Sunday, February 28, 2010

Game design from a gamer's perspective Pt. 4


Lighting is critical to any 3D game title. It is equally as important as the camera. While the camera lets you peer into the world, the lighting allows you to see (or not see) the world. Good lighting let's you see what you need to see. Great lighting sets a mood for the story. Setting up the proper lighting for any scene in a 3D world is critical to the mood you're trying to achieve and is one of the two major ways (the second way being music) to set the tone and mood for the game (and story). Games that have successfully used lighting properly include Bioshock, Chronicles of Riddick, Halo 3, Assassin's Creed, GTA4, F.E.A.R. and The Darkness (just to name the top games). No, Oblivion didn't make this list. While Oblivion has reasonable lighting, it's just not outstanding. Note that lighting needs to be combined with textures and shaders to complete this package. Perfectly lighting a poorly textured or shaded object doesn't do anything for your game. These two things go hand in hand... which is why texturing is next.

Proper lighting entails making sure the color of the lighting is accurate for the scene. Lights should produce a halo effect if very bright. If it can produce the film 'strings' (left and right horizontal streams), it can easily give your game a film like quality. In fact, I've not yet seen a game that's even done this. Lights on the backs of vehicles or during moving sequences should produce streams. Lighting should produce soft shadows, when possible. Lighting should always produce a shadow. Shadowless lighting is odd and makes your game look unrealistic. Also, unless you're trying to achieve a specific mood, shadows should never be 100% black. True ambient (GI - bounce) lighting always lightens up shadowed areas. So, even if your engine doesn't support GI, you can simulate it through the use of ambient lighting.

Texturing and Shading

While lighting is key to setting a tone and mood, shading and texturing brings out the realism. If you want to make a scene look photoreal, you need four things: natural lighting, high resolution textures / shaders, a high resolution mesh (or low res mesh with great looking normal maps) and natural environments. Clearly, in a video game, there's limited RAM. So, you have to optimize the game's memory footprint by taking some shortcuts. So, while you will need to take shortcuts, don't take them unless absolutely necessary. For example, don't make your mesh resolution so low that even a normal map can't fix it. Use a high enough resolution that the mesh looks good on its own. Then, add normal and displacement maps to increase the resolution and add realistic folds and creases. Combining that with great looking diffuse textures and you've got a winning combination.

Games that have successfully produced great looking human models include Mass Effect, Bioshock (at least the Big Daddy), Oblivion, Fallout 3, Heavy Rain, Drake's Fortune, The Darkness and a few others. Unfortunately, there are many more games that use low res textures, shaders and models. For the Mario's and Sonic's of the world, that's fine. But, for a human drama, don't skimp on shaders and textures.

Additionally, don't put tons of effort into your human characters by using 1024 x 1024 texture maps and then proceed to load 128 x 128 maps onto terrain surfaces. Keep consistent. Use high res maps for all surfaces or none. Don't pick your main character to make look great and then cheese out on the rest of the surfaces. Two Worlds was primo at this. High res character models, low res terrains. It looks horrible.. don't do it.


Shaders are what make your objects look like real everyday objects. It adds shine, transparencies, ambience and lots of subtle things to your characters. For example, the use of a specular map on skin surfaces is critical to making skin surfaces shine properly. Without a specular map, the skin surfaces look shiny and plastic. Again, for Mario, that's fine. For human drama, not so much. Always try to make an object look correct by using the proper levels of specularity and specular maps when possible.

Levels of Detail

Level of Detail (LOD) is commonly used in games. It saves memory for distant objects but also gives high res details up close. When possible, use it. Any console programmer likely already knows this, but if you don't, here you go. You can read up on the use of LOD at various sites including Wikipedia.

Character Models

What's to really say here. The best I can offer up is make them look good. Leave this task to your very talented character artists to design. I've rarely come upon games that have poorly designed characters. Occasionally it happens, but rarely. For example, I was a bit disappointed in the models for Crackdown. Most movie tie-in games tend to have poor character models. In many cases, the game designers choose to move the camera very far back from poor quality models. This can make the model look a lot better than it is. For level based games, a small character can work. For up-close-and-personal 3D shooters, that doesn't work so much.

Again, hire quality designers and modelers to produce your models. Texturing and shading them is, of course, a big way to make or break the model.

Graphics modes

I've been very impressed by various games that some designers have chosen to employ. For example, Halo 3 has almost an almost cartoon approach to the models, but placed in a very realistic environment. This gives the game an almost surreal quality. Master Chief looks like armor, but at the same time he looks like a cartoon. This works.

Crackdown and Borderlands, on the other hand, chose an outline system for the characters. While it looks fine for the first few minutes, the graphics quickly get in the way of the story. It's difficult to get past the outlines. If you want to use outlines in a game, use them sparingly. For example, GTA used a design outline approach to intro graphics, but not the game itself. The game itself had a much more realistic look. This also worked.

3D console games really do look best when you use the 3D system to look realistic. Trying to use a stylistic approach to 3D doesn't seem to work well with Direct3D or OpenGL. Perhaps these systems need a facelift, but the recommendation from Gamezelot is not to use outlines if possible. Although, experiment. If you come up with something that has a wow factor, maybe I'll reconsider.

Playability - Terrain and getting stuck

A big part of game design is how well the game plays. Graphics, textures and shaders affect playability very little. Playability includes such things as collision detection, how well the characters move in space and how they interact with objects in the space.

For example, Fallout 3 majorly failed in one aspect of playability. The rebar, rubble and various blocks and chunks of cement get in the way of movement. It's easy to get 'stuck'. Sometimes getting stuck means you can jump your way out. Some times getting stuck means restarting your game. It is crucial to make sure your play testers ferret out any stuck spots. If possible, don't make terrain where you get stuck anyway. It's always frustrating to walk over terrain to be blocked by some invisible thing that you have to jump over or go around. Make sure to thoroughly play test your games for this aspect of the terrain.

