Archive for the ‘Other Games’ Category

Anti-TD Screenshot
A tower defense game where you’re controlling the attackers, instead of the towers? You can even pick which entrance to launch your troops from and direct them in mid-level. Too bad the shortest distance from each possible entrance to the exit crosses the exact same number and types of towers, making this a nearly meaningless choice…

This is a bit of a long post, since I’ll be looking at a game that I like the concept of, but that I think didn’t meet its potential. Once breaking the game down and identifying what I see as the problems with it, I’ll start to explore strategies that I would look at to improve the design.

Tower defense games are a genre of (usually flash) strategy games, with origins in RTS map mods, in which the player places a series of usually upgradable fixed turret defenses in the path of oncoming waves of increasingly powerful enemies. The goal is to defeat all of the enemies in each wave, and place and upgrade more powerful towers to overcome more powerful waves.

The concept is simple but it has a lot of traction, and in its simplicity, and the ease with which clones can be created, it has exposed many variants that make it interesting to me from a design standpoint. The genre comes in two basic forms, one track-based, and one in which you can build your own maze. The successful Bloons Tower Defense and Desktop Tower Defense series are examples from each that demonstrate a high level of polish and careful design. The track-based ones are super easy to make as games go, and are a dime a dozen on the internet. There’s even a series where someone made a “make your own tower defense game” game, where you lay down the track and set values for the towers and so on.

In this post I’m going to look at one of the most unusual games in this genre, Anti-TD. Anti-TD throws the genre on its head: There is a track based tower defense game, but you’re not the player placing the towers around the track. You’re the computer sending the troops out. To win each level, you need to exhaust the player’s lives. I like this concept (I actually thought about doing something similar in the past), but in practice, it’s pretty dry mechanically. Go ahead and try it out yourself.

On face value, this isn’t surprising. Sending a bunch of units you don’t really control to get slaughtered in a gauntlet isn’t very stimulating. But I believe the problem runs a bit deeper than that — the underlying tower defense game just isn’t well designed, and it doesn’t present strategic options to the player. Fortunately, this is solvable.

Your units are all modeled using a speed and a resistance value. I assume resistance is hit points. Typically, in a tower defense game, the relationship between these variables matters. The high speed, low hit points enemies dart past your slower firing towers, shrug off splash damage (as they’re usually a bit spaced out) and reduce the DPS of any towers that do more damage in one shot than they have health, while the low speed, high hit points enemies don’t break a sweat at towers designed to slow you down, and test your ability to focus large amounts of raw damage onto the field.

In Anti-TD, you have units like these, but they don’t matter. The AI builds towers seemingly at random, distributing equally among its tower types and showing no particular strengths or weaknesses. As a result, all strategies tend toward the same result. The towers themselves are missing some key points too. First, they almost all have a high firing rate, so the unique advantage of fast units is essentially eliminated. Second, there are no splash damage attacks, so there is no reason to spread your units out. Third, they have no slow attacks, so the unique advantage of heavy units is essentially eliminated. Instead, the unit selection comes down to this: Which unit has the highest speed multiplied by resistance, and how tightly can I pack them?

Of course there’s cost too, you have to spend money for these units. Gaining money comes from the units traversing the track: Each tile of progress for each unit gives you a fixed income based on the map. This, however, means that speed multiplied by resistance is not just a measure of how many towers the unit can pass before dying, it’s also a measure of how much income that unit will give you before dying. Dividing this factor — I call it survivability — by the price of the unit gives you a measure of how lucrative it is to deploy that unit.

I’ve performed this analysis on every unit in the game. The result is two dominant units.

The first is the TRT-L unit, which is extremely slow but has disproportionately strong armor. This unit is the most cost efficient unit in the game, making it the most lucrative to deploy. Spamming TRT-L units (specifically the lowest level of them, which are the absolute most efficient) to start a difficult will earn you more money than any other unit combination in the game.

The second dominant unit is the Razzo, specifically the highest level Razzo. Not only is it very efficient itself, but it has the highest raw survivability of any unit. It’s not as tough as the TRT-L, but it’s a lot faster, and that means the towers can’t inflict their DPS for very long. Razzos are expensive, and all units are more effective if massed so that towers can’t apply their full DPS to each incoming unit, so it’s best to start with TRT-Ls until you have enough money to send a swarm of Razzos. Typically the Razzos will raise money themselves once they get going, though, especially as they’re likely to complete the map and get a substantial completion bonus for their trouble.

This is a mid- to late-game strategy, since you can’t afford to deploy these units early on, and the income for each step of progress in early levels is too low to justify sending expensive units out, since it’s impossible to recoup the cost, even if they get all the way through. In the super late game, this strategy is then complemented by aerial units. You usually don’t have to think much about aerial units; any tower can hit them just the same, so they’re subject to the same calculation, and they’re inferior. Later on however, they can slip past your slower ground units by bypassing the distracted heavy surface towers. This increases their efficiency, but they also are subject to additional firepower. In other words, the most efficient flier complementing the most efficient ground units will maximize your ability to reach the goal on extremely difficult maps.

