The Omnivangelist

In the first installment of the General Video Game Strategy, we discussed the limitations on the programs which we match our wits against in battle, and the nature of the difficulties presented by video game designers.

For reference, I’ve listed the limitations discussed in part 1 again below,

1. Limitations to computer programs, in terms of their capability, accuracy and efficiency.

2. Limitation to the intelligence of the designer.

3. Limitations in the hardware that hosts and executes the designer’s program.

Over the years playing many video games systems, many tactics have been uncovered which leverage and exploit the patterns and environment which programmers have presented to challenge us.

I will present several below:

Exploiting dimension:

This tactic proved useful in games that feature 2.5D environments. 2.5D means that the environment has X and Y dimensions, and substitutes depth (Z axis) by shaping X and Y toward a horizon, just as you learned from drawing lessons in art class. You can move left, right, up, down and jump up and fall down, creating the illusion of 3 dimenions within a 2 dimensional rendering (hence 2 and a half dimensions).

Some games that have 2.5D environments are,

Golden Axe,

Streets of Rage,

and Final Fight. And there are a multitude of games of this genre with the same environment [River City Ransom (NES), TMNT4 (SNES) ]. These games started phasing out in the mid late 90’s with the Tomb Raider 3D world era. All of these games however shared the same weakness: The enemies were programmed to be difficult when engaging from the left or right. But the enemy is completely benign when approached from above or below. This is because the enemy (nor yourself) has no attacks in this direction. However, when you manage to align yourself underneath the enemy and walk upward, you can control the distance, time and circumstances with which you will engage the enemy. In Streets of Rage, you can simply walk upward into an enemy and immediately initiate a hold move on the enemy from which you can throw. Exploting this allows you to greatly reduce the difficulty which enemies present.

Dimension was also able to be exploited in 2D games. In many 2D side-scrolling games as you moved through each stage enemies would spawn at pre-determined fixed locations. Every time a precise location scrolled onto the screen, an enemy would appear. And contrariwise, when the player scrolled an enemy off-screen, it would vanish. Careful manipulation of how you scroll the screen could be used to control the creation and destruction of on screen enemies. In the best case scenario, the player could scroll to precisely the location where an enemy spawned, then immediately backtrack in order to destroy it, without backtracking enough to have to cross the spawn point a second time. Thus the enemy was eliminated without the need for direct confrontation, using the limitation of the screen space the designer could work with to destroy his own enemies.

Games which this tactic worked well included NES games, Teenage Mutant Ninja Turtles, Ninja Gaiden, Bionic Commando, and Contra. In the photo below, the bird flying across the screen could be cancelled if you Ryu were to jump to the right and scroll the screen to the right such that the bird exits stage left.

Dimension can also be used by the designer to his/her advantage. In Rainbow Six Vegas for example, you can only fire your weapon straight ahead of you. When you are hiding behind and object, your body is flush with the object. Therefore, the objects in the environment you hide behind control the direction which you can fire at the enemy. Isn’t it convenient that most of the objects in the game are perpendicular to the enemy? Well imagine how much more difficult the game would be if all objects were oriented with a 45 degree angle with respect to the enemy! Much more difficult! This would force you to engage the enemy without cover, or with cover but unable to strike while the enemy can strike. Ouch.

Exploiting memory limitations:

Always keep in mind that the designer making a program has a limited amount of memory (hardware) with which to perform his/her intended function. The older games were more limited than today’s games, but even the state of the art games have finite limitations of memory. Did you notice the load time on Metal Gear Solid 4? The game has to load much of the content of the game onto the PS3’s hard-drive because it’s much easier to load code with high throughput from a hard-drive than it is from a blu-ray DVD. CD and DVD readers, though much faster today than ever, are still relatively slow, and most video games today are on compact disc and DVD based media. This is a limitation.

