Team Fortress 2 is a game that revolves around one of two main objectives: defend a fortress as a team, or breach a fortress as a team. Almost every detail in the game’s design and construction is done mainly to serve the satisfaction of those two objectives as one of nine class-based characters ranging from the peripheral jousts of the sniper and the spy, the ever-helpful teammate healing medic, to the abrasive war-cries of the hard-to-miss heavy-weapons guy (or Heavy, for short). It’s a game that’s yet to go sour over the two years I’ve played it on and off, binge and dabbled in, purged or built a map for. There’s no mistake that amongst all this exposure it’s afforded me the chance to observe and comprehend the changes made to it by the developers at Valve, and mostly due to the additional content offered as free Downloadable-Content (DLC). Over the course of its run the game has seen most additions in a major two-fold department: variety of game-mode, and the number of maps each game-mode can be played across. When TF2 originally launched with the Orange Box two autumns ago (10/17/07), it essentially came with two modes (which I’m defining by objective): Capture the Flag, and Capture Point. Technically, there was a third, a variation on Capture Points and the initial bread and butter of TF2: Attack/Defend. This mode set up capture points (checkpoints in a map that may switch hands between the “RED” & “BLU” teams until one team controls them all, which dictates win/lose conditions) in a way that positioned one team to defend their given points, while the other is afforded to obtain them all one-by-one, before time ran out.
The introduction of Payload maps tugged on the heart-strings of Attack/Defend by taking the linear staged-checkpoint format where Red defends and Blu advances, but replace the points with one payload-bomb strapped/shoved into a mining cart. All Valve really had to do was approach the idea of Control Points from a radically different standpoint, which was swapping the measurement of time with the measurement of distance. Rather than stand on one spot for a given amount of time, the player escorts a bomb through various checkpoints until it reaches the enemy base, decimating it as the payload teeters into a pit of radioactive materials. Since then, Payload maps Goldrush and Badwater Basin quickly rose towards the upper echelons of most-played maps among the entire catalog, and Badwater quickly became one of my own personal favorites. Goldrush sits at #2 clocking in near 4750 hours played among servers, with Badwater taking #5 just shy 1000 fewer hours, around 3750. Note that the Payload mode didn’t appear until half a year after TF2’s initial launch, with Badwater cropping up then 4 months later.
Valve records most every live-play statistic that could inform future designs and the results of ones already made, but why cite them? Are we simply trying to gauge which Team Fortress 2 map is the king of them all? Hardly. Consider the now infamous map for Capture the Flag (CTF) mode, 2Fort, which stomps every other map in play-time with nearly 7000 hours. Subjectivity plays a sizable role in discerning the kings of level design, but communities and players have voiced generously about the clash that CTF as a mode of play has with TF2’s class-based dynamics. Through Valve’s various design diaries and playtesting, it is arguable that CTF is plausible in TF2 – and that Valve considered it closely, but it doesn’t quite represent what TF2 does best much at all. CTF is known for most often turning into a game of stalemate, and modified by player-behavior into improv-death match mode. What Team Fortress has always revolved around is design that organically encourages and suggests team-work to enhance an experience. Unfortunate for CTF, it doesn’t quite do that goal justice for the majority of the time. I’ve never met anyone that’s said they like 2Fort, on or offline, but I’m presuming that’s because everyone who enjoys 2Fort is already in 2Fort all the time. For some people, one level can become the crux of an entire game.
I just brought up two points that have relevance in the latter half of this entry. One, before this turns into a bully of level designs (it’s not easy designing levels – I don’t want to simply point and shout), I’m attempting to distinguish that the stats in general aren’t there to unanimously decide what is good design and what is bad. I’m personally shocked that 2Fort is the most often played map, but considering what it provides that other maps have not (a layout suitable for death-match) it has filled a void in the game play for TF2. Perhaps the reason it sprinted ahead of any other map is because, since the release of TF2, it’s been the staple for elongated duke-outs, and not so much flags being captured. Additionally, its ranking in comparison to the newest maps/modes that have only seen the light of day for a fraction of the time, and then to each other at an even more granular level, is why I felt compelled to write about kings and their hills.
The designers of TF2 noticed the desire for a death-match mode and took it quite seriously. The behavior apparent in the amount of time spent in 2Fort was telling enough that players probably wanted this mode to make its way officially into Team Fortress, so lately it has. The first mode that championed smaller maps and faster rounds, a more get-in get-out approach to play-sessions for people who couldn’t spend 20-30 minutes getting something out of Control Points or Payload, was Arena Mode. Added a good year ago with Badwater, Arena’s are reminiscent of Counter Strike, another Valve game where the main gameplay hook was only having one life each round. Until the last player on one team falls, the round persists. Arena does this as well, with its own TF2 twist: a Control Point that opens in the center so if a stalemate so-much as makes a whisper, it affords a new way to trigger a win/loss state. Arena mode also ushered in the second visual-theme/setting for the game, mostly by demand for a change of scenery, the Alpine environment.
Arenas were a good foot forward, and even the consistency with a recognizable TF2 game-twist, it never quite felt just right. A lot of what also makes TF2 tick is not just team-work, but losing often to veteran players and learning from your mistakes quickly, via the after-death freeze cam and spectator mode. When you’re not strictly playing you can spend anywhere from 10-20 seconds watching those who haven’t died carry on and observe their behaviors, and most any multi player game is half about studying the opponents patterns and reacting to them whether that be weapon of choice or the route the player takes. Arena mode doesn’t negate this entirely. It’s debatable that by having only one life per round the player is forced to consider their actions and pathing more critically if they want to win, and having more time outside of play gives more time to observe. However, learning by action and getting right back in to help your team win the existing round is one of the cornerstones that made Control Points and Payload modes of all types really come alive and reward the player for gradually learning to work in teams, eventually.
