The Designer's Notebook: Designing and Developing Sports Games

For most of the last five years, I was an audio and video producer for Madden NFL Football for Electronic Arts. In this column I’m going to pass along some of what I’ve learned about the sports game genre.

It seems to me that sports games don’t get the respect from the developer community that they deserve, and I don’t say that just because I used to work on one. Sports games seldom win major awards except in their own category, and they’re rarely named Best of Show after a trade event. When I talk to wannabes, not many of them say "sports games" when I ask them what genre they’re interested in.

I think there are two reasons for this. First, most developers aren’t jocks. The kinds of people who like to sit twiddling bits in a darkened room have little in common with the kinds of people who enjoy getting all sweaty trying to throw a ball through an iron ring. The fundamental jock/nerd dichotomy is pretty well established by the time junior high school is over, and from then on the two groups are polarized by stereotyping and social pressure. Most game developers didn’t like P.E., didn’t play sports, and don’t play sports computer games. On the other hand, they’re the same kind of geeky kids who read science fiction and fantasy novels and — would you believe it? — a colossal number of computer games are set in those environments. Hence: no respect.

The other reason sports games don’t get the respect they deserve is that they’re unfairly perceived as non-creative. No dragons, no space marines, no story; just people playing a game that you can see on TV anyway. For this I think the press is largely to blame. The game press is always on the lookout for something new to write about, and whatever else athletics may be — exciting, suspenseful, dramatic — they aren’t new. The problem here is that the press (and many game developers and gamers) are confusing novelty with creativity. If we’re really honest, many of the SF/fantasy worlds that appear in computer games aren’t that creative either; they’re rehashes of old ideas and borrowed genres. They seem novel because of advances in technology, but few contain real advances in game design or storytelling.

This isn’t to say that I believe sports games are more creative than shooters and role-playing games, but the creativity is in a different place. It isn’t in the design of the actual game you’re playing, because that’s already given to you. Rather, it’s expressed in more subtle areas: user interface design, player behavior AI, interacting character animations, and intelligent audio to name a few. In the face of ever-more-impressive explosions, that subtlety tends to be lost on the average reviewer.

Here’s why I think sports games deserve more respect: they make a real effort to address the second-greatest challenge in all of computer gaming, and they’re getting better and better at it all the time.

The greatest challenge in computer gaming (and, for that matter, in computing generally) is the creation of credible artificial people. That’s a problem whose solution is still a long way off, and most adventure and role-playing games don’t make any serious effort to address it. The second-greatest challenge, however, arises from the basic premise of almost all sports games: that they are an accurate simulation of the real world.

Sports games’ closest relatives in game development are military simulations. At first glance, it might seem that they have nothing to do with one another, and it’s true that their actual content is completely disjoint. But what they have in common, and what sets them apart from almost all other games, is the claim to realism. Sports games and military sims pride themselves on being accurate simulations of the real world, and that’s why they’re such a challenge to developers.

With most other kinds of games, if the user interface doesn’t work or the AI isn’t smart enough to beat the player, it’s possible to change the game itself to compensate for these problems. With sports games you can’t do that. The game is already designed for you, and you’re not allowed to make any modifications. You must make the user interface work and you must program the AI within the context of the game as it is played in the real world.

All simulations "fake it" to one degree or another — they take shortcuts in their calculations, which cause the simulation to diverge from reality. They generally need to do this because the machine doesn’t have enough processing power to do the job in real time. Wherever they can, they keep the shortcuts out of sight and hope the player won’t notice the difference. In the case of military simulations, the designers can usually get away with it. The number of people who know the true flight characteristics of an F-16 fighter jet is a very small percentage of the game-buying population. If an F-16 simulation doesn’t get it quite right, very few people are going to know the difference. With sports games on the other hand, millions and millions of people know exactly how a star linebacker is supposed to behave, and what’s more, they have the real thing available for comparison on TV every weekend. Sports games are held to a standard of accuracy unimaginable in any other genre.

When you’re designing and developing a sports game, here are some things to think about.