Collision Detection

I won't say that much here about collision detection. In short, just make sure it works properly. Bullets aiming at the player character need to impact and be felt through rumble. The collision needs to make sense. So, test and test to ensure your detection system works 100%. Then playtest it again to make sure your game testers agree.

Easter Eggs

Easter eggs are fun little romps. Whether they are an integral part of the game and story or not, you should always include some. They are fun little diversions that let you step away from the main game and just idle down for a little. It's always good to let the gamer take a break in the environment just to 'play'. By 'play', I mean wander around and look at everything, goof off and generally do nothing. Getting away from the main action for a few minutes lets you regroup for a possibly hard boss battle. Wandering off just to explore the gaming system lets you find easter eggs and these can be as little as a secret message to an in-game award. It's your choice as to what people find, but make it fun.

In-game tutorials

When a game first starts out, many game designers feel that tutorials are the only way to 'help' the gamer become accustomed to the environment. For first-time gamers, this is true. For hard core die-hard gamers, this isn't true. Having to start out a brand new game and spend the better part of an hour wading through pause after pause stopping for pop-up screens and messages is frustrating and annoying. Always give the gamer the option to skip tutorials and go right into playing the game. If the first level is designed as the tutorial, then make sure it's not important to the game and give gamers a way to skip it.

Parts: 1 | 2 | 3 | 4 | 5

PS3 - Heavy Rain

Heavy Rain by QuanticDream / Sony

Parents: This game is rated M for violence and blood.

Heavy Rain tries to be a new genre. Unfortunately, it isn't. The two things that it tries to do to set it apart (controls and storytelling) don't really work to actually set it apart and have been done before. That said, it is probably worth a play through. The first game that focused on interactive storytelling was Shenmue released in 1999 and then Shenmue II released in 2001 Dreamcast and 2002 Xbox. Flashback.. Shenmue was groundbreaking for its time. It had time flow (winter to spring) and led you through the story of a boy, Ryo, caught in circumstances first in Japan and then Hong Kong. The storytelling worked and the game was reasonably extensive. It promised interaction with 'everything' in the world, but really didn't deliver on that promise. Nevertheless, the game is one of my top games still for interactivity and play value. Shenmue was interactive cinema at its best. Unfortunately, Heavy Rain isn't nearly as good.


Flash forward to 2010 and here we have Heavy Rain. Heavy Rain is a detective story plain and simple. There are detectives in the story and the gamer is also a detective. You primarily watch as the story unfolds, but you can occasionally interact with various people as characters. The story, like a film, flips back and forth between several stories at once. You know when the character changes because the face of the current character will appear as the loading screen.

The game switches between 4 different characters that you can control. Ethan Mars (Architect), Madison Paige (Reporter), Norman Jayden (FBI Profiler) and Scott Shelby (Detective). As you control each of these characters, various pieces of the story unfold based on dialog choices and actions you make (or don't make).

The story starts out with Ethan Mars at home about to celebrate a birthday for his son Jason. After a short intro period to get you familiar with the controls and how the game works, it progresses from home to shopping at a mall. Here is where things go horribly wrong. Ethan lets Jason out of his sight while paying for a balloon and Jason leaves the mall after which Ethan chases only for Ethan and Jason to be hit by car. Jason dies from his injuries and Ethan is left in a coma. After 2 years, Ethan revives only to be tormented by guilt.

His second son is then kidnapped after one of his blackouts and is being held by the Origami killer. Here's where the detective story begins. As Ethan, you're following the clues from the Origami killer to find his second son before time runs out (the rain gauge reaches 6 inches). Hence, the title Heavy Rain. This leads to a series of trials for Ethan. As Madison, you help Ethan by patching him up and doing investigating as well. As Shelby, you're a detective following up on independent leads. As Norman, you're an FBI profiler using a high tech ARI sunglasses appliance that lets you see and analyze clues. The disparity between these stories is unclear and is never fully resolved at the end.

Anyway, after the mall (which this sequence cannot be changed), I'm not sure if the story unfolds in different ways based on choices, but it's possible. The way it unfolded for me left a lot of questions unanswered. So, the story writers didn't completely tie up all of the loose ends by the end. Kind of frustrating.

In my story, Ethan Mars kept blacking out and ending up on Carnaby Lane. Yet, later this story event stopped happening and was never explained based on the outcome of my game. So, it is kind of frustrating that this part of the story wasn't revealed or explained why it happened. Perhaps this will be revealed in Heavy Rain 2 (assuming there is a part 2). Also, I was expecting all four characters to interact at some point, but that also never happened (in my game).


To start, Quantic Dream made one of Gamezelot's bad game design faux pas'. That is specifically, messing with a good thing. Instead of camera on one stick and control on the other, they decided it would be best to require you to press the R2 button to actually walk. When you press the left stick, all you do is look around, but you do not move. This controller scheme is bad. Bad bad bad. There is no need to reinvent the controller wheel yet again. Again, don't mess with a good thing. The controller style that already works here is Walking on left or right stick and camera on the opposite from walking. You have complete control over pressure with the analog stick, so walking speed is not an issue. In fact, the part of the game where pressure is an issue, it's actually a lot harder to limit the pressure on the R2 button than it would be using the stick. Quantic Dream, if you're listening, you need to redo the stick layout (or let the gamer at least choose). Immediate 1.5 deduction for 'messing with a good thing'.


