Street Fighter 3 and the Arcade Supply Chain
Arcades Demand Innovation
I’ve been playing a lot of Street Fighter 3: Third Strike (1999) lately. Third Strike is most famous for Evo Moment 37: The Diago Parry (circa 2004). Two decades later, it was a mainstage game at this year’s Evo. Despite its age, Third Strike is the most actively-played Street Fighter game (besides the current installment, SF6). The game’s longevity is largely due to its arcade-era design – not because it was physically in arcades, but because of unique design quirks of arcade games that gave Third Strike a polished and distinctive flair, even today.
Arcades influenced game design in a variety of ways. Notoriously, arcade games were quarter-guzzling machines. Arcade games encouraged new players to try them out with simple and visually compelling gameplay. They encouraged repeat play with punishingly high difficulty, or in Street Fighter’s case, salty runbacks. But one under-discussed factor is capital durability and the arcade supply chain.
Today, game developers sell directly to consumers. Often there is a publisher or game store in-between, but overall the incentives are aligned. Developer makes game. Player buys game. Revenue for all parties involved is proportional to the units sold.
Arcades, however, had an extra layer in-between – the arcade operator. Developers still need to make good games, but they are selling to the arcade operator at a high up-front cost. In turn the arcade operator will buy games they believe will generate incremental revenue in the long run via a steady stream of quarters from players. Moreover, arcade cabinets were expensive, largely due to their non-interoperability. Each arcade cabinet came with its own built-in television and computer. This is drastically different from modern games, where one TV and one console can play any number of game discs. As a result, arcade cabinets were very expensive and durable. Arcade operators needed to believe that a game would have long-lived interest and revenue potential to foot the large up-front cost.
This brings us back to Third Strike. Street Fighter 3 was following in the shadow of the massively successful SF2. Every arcade had a copy of SF2, and every arcade operator would be skeptical of the marginal benefit of a sequel. Would SF3 really bring in any incremental revenue? Or would the Street Fighter fans be perfectly happy to substitute in SF2, and the sequel would simply cannibalize existing quarters? The durability of arcade cabinets exacerbated this problem. Beyond simply lasting a long time, cabinets’ physical bulk made them inconvenient to move, store, or dispose.
To avoid this substitutability, Street Fighter 3 was made to be distinct from SF2. The character roster is particularly revealing. While the mascots Ryu and Ken are present, many franchise mainstays like Zangeif, Cammy, and M. Bison are missing. In fact, 16 out of 20 characters in Third Strike were new. Additionally, Third Strike differentiated its gameplay with its speed and aggression. SF2 was a notoriously slow and defensive game at the top level; Third Strike alleviated that by with new aggressive mechanics like dashes, parries, and universal overheads.
The uniqueness of Third Strike is apparent when you look at the Street Fighter franchise as a whole. While each installment does make changes, Street Fighters 4, 5, and 6 all focus on the grounded footsies found in SF2. SF4 and onwards were made in the console-dominant era, and players were happy to purchase mild iterations on their favorite games. Much like Call of Duty, Street Fighters 4, 5, and 6 functioned like a proto-live-service, where sequels offered additional content without fundamentally changing the core game. As the final arcade-first Street Fighter, Third Strike was the last installment to dramatically mix up the formula.
Third Strike also stands out in the modern gaming landscape thanks to its late 90s design quirks, albeit not arcade-specific. The art style is a standout. The late 90s saw a transition from 2D to 3D games. While those early polygons were crude, eventually even 2-dimensional games would be rendered with 3D graphics engines. Third Strike was the last Street Fighter to be rendered in 2D. The artists were masters of hand-crafted 2D sprites, and Third Strike has some of the best pixel art around. While the market indicates that most gamers dislike pixel art, there is certainly a niche to be served here.
On the gameplay side, Third Strike is a precise and unforgiving game. Most modern games give leniency to the player via input buffers and coyote time. Third Strike, in true old-school fashion, does not. When you press a button, the input registers. If there is a 0.05 second window to connect a particularly tricky combo, you better nail that timing. While it certainly feels bad when you can’t execute the string of actions you imagine in your head, this lack of hand-holding also increases the skill gradient of the game. Not many players can land that 0.05 second link, so those who can really stand out. There is a recent trend in games being intentionally difficult, exemplified by Dark Souls, and some contemporary players gravitate towards Third Strike for that execution difficulty.
Third Strike also reaches a wider audience thanks to its low computational load. As a 90s arcade game, Third Strike can be run on practically any computer with emulation. On top of this, the quick performance allows additional online play improvements, such as rollback netcode. As a result, plenty of Mexican kids on toasters are playing Third Strike online thanks to Fightcade.
While I have spun a (hopefully) compelling theory here, it does beg the question – why are there no other long-lived arcade games? If the mixture of arcade innovation and 90s game design are so potent for Third Strike, why don’t we see other games from that era continuing to survive? There are a couple legacy cult esports like Starcraft: Brood War and Smash Bros Melee, but none of the single-player arcade games seem to have survived. Maybe the esports component is crucial to its longevity, as legacy breeds reputation which breeds further interest (see: Chess). It’s entirely possible the theory presented here is totally misguided, post-hoc retrofitting an explanation onto the observed phenomenon that is Third Strike.
A bit of trivia – Evo Moment #37 is not actually a series of several dozen “moments”. The original uploader gave it that title as a joke, implying that there were dozens of similarly memorable moments that were simply not recorded and uploaded.