amazing: discussions typically pockmarked with opposition and contradicting ideas give way to strong shared conclusions. Not only that, but the amount of trust and respect that is built between designers that originally had differing opinions only to reach a refined conclusion through this process is inspiring and powerful.

Will collect your bad ideas for achievements

After playing Assassin's Creed, I had to spend some considerable time working out why I didn't like the game, because I was also very impressed by it. One reason that surfaced was how much I hated collecting flags for nothing else but Achievements. Now, I haven't played the PC version, but I'm guessing there probably aren't any achievements - so are we expected just to collect flags? Where's the gameplay in that? It's like an Easter Egg hunt, without the candy.

Let me look a step upward at a game that incorporated collecting in a slightly improved manner: GTA. To be fair, I hated collecting in GTA as well, but at least the reward was in-game, and was meaningful - The gameplay: Collect 100 dubious brown bags, and you would unlock a weapon or armor for all time availability.

After all this thinking about collecting, I realized this simple truth:
 
Collecting stuff gets harder as you progress.

Take the real world example of the Easter egg hunt. Say that Mom hides 10 eggs around the yard, and you have to find them. The first 3 might be really easy to spot, the next 3 take some more time, and the next 3 take even more time. But it's the last egg that gets you to stop and wonder if Mom really only hid 9 eggs, but told you there were 10 as some cruel joke to get you to play outside longer. The point is that as each egg gets found, and subtracted from the world, the chances of finding the next egg diminish (and therefore, the time it takes to find the next egg increases).

So how might a good collecting system be implemented? To challenge myself, I'll use Assassin's Creed for my examples to prove my point that something better could have been done using the existing tech (with maybe a few days of extra work).

• A) Unlocks at regular intervals of increasing value.
  
  • At 25 flags, concept art is unlocked
  • At 50 flags, hiding from enemies takes 50% time.
  • At 75 flags, Altair can hold 10 additional throwing knives.
  • At 100 flags, Altair's sword damage is increased by 25%.

• B) One big unlock, but introducing a system to 'detect' hidden items at regular intervals.
  • At 40 flags, flags in a small radius are added to the map when Altair climbs a perch point.
  • At 70 flags, flags are simply added to the map.
  • At 100 flags, Altair's throw damage is increased by 100%

I wonder if the function for finding items in a world is linear, or exponential...

As stone sharpens stone

So before I get straight into posting, I should probably explain the point of this. After a roundabout career path, I've finally landed a job in game design. I really can't shut my brain off, and most of the time I'm up until 3 am thinking about game design anyway, so I might as well have an outlet for all those thoughts. Lately, I've been keeping a notebook for most of the thoughts that hit me while I'm playing a game, but writing tends to be too slow and hard to share.

So the point. This is a medium where I can throw out game design problems, solutions, and conversations and give anyone the chance to comment, reply, argue. As stone sharpens stone.