The thing that's most frustrating about this game's save system is that there is no 'Restart Checkpoint' option. Yes, this game uses checkpoints to save. Once a checkpoint is saved, you cannot easily and quickly reload a previous checkpoint. Instead, you're forced to quit to the main menu and then restart that way. How about we give a menu option for restarting the checkpoint next time, hmm? Also, checkpoint saves are fine IF you can also save your game separately at various points. With a proper save system, you can go back and make alternative choices from a previous save. No, can't do that in Heavy Rain. In an interactive story such as this one, this is the bare minimum of a save system. Another point deducted for not allowing saves.

Segment Types

The segment types in this game include cinematic (watch only), free roaming (interact with the environment) and action timing sequence (pressing specific buttons using specific sequences with timing events attached). The free roaming is really the best part of this game. It lets you interact with the world, people and objects. Unfortunately, there's so little of it in this game that you really feel cheated. The action timing sequences come out of nowhere and without warning. It starts and you better be ready. This is one of the downfalls of this game. If you're not prepared to start an action timing sequence, too bad. They start without warning whether you are ready or not. Of course, if you catch it in time, you can quit to the main menu and restart your checkpoint.

My least favorite part of this game, and any game really, is the action timing sequence portion. I really despise this style of gaming. Just give me real live action fighting. Don't throw a bunch of random sequenced button presses and require me to do them in order timed or not. Live action gaming is best. Timed button sequences only serve to frustrate and annoy. It means you have to know the controller by muscle memory. Worse, this game took this timed button sequence to a whole new level requiring contortion to not only press the buttons in order, but press them simultaneously in order. So, you may end up pressing X, O and Triangle buttons all together and then L1 and R2 and possibly even L2. But, you have to press and hold them in sequence. This is highly frustrating and annoying. As a game designer, don't do this. This entire gaming sequence completely gets in the way of storytelling. This is yet another Gamezelot faux pas. Perhaps Quantic Dream should read Gamezelot's Game Design from a Gamer's Perspective series of articles.

Controller Motion Sequences

Ok, while I know it's a novelty on the PS3. I really despise it. There is no need to use an accelerometer as part of the game controller system. This idea is long past its time and needs to go away. The novelty has long worn off and now it's just annoying when used. As such, any game the uses the accelerometer should also provide a way to not use it. That means that the gamer should be able to disable that part of the controller system and use a button in replacement of that controller event.

Character Interaction

The characters interact with much of their environment. However, characters can usually only interact with items that are necessary to move the story forward in most cases, as it should be. Of course, the more things you can interact with story-related or not, the more realistic your game becomes. Although there are some things you can do, like juggling, that never make a reappearance as something important later.


The soundtrack audio is typical of a modern thriller. The airy piano segments really set the somber mood of the scenes. The orchestra swells and undulates during action scenes. The music is probably one of the best parts of this game. It really brings out the action properly.


The characters are well done in features, textures and mesh design in most places. There is occasionally some letdown (bandages on Ethan). The lip sync is done reasonably well, but still a bit stilted in places. The character's textures, especially on the loading pages, look great.

However, instead of doing them in-game, Quantic Dream should have pre-rendered them in Vray or another higher resolution system to make them look spectacular. The PS3 does a reasonable job rendering characters, but if you're planning to have a closeup of a face, then do it using pre-rendered high res imagery. Make these graphics knock-your-socks off good. There's no need to sacrifice the quality of this imagery simply because you want to prove how good an in-game render can be. An in-game render will never look as good as something that comes out of Vray or any other professional 3D rendering system. Why sacrifice this part of the game?


This is basically interactive cinema. Although, it feels like Shenmue, it has a lot in common with the Sims and virtual character games. You watch much of the time and play part of the time. The timing action sequences start too often without warning. There are consequences to failing or winning (or even partway winning). So, depending on how the action timing sequence ends and whether you fail or succeed, the story moves on. This is good in that you get to see the events unfold no matter what, but it's frustrating that you can't always easily go back and redo the sequence exactly how you want it to be. If you catch it in time, you can quit to the main menu and reload the checkpoint to start that sequence over. But, this doesn't always work.

Finally, the game is very short. You can play through this game in only a few hours without even trying. The story event continues fail or succeed. If one of your characters dies, the game moves on without them. Overall, for as involved as this game should have been, the story was way too short.

  • Sound: 9/10
  • Graphics: 8.5/10 (good, but not perfect, lighting: excellent)
  • Gameplay: 7/10 (timing action sequences dragged this one down)
  • Story: 7/10 (plot holes left unfilled)
  • Bugginess: 5/10 (glitching and slowdowns)
  • Controls: 6/10 (badly mapped controls, no way to remap)
  • Bang-to-buck: 1/10
  • Play Value: $5 (too short, rent)
  • Overall: 5.6/10 (timing action sequences, poorly mapped controls, no way to turn off accelerometer controls, no player saves, checkpoint saves, way too short)

Sunday, February 14, 2010

Xbox 360 - Mass Effect 2

Mass Effect 2 by EA Games / Bioware

Mass Effect 2 is the sequel to Mass Effect (Gamezelot gave Mass Effect 5/10). Mass Effect 2 picks up where Mass Effect leaves off, story wise. But, the story this time around is questionable at best. Who writes these game stories anyway? I think they need to hire some better writers.


Mass Effect 2 picks up almost immediately where Mass Effect leaves off. Sheperd is commanding the Normandy. However, just about the time all seems great, a large ship cuts the Normandy in half causing the pieces to crash into a planet. We come to find out that the large ship with that huge beam cannon that rips open the Normandy like a can opener is operated by the Collectors. Now starts the Mass Effect 2 story.