Take for example Grand Theft Auto 4. When you are driving around Liberty City in your car, there are other cars on the road (usually in your way in fact). These cars consume memory and computational time to render on the screen. Also, the textures which detail these cars require memory, and this memory is rapidly accessed memory, which is even more scarce in hardware (it’s either, RAM, video card memory, processor cache or some similar specialized memory). Even a game like GTA4 has limitations. Have you ever noticed that the types of cars on the road at any given time are similar to each other in terms of car quality? You will likely see poor or middle class cars on the road when you are driving a low-end car. On the other hand, when you are driving a sports car, you will likely see other sports cars. This is partially intentional as the developers don’t want you to always have access to sporty cars, but the other half of it is that the game can only render so many cars on the screen at once, and to have memory available for rendering other objects, the memory spent on textures for cars will be economized such that the cars will share similar textures.

Similarly, in many video games, many enemies will be copies of each other in order. This saves memory and computational capacity so that the designer can instead put more enemies on screen and still maintain a high frame rate on your display.

Believe it or not, but there are ways to exploit this. Think about it.

If you drive a firetruck or some very large detailed vehicle in GTA4, are you not taxing the processor and memory more than if you drive a car comprised of simpler textures? And does this not place a larger restriction on the remaining memory available to the designer with which to render enemies to challenge you?

Looking again to Streets of Rage 2 (Genesis), if you play this game, you will notice that only a certain amount of enemies will appear on the screen at one time. If you refrain from killing enemies and allowing new ones to spawn, and instead simply walk away from the enemies on the screen (they will of course follow you), you can keep new enemies from spawning along the way as you travel and scroll the screen. New enemies will not appear because you already have the maximum number of enemies available on screen that the developer can handle. By understanding the memory limitations imposed on any game, you can often control the time and placement of where enemies will spawn. This is a powerful tool for providing you with the advantage of choosing the parameters and condition by which you will engage the enemy. And you thought that the designers controlled this! Wrong. Other games that this tactic could be employed are Metroid, Ninja Gaiden, TMNT - many NES and genesis games. Amazingly, programmers for the Atari 2600 seemed to be very careful in not exceeding memory and processor capabilities of the hardware.

Have you ever noticed in Metal Gear Solid, that after killing a certain amount of enemies in any given area, the bodies of some of the fallen enemies will start to disappear? This is the system reclaiming memory. There is a limit to how many bodies can pile up. In fact, the bodies need to be removed to create memory for more living enemies. So your thought process should be, “How can I take advantage of the memory limitation?” Perhaps planting claymores on the screen will help. If I plant claymores, the computer needs to remember them and their placement. This takes memory and leaves less memory for the AI to use for spawning enemies against me.

In Assassin’s Creed, the system will memorize certain things in the environment. Say for instance there is a dead body on the street from a guard that you assassinated (sweet!). Oftentimes, to remove a dead body on the street, you’ll have to run to a location far away from the dead body to force the computer to reload the memory and refresh the environment. The distance that it takes you to travel in in order to refresh the scenery is a great measurement of the memory limitation on the designer. It tells you that he chose to allocate this fixed amount of area to memory in order to achieve his desired effect. If for instance, you aroused the suspicion of a guard in the game and simply was able to walk around the block to get the guard to forget you, that wouldn’t be very challenging, nor much fun, so the designer is basically “forced” to grant a certain amount of memory for this purpose. However, what if the system doesn’t really have enough local memory available to support the rendering of the beautiful world found in Assassin’s Creed for the local area AND remember all the detailed alterations to the environment you’ve done? Answer: The developer will need to cut corners. See that poor slug lying on the ground in the Assassin’s Creed photo? He’s a memory hog, and therefore your ally!

So how did the developer cut corners? Well, it depends. But this is what you want to find out. Perhaps if you are being chased by enemy guards, the system will focus more on memorizing that you are being chased on other tasks, prioritizing what memory is allocated to what within the game. Perhaps the memory is allocated such that the lower level of the city is separated from the upper levels, and if you climb ladders, you will trick the computer to refreshing memory more rapidly. Perhaps pausing the game and loading the item menu will force the computer to store all the info going on in your current guard chase to hard drive so that the menu data can instead be loaded to memory. And perhaps this “storing” of the data will be lossy and allow the environment to be slightly more to your advantage when you unpause. This is the type of thinking that goes on…

Often, my battle with a game such as GTA4, MGS4 and Assassin’s Creed are a battle with the mind of the designer and his limitations due to hardware and algorithm complexity. Consider it my way of getting an added challenge out of the game, on top of the regular challenge of gameplay