Enter King of the Hill (KotH) mode, just shy of exactly a year after Arenas surfaced. Amidst snowy modifications to the existing Alpine setting, KotH recently brought us a brand new map, Viaduct, and a recent mass-conversion of maps both official and community map. Viaduct is an amazing map, and my absolute favorite in the game now. It nails the type of map that represents what TF2 is in such a minimal amount of time and space, with a simplified objective: take control of the Capture Point for 3 minutes, and your team wins. The point can switch hands though until one team reaches 00:00, and it helps what each class brings to the table breath. While Viaduct is easy enough to champion for why KotH works in TF2, it’s more the mode itself that deserves recognition. Valve yet again took a very fundamental root of an existing mode/technology, and iterated upon it to enhance it. By taking the composition of respawn-times, rules for doors locking, the timer for when the Capture point opens and the time each team has to obtain it, KotH mode presents itself as a functional and well thought out revision of what Arena may have ultimately wanted to do for the game.
I’m not the only one thinking this way, as it’d seem numbers suggest. Let’s revisit the idea of map/mode conversions, the statistics of hours spent in each map, and also the fact that all this additional content is free to all players of TF2. Is this really what players wanted, and is there considerable mileage to be gained from such micro managed iterative design? Just a few paragraphs above is an image taken from Valve’s last update to the game, which not only had KotH mode, but converted maps that were only released 4 months prior as Arenas. It’s also not difficult to assume this is exactly what everyone else who made Arenas was thinking with the Big King Map Pack, a packaged suite of Arena maps that were almost instantly converted to KotH once the mode was available for reference in Source’s SDK, being just one example I can readily point towards.
Within only two months under its belt, Viaduct, as well as the Sawmill/Nucleus conversion, has climbed its way past maps that have been available for play for two years. Not only that, they’ve garnered more hours spent in them than nearly every Arena map to date, almost twice as much in some cases. To break down the numbers (which I’m rounding):
- koth_Nucleus: 800 hours, 14th most played, available since 8/13/09
- koth_Sawmill: 700 hours, 15th most played, available since 8/13/09
- arena_Lumberyard, 700 hours, 17th most played, available since 8/19/08
- koth_Viaduct: 600 hours, 18th most played, available since 8/13/09
- tc_Hydro: around 500 hours, 20th most played, available since 10/17/07
- ctf_Sawmill, 490 hours, 23rd most played, available since 8/13/09
- arena_Sawmill, 480 hours, 24th most played, available since 5/24/09
- arena_Granary, arena_Well, arena_Ravine, each between 470-480 hours, 25th – 28th most played, available since 8/19/08
It’s clear that KotH mode was successful enough to surpass maps that have been around since a years time or more, and that it’s something players were pining for. This applies for both new maps and converted ones as well. Is there perhaps a unanimous agreement among players and creators for the game’s content that this is what people really wanted out of TF2 in quick spurts? It’s simple enough to convert an Arena map if you know what you’re doing, and know how to make entities work for you in Source/Hammer, so perhaps the appeal is well-justified on both ends. The answers to these questions may be relevant to the future cycle of TF2, but what interests me more is the success is founded upon a core principle of modern computer game designs, that being iterative design: the gradual revision of a realized idea and its construction. For starters, most of these changes over the span of TF2’s course seem to have happened in response to player behavior, regardless of whether or not it was intended. Players wanted a more vertical slice of the play TF2 offers, and they got that under Valve’s conditions eventually. Additionally, it shows some sign that the community is always willing to take matters into their own hands. Take for instance a number of maps that tried to implement two payload carts into maps before the first official one, Pipeline, was ever known about or released.
TF2 has continued to be a unique example of iterative design disciplines in today’s gaming industry. There are a dynamic set of ingredients that cause this: 1. There is a robust and accessible library of player-generated content available, as well as the means to teach yourself the techniques. 2. All this content, Valve-made or community made, is free on the PC. Everyone on the native platform gets a fair shake at this content, so it’s up to the player base to instigate these numbers to grow or surpass each other. 3.) Creative people watched & listened closely. By playing the game a certain way, players showed what they really started to want out of a round or two of TF2. By taking it one step further, both Valve and community developers afford these changing behaviors to happen, but in a more intentional way that’s true to the core principles of the game.
Every game that has an extensive online-multi player system is subject to becoming a living project, where balance and reiteration of familiar elements is a common occurrence. Often I’ve heard the critique that games which do this aren’t worth the initial price, that they are simply beta-tests committed to disc with a dollar sign strung to it. Whether or not some feel they can justify these claims, this is the route games have come to for now. If a developer finds the window to make their game more valuable, to make it more reliable on hardware or add to the number of experiences you can have within it, they will do this. It’s staggering how less expensive it is to make content for an existing engine or pipeline than it previously was to start from scratch, and with TF2 taking over a decade to develop that seems only fair.
If there isn’t something for everyone, eventually someone will fill the voids in a computer game of this genre. People showed what they enjoyed by enjoying it, and a lot of talented people have taken note and got to work. We’re in a day and age with game design that we can provide each other with very articulate modes of play and respond very directly between player and developer. Some developers have finally caught on to taking their players more seriously, in concept, than their game alone.
The game continues to give us new locales to explore, weapons to unleash, achievements to obtain, and generally new opportunities to feel slightly empowered for just a few minutes a day, if only to entertain us with frenzied team-work and comedic cartoon-violence. Although player-anonymity is an exclusion to the rule, Team Fortress 2 treats its players like kings — each one eventually finding their “hill” to claim.