So, instead of Sheperd getting to a rescue pod like everyone else, he goes down with the ship and perishes. Or, so we think. Cerberus manages to obtain Sheperd's remains and manages to rebuild him into a whole human again. That's where your character, Sheperd, wakes up. So, here begins the story again. This time, however, because Cerberus brings back Sheperd, he takes allegiance to Cerberus (sort of) and the fact that Cerberus builds him the Normandy 2. Cerberus even manages to get Joker and the doctor back, but not the rest of your 'party'.

Of course, the whole point is that that Illusive man wants Sheperd to take out the collectors. Why exactly Sheperd was reassembled, we're not sure yet.

Importing Character from ME

This was a pointless exercise. The only thing this does is bring in your character's likeness. It doesn't bring the experience or anything else. It's all a cosmetic thing without any substance, so why bother?


Basically, there are three main game play zones. Combat, free roaming and mining. Combat zones get you through the objectives. Free roaming (most on the ship) lets you manage your crew, objectives and research. You can also free roam the Milky Way to find your next objective locations. By using probes, you mine the surface of planets to obtain elements. Elements are used in research upgrades. Upgrades give your armor and weapons better abilities.

While some of the new aspects work, some don't. The developers improved parts of the game, but overall it really is no better than Mass Effect for play value. For story line, it's far worse.


Well, what's to be said. You can outfit your character with armor, for what it's worth. The reality is, of course, that it's all for show. Also, the only armor that's upgradeable is the armor that comes with the game. Any imported armor from the downloadable content (DLC) area is locked. By locked, I mean you can't do anything to alter this armor. It is what it is, like it or not. You can't add upgrades or in any way strengthen DLC armor.


The controller is pretty standard. Movement on left stick, camera on right. Various buttons for action (firing, changing modes, etc). You'll pick it up pretty fast because it's fairly standard.

Controlling Sheperd, on the other hand, can be rather tedious. There are lots of bugs in the control system. For example, there are times where Sheperd will move in random haphazard ways even when you're holding the stick firm in a given position. Next, when you're in a cover position or when entering a cover position, I've had several times when he'll enter cover and then immediately jump over the cover into a clear position to the enemy. Frustrating!

Death Halo

Mass Effect is another in this growing series of games that prefers the death halo to putting an actual health bar on the screen. No No NO!. No death halos! Whoever dreamed up with this idea needs to be slapped and slapped hard! Get RID of the death halo. The reason it doesn't work is that 1) it covers the entire screen so you can't see what you're doing, 2) the controller becomes less responsive so you cannot move out of the way of fire (let alone see what's even firing at you) and 3) it's just plain annoying. The first two points are enough to prove that this idea is stupid and doesn't work.


This is one in a handful of recent games were the camera has, yet again, become an issue. When you're able to shoot at a distance, this mode works great. However, when you have enemies that get right up next to you, the camera fails horribly. First, the enemies that like close contact get so close to you they're touching you. So, there's no way to position the camera so you can even see them. Second, because they are so close, you can't even use a gun. You have to use a melee tactic. The trouble is, the zombie creatures (that love to do this) get right on top of you and congregate. When these enemies do this, you cannot even move. The zombies get right in your way and prevent character movement even when they aren't hitting you.

Between the poor camera system AND these stupid close contact enemies, this game completely fails with the camera. At least let me move the camera to a more distant position so I can see what's going on around me. For this combined problem alone, the game loses 1.5 points from the overall score right off the top. Camera problems should have long been resolved in games at this point. Going back to having camera problems means the designers were not thinking properly about the game design.


The missions are way too short and very repetitive. Different characters, same results. Move your character from point A to point B and do something at the end. Between points A and B, you fight a lot of enemies from behind cover. It's all mindless shooting and nothing really to think about. No puzzles, no thought provoking ideas, nothing. Just a lot of shooting. I was very disappointed with this part of the game. At least add some puzzle levels to the game. At least make the gamer use some kind of thought process more than pulling the trigger. Alas, it didn't happen. For an RPG style game, it's really pretty one-tracked.

Cinematics and scene skipping

With some cinematics, you can skip them. With others, you can't. There is no rhyme or reason why some are skippable and others aren't. On top of this fact, once you're in certain parts of the opening of a mission, you cannot pause or start the level over. You must wait through perhaps 5 or more minutes before you can get to a point where you can either save or start over. These points where the game is locked out of the save/load screen is very frustrating. It's especially frustrating when the game has automatically loaded an autosave and you want to actually load a different saved game. Yet another 1 point is immediately knocked off for this problem.


Save early and save often (when you actually can). There are times where the save screen is not available (see above)... and it's usually about the time when you want to save. Also, if there is even one enemy present on the playfield, you cannot save your game. I've had several glitches that have prevented me from saving my game because the game thought there was an enemy on the play field.


Yet another problem is with the loading after Sheperd dies. There is A for resume and X for load. If you press A for resume, it is random what game it will load... meaning, it could load the 'Restart Mission' version or it could load your last save. Because of this roll-of-the-dice loading technique, you may end up having to load the game twice. Loading the game twice (once as a mistake, once intentional) can sometimes take as long as 5 minutes. Again, another 1 point right off the top (so far, that's 3.5 points right off of the score).


The game could have been loads better and was slightly better than Mass Effect. They tried, I'll give them that. But, EA must have held Bioware back. This may be the last Mass Effect game I buy unless I know that they've put much more effort into the next game. The gameplay leaves a lot to be desired, it's very repetitive. Worse, though, are the glitchy bugs that make Sheperd uncontrollable at critical times. Even worse, in many cases the game intentionally drops enemies right behind you. I mean, immediately right behind you. So, you're forced to run away for cover. Totally unacceptable considering the camera problems. If, as a developer, you're going to play these combat games, than at least provide the gamer with a HUD scanner to see where your enemies actually are.

