Level design in video games fall into one of two camps, either linear experiences or randomized levels. I’ve seen both good and bad examples of both; while they both have advantages they also have major disadvantages as well.

With linear levels, the designers can craft unique tight challenges balancing difficulty with periods of rest, a well crafted level can be a thing of beauty. However the charm only lasts the first time you encounter the level, on repeat visits you know all the tricks and the excitement is gone. Random levels keep the player guessing as the geometry of the game space is different each time, unfortunately this removes the tight level design of linear titles.

Also I’ve yet to see a random level generator deal with the issue of huge areas with absolutely nothing going on. I had a thought just now about a way to combine these two opposing elements to allow an interesting take on level design.

Before I give my idea I would like to mention one that has already met halfway in some regard between these two. The director of Left 4 Dead adds randomness to a linear world, while still keeping the tight game play. The game space of each map is purely linear, with crescendo events and end points, however where enemies will come from and what items the player will find is randomize with each start.

This keeps the player on their toes as periods of challenge and rest are different each time. However the director is working on linear maps; no matter how random it is set up to be there is a limit to the number of different elements when dealing with a set constraint. Now here is my idea for creating randomize linear levels.

A level in a video game can be split into sections either by changes to the environment (Move from a gas station to a warehouse) or by explicit changes to the game play (chase section, followed by puzzle, followed by boss fight), now what if instead of just randomizing item and enemy placement we randomize the sections themselves? Let’s take a normal city that is destroyed for example, the player will need to get from their house to city hall in the middle. The player must follow a linear path; however that linear path will change each time.

For example on one play through the player finds themselves near a supermarket which is demolished to the point that they can’t get through it, so instead they cut through the back alley. Now on another play through the super market is fine and the back alley is now collapsed, do you see what I’m getting at. The game would still have the tight linear design yet randomized to keep the experience fresh.

However I can see the major flaw in my plan, one complete level in a game would have to be built several times larger then in other games to accommodate the different randomized sections. Which would increase development time of the game, but I think it could be worth it.

The actual size of the level would not change, if the level takes about 15 minutes on one play through then that should be the same regardless of any changes. I could see this working for games that are built on replay ability over huge 30+ hour games.


Posted By

Since picking up The Orange Box I’ve been hooked on Team Fortress 2, Valve’s first excellent shooter; with the release of L4D one wonders what will be the better shooter. Personally I think that while starting out L4D will draw some of TF2’s fan base just from the newness factor; however I don’t think both games are in direct competition with each other. While both games are multi player based shooters, they both strive for different game play experiences.

Team Fortress 2 is more skill based in my opinion, with an expert player using his/her best class should be able to fight less experienced players and win. While the game is set up for a team based experience, a few good players can make a difference between winning and losing. L4D however is another story, you need your team to be in sync to stand a chance on expert mode, and one player is not going to cut it regardless if he/she is the best player in the group. The learning curve for L4D is higher starting out compared to TF2, thanks to friendly fire and the randomness of the zombie attacks. However as of right now before any content update from Valve TF2 has more variety then L4D with the various types of game play.

The main difference between the competitive natures of TF2 vs L4D is that in L4D the teams are playing different games. They are asymmetrically equal in L4D as the infected team is completely different compared to the survivors. It would be like in TF2 if the red team couldn’t use scouts, medics, or heavies, and the blue team couldn’t use demo man, snipers, and spies. Map design has a greater effect on matches in L4D as the infected must use the geometry of the land to set up ambushes.

Which leads us to this,” what game do I prefer?” Even though I’ve been playing nothing but L4D for the last week, I’m starting to feel the pull back to TF2. With only 4 campaigns I’m starting to see the repetitive nature of the game and versus mode desperately needs more maps. The director does an admirable job of keeping things interesting but there is a limit to how random a linear level can be. On some maps I’ve played them enough times to know where all the ambush locations are and it’s just a matter of going through the motions. Both games are great and I’ve more then gotten my money’s worth with each, however right now I think TF2 has more longevity then L4D as proven by the fact that people are still playing it after a year, whether L4D will still be played as much in a year remains to be seen.


Addendum: As mentioned this comparison is done with the more up to date version of TF2 vs the retail content of L4D. There is no guarantee that my complaints regarding L4D will be around in a year and perhaps a second comparison will need to be done once content patches are released for it.