I think the game really just needs an injection of interesting choices and sense of achievement. First thing I would do would be to strip out some of the more meaningless choices: Instead of having a large table of units that are, for the most part, just points on a one dimensional usefulness spectrum, I would suggest a handful of behavioral archetypes. Your classic medium strength and speed unit, a sprinter, a heavy, a class or two of fliers, some spawner/carrier units, maybe some that have special defenses like shields that deflect certain attacks better than others or buff others. Start the player out with only a couple, but as the player progresses in the game, let them allocate upgrade points toward unlocking new types. Instead of offering differently priced strengths of these units, offer just one, but let the player algorithmically scale their global power level between waves.

Then, I would focus on giving the game more of a strategic element. Path choice in Anti-TD comes down to picking the path that crosses over cool buffs, a feature I very uncharitably suspect was added after the initial prototypes of the concept didn’t deliver on being fun enough. Instead, I would let the AI build unbalanced tower layouts. If the player favors a certain path, you could let the AI drop more towers on that path, but another option is to simply have the AI’s towers upgrade when they reach a certain amount of experience. I’d also differentiate the towers significantly, and allow certain tower choices to be better counters for certain unit choices. Have a tower that excels at fighting fast units, a tower that punches through heavily armored units, and so on. Combined with experience-based upgrades and a reactive building algorithm, this would provide a negative feedback mechanic on the most effective strategies. It might even be worth encouraging the AI to be biased in favor of unbalanced tower type distributions too — after all, one of the weaknesses of Anti-TD is that the player can’t exploit weaknesses in the tower pattern because there aren’t any.

Finally, there are a few miscellaneous things that I think would really give some extra punch to the game and the sense of playing as the enemy in a tower defense game. First, I’d give the AI player a cursor and GUI, and let you watch the computer actively placing (and perhaps upgrading) towers in the path of the player’s units, responding to each wave by either sitting still and letting it come, or desperately trying to toughen the defenses. Second, I would force the player to prepare and release enemies in waves, with feedback on the outcome for the player and the AI after each wave. Every ten waves or so, the player could prepare a special boss wave. During waves, the player’s unit deployment GUI could be hidden entirely, instead using the AI’s tower building GUI. Between waves, the game could pause and the tower building pane overwritten with a darker, more evil unit deployment GUI, for preparing the next wave. Third, I would display direct feedback on things the tower-building computer player is thinking, from the more direct exposure of AI reasoning in “X is worried about fast units” to a more flavorful exposure of confidence in “X is stepping away to get a soda”.

As much as I knock Anti-TD, I think it’s an interesting concept and a fun one to explore. As games like Dungeon Keeper have long before shown, the fantasy of swapping roles and playing as the enemy is one that resonates with people. The difficulty is that these games have the potential to only be fun in one direction. I think half the real challenge is just to find the fun in level design, and expose that to the player. Of course, the designer is really on the player’s side when making levels, so a bit of creativity is needed in the adaptation, but in most cases I believe it can be done. For Anti-TD, it’s a first attempt. I see it like a prototype; it’s fun to look at how the concept can be expanded and changed to hopefully design a more fun game.

Read Full Post »

This is a repost from my personal journal, dated April 1, 2007.

Yesterday I was dirty
Wanted to be pretty
I know now that I’m forever dirt

We are the nobodies
We wanna be somebodies
When we’re dead,
They’ll know just who we are

Some children died the other day
We fed machines and then we prayed
Puked up and down in morbid faith
You should have seen the ratings that day

Yesterday, we had an ethics debate in class about Super Columbine Massacre RPG. Some people were familiar with it, others had never heard of it, and were horrified to know it existed. I was aware of its existence, but not terribly familiar with it. The game places you in the shoes of Eric Harris and Dylan Klebold, on the morning of April 20, 1999. Initially published anonymously, one survivor was so infuriated that he hunted down the identity of the creator, and outed them. When he publicly posted the identity of the man who developed this RPGMaker game, he said, “One of the girls who died was a friend of mine. Rachel. We were in the same church group. Anyone playing this game can kill Rachel over and over again.” As an educational exercise, I took it upon myself to download and beat the entire game today.

The game is more like an interactive film than a proper game. Gameplay is simple; most of the game involves walking through a school building, seeing students and teachers running wildly, approaching them, and then selecting the weapons to murder them with in a Final Fantasy-style RPG battle. If the innocent people fight at all, they do minuscule damage.

The game’s appeal and interest does not come from the action. Instead, like most eastern-style RPGs, you’re playing from cutscene to cutscene in a sort of interactive movie drawn out by repetitive battles. In the introductory sequences, you control “the boys” as they place propane bombs in duffel bags and film a last video of themselves the morning of the massacre, then sneak the bombs through the school and set them up under two tables in the cafeteria, before rigging your car with a third bomb, and hiding on a hill to watch and wait.