Oh, and considering the time era that this is supposed to be set in, it's absolutely ludicrous that Sheperd's suit doesn't contain a HUD scanner.

  • Sound: 8.5/10
  • Graphics: 9/10 (very good, not perfect)
  • Gameplay: 7/10 (fair)
  • Story: 7/10
  • Bugginess: 5/10 (lots of glitching)
  • Controls: 5/10 (could be lots better)
  • Bang-to-buck: 2/10
  • Play Value: $10 (rent or buy)
  • Overall: 6/10 (slightly better than Mass Effect in limited ways)

Sunday, February 7, 2010

Game design from a gamer's perspective Pt. 3

Loading Screens Continued

You should always keep the gamer in mind when you design your game. Always! You should never program the game for your convenience. Always program for your target audience. So, when you design your loading screens, you need to consider how long it takes to get data into memory. If it takes longer than 2 minutes, it's taking too long. If you don't have a choice in this matter and it is what it is, then design a mini-game or alternative screen that the gamer can play with during the loading period. Alternatively, use of cinematics during this time is acceptable. Do use cinematics as ways of moving the game forward combined with loading the world into memory. Don't wait through loading screens only to watch a long cinematic and then get into the game.

Health pickups

If you design a game where the character will lose health, then always design a system to obtain health on the level. Batman Arkham Asylum is yet another shining (bad) example of how not to do this. Batman's designers chose to obtain health pickups by subduing other creatures. So, as long as there was something to subdue, you could get health. The problem with this 'feature' is that there were never enough enemies to subdue (Batman doesn't kill in the game). Worse, when you did subdue enemies, they gave you a pittance of health back. Meaning, you never could fully ever replenish your health bar. Don't do this. If the character needs health pickups, put enough health pickups around the level and make them blatantly obvious what and where they are. If that means attaching a medkit to the wall, do it. Also, health pickups should come in at least two sizes, small and large. Small increases about 1/4 of your health. Large sized should fully or almost fully replenish your health. I despise playing games where my health meter is always 1/4 full because I am unable to find health pickups.


As we move more and more into cinematic games, it's now becoming increasingly important to weave gameplay together with story. Unfortunately, I've yet to find a single game that has successfully achieved this goal. The closest game was probably Oblivion. Oblivion's method was to weave each quest into its own substory. There was the 'main quest'. This quest basically starts and finishes the game. You can do this quest at any time, but because of the way it changes Cyrodil's landscape, it is best to wait until much later after completing many other smaller quests to tackle the main quest. Even still, there were smaller quests that needed to be completed before you could complete the main quest. Thus, this tied in at least exploration and completion of other goals into finishing the main story line. That unfolded other stories that helped solidify the main quest's story and ensure the gamer a much more immersive gaming experience.

A story is a story is a story. It doesn't matter whether you read a book, watch a movie or play a video game. The story is important to the entire affair. If the story is well written, has twists, feels complete and has a satisfying ending, then that's all you can ask. Games with weak stories or without stories at all are limp and lifeless like Dead Rising. Mindless games without a compelling story verge rapidly on boredom. A solid well thought out story not only keeps the gamer interested, it gives enough subtext for the gamer to become immersed in the world. That's exactly what a good story should do. Once you've gotten the person immersed in the story, they will continue to play for a long time. Games with great story lines include Oblivion, Halo 3, Halo 3 ODST and Fallout 3. A good game has great gaming elements. A great game combines the gaming with a solid story. A perfect game devises a way to marry the story and gameplay seamlessly (we've not gotten here yet).

Granted, not all game styles need a story. For example, sports games and racing games don't really need them. But, it helps if there is some kind of story involved. For example, Gran Turismo is pretty much a straight racing game. The goal is to race your way through each event and improve your car. That's pretty much one tracked. Now, if you took that gaming element and wrapped it around a person and events in their life, that would make Gran Turismo take on a whole new dimension in gameplay. You could even open up free roaming aspects to let the person walk around the cities and find things to enhance the car or their driving skills. Just think of a combination of Grand Theft Auto and Gran Turismo together in one game!


Enemy and Boss Tactics

Perhaps this should have been under the bosses section, but I also feel it needs to be under game play. When designing how a boss works, do not use an unkillable/undamageable boss simply to whittle down player health for later smaller enemy attacks. This is frustrating and annoying. If you're going to let the gamer battle the boss, then let them battle it. Don't do some intro game play where it's impossible to kill (or even wound) the boss. If the boss can't take any damage during a specific area of play, then it should not be there. Only enemies on the play field that can actually be killed or damaged should be actively engaging the player. If something can't be killed or damaged, then it should not be there... or at least, it should not be launching damage salvos at the player character. This is similar to Perfect Aim. Don't do it. Only killable or damageable enemies should be actively engaging the player character.

Here's an issue where changing controller mappings sometimes works, but most times doesn't. As a game designer, be mindful of your intended gamer audience. Between the PS3 and the Xbox 360, the controllers are similar enough that you can map similar styles to each of the buttons. For example, the analog sticks usually are reserved for camera on one and movement on the other with the buttons mapped to attack, jump, climb, etc. This is the perfect combination. Because so many games have used this style of mapping, don't mess with a good thing. If you want to design a 3D shooter or RPG, use this style. Don't think that because you decided to put the camera movement on the bumper buttons that that will make your game better. Don't do it. Stick with what works and is accepted.

If you decide to modify the usual and standard controller layout to something odd, then at least give the gamer the ability to change this mapping to a controller style of other familiar games. Don't lock the gamer to a non-standard button mapping system and force them to play the game that way. Having to relearn the controller layout for your game is just bad design and, worse, may even doom your title into obscurity and lackluster sales.