Posted By

I have officially logged over a day trying to survive the zombie apocalypse that is Left 4 Dead, after spending time tearing my friends a new one in versus mode and surviving expert I have my review.

Let’s get single player out of the way, don’t bother with it. While the AI bots are serviceable it really goes against the core element of L4D, the insanity of co-op. L4D ships with 4 campaigns (20 maps total) with two of the four available for versus mode which I’ll get to in a bit. The five maps of each campaign fit into the theme of the campaign, whether you and your friends must cross a city to reach a hospital alive or getting to a small town the map design has a lot of variety. Co-op as I mentioned earlier is the name of the game; every design decision has been built towards forcing everyone to stay together. Alone the player will be easily swarmed by the mobs of infected and will discover that they are unable to escape thanks to being reduced to a crawl while being attacked. The special infected (boss characters that will appear), 3 of them can disable someone while the fourth is so strong that it requires a group effort to take him down. Survivors who are knocked out can be revived with a helping hand from a friend. L4D rewards groups who are in sync with each other, as a good group can take anything the game throws at them. While 20 maps might not sound like a lot, the game has a trick up its sleeves to make you want to come back for more.

The much touted “Director” was created to keep the players guessing with each play through. Every time a map is loaded the director will determine where infected (both regular and special) will show up, item placement and will send mobs of infected to attack the players on a regular basis. Besides the big defense moment on most of the maps, ambush locations will change on each play through forcing the team to quickly adjust their plan. So far the director has been doing its job well, multiple play through (which on expert will happen a lot) have kept my team guessing and praying that we don’t run into a tank. The game comes with 4 difficulties that are sure to give everyone a challenge. Friendly fire isn’t so bad on the lower levels, but on expert one stray shot can take out your friend. Moving on let’s talk about versus mode, for when you want to become the chaser and not the chased.

Versus mode pits two teams of four against each other to see who can score the most points over a campaign. Each map has one team as the survivors and the other as the special infected who will spawn into the level. Points are earned by the surviving team and are based on how far they get, total health if they finish and more. The infected players have the task of stopping the other team by reducing their health to zero. The survivors have a lot of advantages to keep them going, firepower, huge health, and the ability to revive someone who is knocked out. The infected however cannot be completely killed and will always respawn in about 20 seconds meaning the survivors will need to move fast lest they be overwhelmed. Team work is more important for the infected as alone, they can be easily dispatched but working together they can use their unique talents to destroy the group. Overall versus mode is a lot of fun; however you really need 8 people to make the most of it.

L4D is not perfect and there are still a few bugs in the system. The match maker system for finding servers is not a good replacement for simply finding one for you; there are still a few imbalances in my opinion. The difficulty curve isn’t smooth as advanced is pretty easy for the FPS vet, while expert can make you want to tear your hair out every time a tank knocks you out in one hit. The achievement system is known for not giving the player the ones they’ve earned and some of them would make me livid if I didn’t get them due to a bug.

At this point it remains to be seen what new content Valve will add and if L4D will have the staying power like Team Fortress 2. Still for now L4D is a great game and a must for people with friends looking for an exciting time. I give L4D the correct use of the word “for”.


P.S. If you haven’t figured out the reasoning for the title of this entry, then you are beyond any help I can give sadly.

Posted By

Today I once again watched one of my favorite horror movies of all time, John Carpenter’s remake of The Thing; every time I watch this a little piece of me dies inside over the fact that we’ve yet to see a game based on it. Of course there was the third person shooter a few years back, but other then setting enemies on fire it missed the point of the movie, paranoia. Now through the magic of imagination and nothing to do on Saturday, I bring you my version of a video game based on The Thing.

My game would still take place in Antarctica however this would either be a remake of the movie, or a retelling of the story. The game would begin before any sightings of the thing giving the player a chance to interact and watch their fellow crew mates. Depending on where the player is in the base at that time, they may or may not witness the thing’s arrival. Once the thing has arrived it will be up to the player to convince everyone else that he is not infected and attempt to lead them.
Each person and animal would also mutate into a different Thing once discovered, but the same method will work to kill them, burning them alive.

Heading outside, the player will need to watch their body temperature as it is below freezing out there, and of course the only way to move around easily is using the guide lines. I can envision the player and any surviving members going to other camps some destroyed others perfectly fine, and have to figure out if the thing is at work.

Granted this idea is no where near completion, it gives me some solstice that perhaps someone out there is working on a game like this.


Posted By
Return to Top ▲Return to Top ▲