After none of the bombs explode, you make your way through the parking lot, engaging in inescapable battles to the death with every high school student you touch, all of them helpless before your array of deadly weaponry. You’re free to enter classrooms or pass them by, killing janitors, teachers, and all brands of boys and girls, such as the “Jock Type”, the “Sheltered Girl”, the “Church Boy”, and the “Openly Gay Man”. Cutscenes with dialog and/or flashbacks appear at critical moments, such as the first time you enter the cafeteria, seeing the controls that one of the boys used to run the lighting in a school play, etc.. After each major milestone, the music that you murder people by changes.

You get additional special abilities in combat after achieving certain acts, such as triggering the fire alarm, asking the girl in the library if she believes in God just before murdering her. You can download additional special abilities on computers, such as the ability to manufacture co2 bombs on the fly, if you use up all of your initial supplies. If you run out of ammunition, you can return to your car to get additional clips. Occasionally, you’ll encounter battles where groups of people will team up to try to stop you, but these are never challenging, as it’s usually a couple of sports players against madmen with automatic weapons and homemade grenades.

Once you get to the back of the library, you spot the police outside, and after a short shootout, can choose to commit suicide. Everything you do is explained in character and in the moment by the dialog between Eric and Dylan, forcing you to act in character. You can’t leave without carrying out the massacre; you can’t have mercy on people after you get into battle with them — you either kill them or die. Even when you commit suicide, the choice is to either kill yourself now, or shoot more people first.

After your death, you are treated to a long montage of images, including graphic photographs of Eric and Dylan’s bodies lying on the floor of the Columbine High School library, and pictures of them as children. Eventually, you are placed back in the game in hell, and fight through armies of DOOM enemies, picking up weapons like the chaingun, rocket launcher, and BFG 9000. You’ll collect pieces of the Satanic Bible to gain dark powers, fight demons, and eventually chum with Friedrich Nietzsche, who gives you in introduction to Satan himself, straight from South Park. Satan shows you a scene depicting some of the responses to the tragedy, and then Satan tells Eric and Dylan that they can “blow it all up” in good time, but for now, just to keep him company. Eric and Dylan agree, and the game ends.

PC World rated Super Columbine Massacre RPG as the second worst game of all time. They say “Whether Ledonne’s site has any constructive value whatsoever is still up in the air. But as a game, Super Columbine Massacre RPG is appalling.” This is contrasted against a recent controversy (inspiring the debate I had in school) about the Slamdance Festival dropping the game after initially declaring it a finalist, for the first time in its history. The controversy caused the entire game side of the festival to collapse this year, after one sponsor withdrew and half of the remaining fourteen finalists dropped out in protest of the decision to remove Super Columbine Massacre RPG from the running.

I was talking to about this, and he observed that social roles allow us to excuse games with minimal plot. Absent context, this is a game in which you play as a team of two heavily armed people armed with guns and bombs, who must enter a building and kill everything inside. This is essentially identical to games like DOOM. There are two major differences. The most obvious is that in games like DOOM, you play as an upstanding citizen killing demons, aliens, and monsters, and in this game, you play as two disturbed teenagers massacring their classmates. The second is that this game has more plot, deeper characters, greater intelligence, and is based on the real world.

I guess it comes down to whether you’re looking at it as a lump of gameplay, or a lump of art. As a lump of art, I have no problem saying that Super Columbine Massacre RPG is much more meaningful than DOOM. As a lump of gameplay, the reverse is true. Super Columbine Massacre RPG is entertaining the way rubbernecking is entertaining. It’s fascination with disaster scenes. It’s the thrill that some people get when they see a scene in a movie in which everything they know is being obliterated in a wave of incredible violence. Society collapses, skyscrapers fall, cities explode, schools get shot up. But instead of just being an emotional retelling of the Columbine massacre, it uses video games as a medium, to tell its story from the perspective of being inside of it. Simultaneously, it uses the fact that it is a video game depicting a brutal massacre to cause you to think about the barrier between games and reality. Just as DOOM was blamed by some for part of the Columbine massacre, later school shootings were blamed in part on the fact that the shooters had played the Super Columbine Massacre RPG. Even the fictional hell scenes form a coherent first person fantasy, and I really feel like I’m walking through the boys’ dreams, a hell populated with images from the games they played and the books they read.

I can understand why Slamdance would declare the game a finalist. They were looking it as art, as social commentary, as an exploration of psychology and gaming itself. I can also understand why PC World would say it’s one of the worst games of all time. Still, I don’t think it’s appropriate to remove a game from the emotional and artistic context, and for that reason I disagree with PC World. I don’t plan on ever playing this game again, but it’s not one of the worst games I’ve played, not by a long shot. It’s just psychological.

I’d like to invite you to share your own opinion, whether you’ve played the game or not. You can find the game at its website.

Read Full Post »