Also, don't lock gamers into using the Xbox Kinect, the PS3 Move, PS3 Sixaxis or the Wiimote motion controllers for your games. Give the gamer the option of using a standard controller to play. Simply because you, as the designer, feel the game may be better experienced by using these motion control systems may alienate an entire paying audience who can't (or won't) play games using these motion control systems. Spend the time to add a standard controller layout to your game along side these novelty controllers.


I know a lot of people are familiar with the use of fixed cameras through franchises like Resident Evil and early Tomb Raider titles. As much as Resident Evil liked this style of camera, when designing your title, don't do it. The trouble with Resident Evil becomes quite apparent very rapidly once game play starts: Enemies can hide out of the camera view and damage the player. The player can't even see the enemy to kill it. This, once again, goes back to Perfect Aim (and Perfect Vision). In game, the enemies aren't reliant on the camera and can 'see' exactly where the player is. The gamer, on the other hand, is limited by what's shown on screen by the camera. Fixed cameras prevent the player from seeing critical parts of the game playfield. Don't use fixed cameras. Always use a floating camera and let the player move the camera to wherever they need it to be. The camera is the only way to view into this game world. Don't cripple the player by imposing stupid restrictions on the camera. The camera should never be used as a challenge element. It should only be thought of as a way to peer into the world. If the player can't see the world in the way they need to see it, they will get frustrated and stop playing the game.

Recently, I've found few games that cripple the camera. However, every once in a while a game will come along and try to do something creative with the camera. In fact, by crippling the camera the only thing the developers have done with their game is doom it to obscurity and failure.

Audio and Soundtracks

Music is an important element in any game. It's what sets the mood and tone of a given level. It can swell to indicate enemies approaching. It can diminish to indicate the battle is done. There are lots of ways to use music and sound effects in creative astonishing ways. The trouble with some soundtracks, though, is monotony. Be careful to use the right music at the right time. Don't use hip-hop music when doing a medieval genre game (not that I've seen anyone do this, yet). That's an extreme example, but be aware of what you need. I prefer orchestral music in games as it's soothing at necessary times and intense at others. Heavy metal can be used in some instances successfully, but it can be out of place at other times. So, be cautious of using heavy metal in a game you design. The same can be said of techno, so be careful with this genre. The style and type of music, though, can definitely help or hurt your storyline. It can make the difference between an emotional scene or making it stupid.

Coming Up:
  • Characters
  • Computer Graphics: 3D models, textures, lighting
  • Playability
  • Easter Eggs
  • In-game Tutorials
Parts: 1 | 2 | 3 | 4 | 5

Saturday, February 6, 2010

Game design from a gamer's perspective Pt. 2

Time Wasting Continued

When designing a game, you're already asking the game player to forfeit part of their lives on your game. But, as long as the gamer is having fun, they don't see that as a loss. However, what is a loss are all of the time wasting moments that are not in-game. For example, loading screens, cinematics, transition scenes, etc. These are all non-playing aspects of the game. Anything that takes the gamer away from the action and forces them into a 'watching' or 'waiting' mode is dead-time. This is true time wasting at its best. To avoid these dead times, always allow the gamer to do something. If that means designing a mini-game or putting up an experience management page, allow the gamer to do something during dead time.

Dead time is what can also make or break a game. Too much dead time and the game becomes boring. Gamers will move on to more action oriented games and leave yours behind. If you want to keep the gamer excited, be acutely aware of dead time. A good example of how to prevent dead time is Assassin's Creed II. During loading screens, you can at least play with Desmond and make him run and so forth. This gives the gamer at least something to do during these dead times.

Also, do not waste a game player's time by making them do pointless activities. Think heavily about each part and level of your game. Consider if it moves the game forward or if it's exciting for the gamer. If it neither moves the story forward nor is exciting for the gamer, remove it from the title or spice it up and make it worthwhile. A few good examples of this are chests, cupboards or containers that contain nothing. If you're going to create a container, put something in it. Make it worth the gamer's while to search for things. If you're planning on creating a level, put something on the level to make it worth being there (a boss, treasure, something.. anything). The prime example of a level being done badly is in Darksiders. The Vulgrim Tunnels are pointless. You walk down a path from start to finish with nothing in that void (literally). Unless they're trying to give your character exercise or test your walking skills, there's nothing here to do. Don't do this in your game.

For loading screens, though, it is suggested to load a mini-game, a shop or an experience screen and let the user manage their character until the loading of the next level is complete. Even better, use incremental loading during the game and never have a loading screen at all. For dungeons or other rooms, begin to prefetch these in the game when the gamer gets close enough to the room. So, if the gamer opens the door and enters, the room is nearly already loaded. This can cut loading times in half or more.


There is plenty to be said about bosses. Let's just cut to the chase, though. Bosses can be done right or done wrong. Most times, however, the reason a boss level is done wrong is not because of the boss itself. Many times it's done wrong because the game designer didn't work through the proper method of how it should work.

In Batman Arkham Asylum, for example, the bosses in this game weren't the issue. They were hard, yes. But, the issue is that the game relied on trial and error play to win the level. This means that your character must die and start over many times in order to find a strategy to defeat the boss. This goes back to time wasting. Making the gamer restart the boss battle over and over and over is a utter time waster. Don't do this. The problem with Batman, though, goes back to the fact that Batman's health drops rapidly. Far too rapidly for the amount of armor he is supposed to be wearing. Secondarily, the game throws 12-15 enemies at Batman all at once. Then there are other issues, like certain moves the enemies can do that take away nearly 1/4 to 1/2 of Batman's health... merely by kicking or punching him.

So, in Batman, the way to win is let the enemies take out themselves and simply run around and avoid being hit. Yes, the enemies will attack each other. So, let them. The trouble, of course, is that you have to get them in a position to do so. They won't do it on their own. Batman must lure them.