A more severe form of exploiting memory was available in games such as TMNT(NES) which suffered from such overloading or memory and processor capacity that the game would suffer from slowdown. Slowdown allowed you more time to react to events, albeit not in a natural way. Games like Megaman on otherhand, allowed you to exploit slowdown by “pause-spamming”. Rapidly pausing and unpausing the game would have a similar effect to putting the game in slow motion, which was an advantage of its own. But in Megaman, it was even worse. If you shot a special weapon in the game, and paused the game while the energy beam was in the enemy, and unpaused, the beam would register another hit and damage the opponent. This could be repeated ad inifinitum to the point where certain bosses could be easily defeated with this tactic. Obviously the designer either overlooked this possibility, or placed it in the game as an easter egg. Either way, understanding the nature of limitations on the designer can produce wonderful leverage for you.

Exploiting Enemy/Character patterns and Cancelling:

Of course any game which presents puzzles will display some sort of pattern. This is common sense, and as humans are excellent pattern recognition systems, I will leave the general topic of pattern recognition to you. However, there are a few special exploits of enemy patterns which are unique.

Many fighting games and even games some shooting games fall into this exploit. You get into a fist fight with an enemy and notice that when you spam the punch button you will get off five or six hits, something like Left-Left-Right, Left-Right-Uppercut. After this your punches are halted (your player specific combo is over) and then the enemy begins his/her counter-attack. However you may not have noticed that if you press the punch button at a specific cadence, one punch at a time, you will never enter into your built-in combo and the enemy will simply remain stunned. This is the difference between doing a 6-hit combo once, or a 1-hit combo infinite times. This type of exploit exists in games such as Final Fight, Streets of Rage, and even more recent games like Godhand. In Godhand, if you are fighting a single enemy, they are toast when using this single hit combo. The only way to upset it is to have multiple enemies on the screen, so I will often single out an enemy not near the pack and finish them off one at a time.

Final Fight was even more exploitable at this than other games. If you had a turbo controller (whichi were popular at the time) you could “juggle” enemies that were both on your left and right by holding down the punch turbo button than spamming both left and right very rapidly: face left - punch, face right - punch, face left - punch, repeat, etc. This would keep enemies on either side of you stunned until you basically finish them off in addition to completely occupying the space surrounding you, making other enemies practically wait in line for their turn to get an ass spanking.

Cancelling is a more well-known technique perhaps familiar to players who became competitive in Street Fighter 2: World Warrior and other fighting games. Cancelling is the act of starting one attack or attack sequence and then performing another attack or attack sequence right on top of the first. Many times a powerful attack in a fighting game will require a recovery period afterward, which is usually used to balance out powerful moves (advantage but with a disadvantage). It just so happens that many powerful attacks can be interrupted by a simple attack, granting the player with both the strike and effect of the powerful attack without the recovery time. The 4-hit Dragon Punch in Street Fighter 2 is an example. Generally, the standing fierce strike has a modest recovery time, but if you cancel this strike into a dragon punch, you will very rapidly combine the standing fierce and 3-hit dragon punch into 4 hits, with little recovery time. This technique is also exploitable in Godhand. It’s not that the designer unintentionally let this exploit slip into the game. Sometimes this is intentional, but often it is not. This type of exploit slips into games simply because there are so many possible attack combinations and the code programmed to perform these moves may behave differently when concatenated in specific ways. The fact that there is only a certain amount of play-testing available to the designers of a game in which to balance out the power and weaknesses in the game is limited. Developers, like anyone with a job, have to often cut time and release a product before being able to fully wring out the wrinkles. These “wrinkles” often provide the players with advantages.

In the next installment of the General Video Game Strategy, we will discuss from a more philosophical perspective, and discuss some techniques for decision making.

Note: “Cancelling” is mentioned to an extent in a tutorial video for Street Fighter 2 on the PS2 release of the Capcom Classics Collection.

These tutorials are available here:

Beginner: (search: Super SF2 Turbo Beginner Tutorial (Part 2) by David Sirlin)

Part 1:
Part 2:

Intermediate (search: Super SF2 Turbo Intermediate Tutorial by David Sirlin
Advanced (search: Super SF2 Turbo Advanced Tutorial by David Sirlin)