One thing about enemies and bosses that always drives me nuts is perfect aim. In a digital world, we are limited by our view into this world through pixels. Because pixels are a fixed dimension, as objects get smaller they also get harder to read. But, the game world itself has clear vision to infinity. So, while the gamer is crippled by the pixel screen, the in-game enemies have perfect vision. This is an unfair advantage. So, for example, enemies can always seem to find, aim and target your player character perfectly. On the other hand, finding and aiming at enemies can be a challenge due to pixel size limitations.

When designing a game, please keep this in mind and compensate for the 'perfect vision' by at least making the enemies miss or aim incorrectly at least some of the time. Perfect aim gets old and tired in games. Don't do it. Do make the enemies miss more often the farther away they are from the player character. Basically, whatever rules you impose on the player to negatively affect aim should be applied to enemies and bosses. Even more than this is that extra care should be taken to ensure there is no such thing as 'perfect aim' in your game.

Enemies always targeting the Player Character

An offshoot of the Perfect Vision problem is game designers who design games so that the enemy AI always targets the player's character. Even if you have a team helping you, inevitably most games have their bosses and enemies target the player's character. This is stupid, unfair and unrealistic. If you have a team, the enemies should equally target each of the team. The player's character should have no more weight in being attacked than any other player. The trigger to whether a specific character should be attacked is if the character enters the enemy's block of influence. This means, as soon as any character gets in range, the enemy should 'see' and begin targeting that character. A perfect example of this being done wrong is in by Bioware in Mass Effect 2 and Dragon Age Origins. In these games, once an enemy locks onto your character, it stays locked until you kill it or run away.

Character Deaths

Inevitably, when playing bosses, your character may die. How your game design team chooses to handle a character death is as important as any other aspect of the game. In some cases, it's even more important. Some games (Too Human) like to produce long drawn out unskippable death sequences. Don't do this. Instead, if you must show some kind of cinematic, make it quick and let the gamer skip it. If you are going to show a cinematic, use that time to reload the level. Don't wait until the cinematic is complete before reloading the level.

As far as level reloads, I never get this part of gaming. You already have the level loaded, yet you have to reload it again from scratch? This can take 1-2 minutes. Why do this? Instead, use the existing loaded level and simply reset all of the objects on the level. That's got to be faster than dumping the level and reloading it from disk. After all, you were already playing the level and it's already in memory. It may be slightly more complex to code this, but shaving the time off of the gamer's dead time gives the gamer much more in-game time. This is what you want. So, it IS worth the effort to code it. Smart coding leads to higher quality gaming.

The next issue with character deaths is using this as part of the gaming experience. The player characters should never die. Never. The goal is to keep the gamer in the game experience as much as possible. Making the gamer wade through death sequences time and time again goes back to time wasting. Don't do it. Do everything you can to avoid the character dying. If the character must die, then the reload to start the level should be instantaneous. No sequences, no specialty things, just fade out and fade back in. Let the gamer start immediately. We'll come to why this is important for the story shortly.

Loading Times

When a game needs to load a level or reload after a character death, it should be as fast as possible. There is no need for 3 minute loading screens. This is excessive and wasteful of the gamer's time. Granted, sometimes it can't be helped, but smart incremental loading can be designed. Oblivion and Fallout 3 are perfect examples of incremental loading of the environment. With incremental loading, the game never has a loading screen and instead is constantly fetching in-game data to build the next part of the environment. Waiting through loading screens is not mindful of the value of the gamer's time and suggests convenience on the part of the programmer and designer. Don't do this.

Coming Up:
  • Health Pickups
  • Story
  • Character Control
  • Controller Mapping
  • Audio & Music
Parts: 1 | 2 | 3 | 4 | 5

Friday, February 5, 2010

Game design from a gamer's perspective

This is a multi-part series on successful game design from a gamer's perspective. This article series will encompass such topics as story, models and texturing, lighting, artwork, gameplay, genre choice, audio and many other subtle aspects of creating 3D based video games. This article will also discuss what game techniques work and what to avoid. So, let's get started.

Successful games

Let's get right into the meat of this and discuss what makes a successful game. Clearly, success is nearly always measured by dollars. Specifically, how many units sold and how much profit was made. Even more, did the game pay off its debts that were generated during creation and did the revenue rise above those expenses to actually make money? For executives of gaming companies, this is the goal of a video game. But, was the game actually successful for the gamer? For the C-level executives, I'm sure they'd respond a resounding yes (assuming it exceeded its dollar goals). From their point of view, apparently enough gamers purchased the game to make the dollars at least work. But, was the game a success for the gamer? That's a completely different question.

From the gamer's perspective, the game may not actually work. An overhyped game from a brand resting on past successful history can produce games that appear successful, but only because a gamer was 'tricked' into the purchase. So, success of a game is measured both in dollars and in how the game was received by the gamer. The adage is still quite relevant, "Once bitten, twice shy". If you burn the gamer with a bad title, you likely won't get much respect from future titles. So, don't burn the people who keep you in business by producing bad titles.

So, that means you should measure success in two ways. First, money. Second, longevity. Money describes how well the gamers decided to adopt the game immediately. Longevity describes how long the gamer played the game before giving up or trading it in. If there are massive trade-ins within a few days, the game failed as a game. That's when the developers need to understand why it failed.

No longer can developers sit in a bubble and develop games without listening to gamer comments. With social networks like Twitter and blogging, it is more important than ever for developers to review forums, read critical reviews, listen to complaints and understand just what problems gamers have found in a game. These are issues that must be addressed, preferably in patch updates to the existing game if possible. If not, then these issues definitely need be addressed in any sequel games.

So, while sheer numbers may describe immediate monetary success, this does not tell the whole story. Executives who are simply bean counters fail to see the bigger picture. Gamers are finicky and will choose with their wallets. Once bitten, twice shy applies to game titles. More than this, it also applies to game development company loyalty more and more frequently. As a gamer, I know what companies to avoid. For example, I simply will not purchase any more Square Enix titles. I've been burned too many times from this company. I've about had it with EA as well. The quality of EA titles is so widely varied that it's too much of a risk. So, before I buy any EA titles, I must read reviews and play demos, if possible. I also feel this same way with Activision's and Atari's hit and miss strategy.

Game development: Ideas that work and techniques that don't

As a game developer, it's important to solidify the game style and format up front. All too many times, the gaming engine that is chosen dictates the game's play style. The choice, for example, of using the Havok engine may have serious consequences on the success or failure of the final game. So, choose your engine wisely based on game genre and understand its downsides carefully. For example, licensing the Havok engine for a full RPG is probably not a good idea. At least, it's not a good idea without some recoding effort.

Health Status Indicators

As an example of what doesn't work, there are some licensed engines that don't offer a health meter. Instead, the engine opts for a blood or out-of-focus halo around the screen. As health diminishes, the halo increases. For a gamer, this aspect can make the game frustrating and unplayable. Halos obscure the play view, so you can't see what you're doing (see Perfect Aim / Perfect Vision). Worse, some games inhibit the character's ability to play after a certain point. So, you can't move or respond correctly to enemies. This just leaves the gamer to stop right there and reload from a saved game. There's no point in continuing to play when you can't even control the character properly. Mass Effect 2 is the perfect example of a blood halo done poorly. In 2010, if you can't provide a health meter on the screen, don't bother creating the game. There is no reason not to provide a health status indicator (more than a blood halo). All too many times, especially in health screen halo engines, once you see the blood halo, your character has 1-3 hits left before completely dying. Worse, you can't properly see the screen to maneuver your character out of the way.

On the other hand, having an actual meter on the screen so you can see how many hit points you can take before dying is much more useful. It helps the gamer decide how strong a given enemy is by how much damage they deal. Having this information allows the gamer to create a strategy to beat that enemy and know their relative weapon strength.

Save Game Locations

There are many styles of game saves. These include checkpoint saves, save anywhere, pause save screen options and in-game save points (obelisk saves). Game developers need to understand that saving the game is not and should not be part of the game play. Don't weave in saves as part of the story or challenges. Saves are there for convenience to the gamer. They are there to allow the gamer to save progress and also allow the gamer to stop the game at selected times and/or prevent losing the work up to that point. Therefore, game save points should never be treated as some kind of obstacle, challenge or in-world treasure. Never.

It is preferable if game saves be allowed anywhere in the game. This style is the most efficient for the gamer and allows the gamer to prevent starting over time and time again. The game save style to completely avoid is the one that forces you to play through an incredibly long, hard and complex level with lots of chances for death before you reach an in-game save point. This style of gaming is frustrating and extremely bad design. It ensures the gamer will give up before they finish the game. Don't do this.

Checkpoint saves can be useful as long as there are enough checkpoints. Again, this goes back to in-game saving. If your team has decided that checkpoint saves will be the only mechanism for saving progress, then your team better make sure there are enough of them along the way. Otherwise, your game will end up in the same boat as the immediate example above. Keep in mind the pitfalls of using checkpoint saves, though. Checkpoint saves overwrite the previous save. So, the gamer can only start at the most recent checkpoint. This means they cannot step back two or three checkpoints and redo those sections of the game. If your game is the type where you can make choices that affect the outcome of the story, then checkpoint saves are not appropriate for this gaming style. Checkpoint saves are intended for mindless zombie killing. They are used where outcome of the game is irrelevant. For RPGs where choices can be made that affect outcome, do not use checkpoint saves. Instead, RPGs should always use save-anywhere saves. This allows the gamer to save before critical choices or battles and restart the battle from seconds before.

It's fine to combine save styles, though. If you allow save-anywhere saves and want to also create checkpoint or restart mission saves along side, that's fine. Again, though, beware of pitfalls. If the character dies on the level, let the gamer choose which save point to restore from. Do not automatically start loading from the checkpoint immediately after a character death. This is especially true if there is a newer save-anywhere save present. This is waste of the gamer's time as he/she will need to wait through that load sequence only to reload again from their own save. Mass Effect 2 is a prime example of this behavior. Do not do this!

Other saving issues include how much data is saved to the save file. For example, the best games store everything including character position in game, character level, inventory items, etc. With this save type, your character starts exactly where you left off. This is the best style of save format there is and the least disruptive to the gamer. Always choose this style of save when designing. Other save game file formats include starting over at the checkpoint. So, this file format stores only the start point and nothing else. In this case, the gamer starts the level over with a clean slate (no previous weapons, armor or whatever). This is frustrating because all of the stuff you'd found to that point is lost. Don't do this... especially on an RPG. Don't force the gamer to backtrack in a game to get goodies a second, third or fourth time. For shooters, it may be acceptable, but even here I wouldn't recommend it.

Time Savings

As part of the game, don't waste the gamer's time on irrelevant or unnecessary things. These things include long loading screens, long death sequences or unskippable cinematic sequences. In this goal, the pause button on the controller should work 100% of the time in game. Granted, certain times you can't, like intro loading screens, game saving sequences and other operations that can't easily be interrupted. But, as long as it's in game, pause and load panels should be available 100% of the time. Again, Mass Effect 2 is the prime example of not doing this correctly.

Coming up in part 2:
  • Bosses (when is too much or not enough)
  • Character Deaths
  • Loading screens
  • Loading times
Parts: 1 | 2 | 3 | 4 | 5