The Rules

The sports game customer is a rather special person: a sports fan who is also a video game player who wants to play the sport on a computer, not just watch it on TV. These people know the game well; they’re not just casual watchers. Attention to the detail of the sport is critical. You must implement every rule and get them exactly right (and in American football, this is really complicated), or the customers will roast you.

Some of the rules may be more complicated than you realize. For example, most people probably think that whoever wins the coin toss at the beginning of a football game earns the right to choose whether to kick or receive the ball. However, that’s not strictly true. In fact, what they earn is the right to choose to kick or receive, or to choose an end of the field to play on (and on a windy day in Cleveland, this can make a big difference). Since the TV networks seldom show the coin toss, you might not notice this — but you have to get it right. Be sure to get the official rulebook from the league whose sport you’re simulating, and study it closely.

There are a couple of important decisions you have to make about the rules. One has to do with the balance between accuracy and user frustration. (When talking about sports games we have to talk about the customer as a "user" rather than a "player," because "player" refers to the athletes who are being simulated in the game.) In games like soccer and football, fouls are fairly common and they’re caused by actions the players take on the field — actions that the user may or may not have control over. In football, for example, holding penalties by the offensive and defensive linemen are among the most common. The user doesn’t ordinarily control these players directly, so the computer could perfectly well simulate penalty-free play, in which the players never commit fouls. However, penalty-free play isn’t how the game really looks on TV, and would seem rather artificial. As the designer, you have to decide: should the game call penalties at random to simulate their frequency in the real game, or is that frustrating for the user because the penalties are not his fault? (Of course, frustration at penalties his players caused is part of a real coach’s experience, too!) In the end what we decided to do with Madden was to allow the user to set how "tough the referees were," i.e. how frequently the game called random penalties. This enabled the users to choose for themselves between verisimilitude and low frustration levels.

Another question to ask about the rules is what to do if you have old-time teams. A lot of games let you play with teams made up of players from yesteryear — DiMaggio’s Yankees, for example. This enables the users to set up fantasy matches that never really happened at the time. But do they play according to the rules used today, or the rules current at the time? Since the rules are usually hard-coded into the software, it’s a lot of extra effort to have multiple variants for different eras in history, especially with a game like football where the rules change frequently. Purists would love it, but are purists your target audience?

User Interface

Camera control and user interface design are critical issues. In other games, the user interface and the game are designed together. In sports games the user interface must be made to fit an existing athletic activity, and many athletic moves do not translate well to a D-pad and buttons. In most soccer-like games (hockey, water polo, basketball) most of the vital action takes place in one location — where the ball is. In baseball, the ball and the baserunners can be 300 feet apart, which makes for some tricky decisions about camera control.

Camera control is especially problematic for single-machine multi-player games. A lot of sports games are played by two people on a single console, watching a single TV. The optimal view for a given player may be completely unusable for his opponent. To be fair, the same view must be equally playable for both of them. You need to think this over very carefully too. Buy a few games and study them. Pay close attention to button layouts, changing modes, and camera behavior.

Sometimes you simply have to make compromises for the sake of playability. In baseball, a major-league fastball thrown at 95 miles per hour requires just over 0.4 seconds to get from the pitcher’s hand to the plate. However, the ball isn’t actually within reach of the bat the whole time. Of the 60’6" that the ball travels, it’s only hittable within about the last six feet, which reduces the time available to hit it to 0.04 seconds. That’s why major league baseball players are paid so much money, and why they’re doing well if they get a hit one time in three. Making a game for the general public, you simply have to fudge these numbers a bit. Video games always throw the ball more slowly, and they allow you to hit it over a much wider area.

Modes of Play

Another thing to consider is exactly what experience you’re going to sell the user. Team sports especially have a wide variety of roles you can play, and over the years we added more and more of them to Madden NFL Football. These are some of the things a user can do:

Look at all the events that appear in a team’s annual calendar — hiring rookies, trading players, training, exhibition games, regular season play, and finals — and think about how many of them you want to implement. Can users build a perfect fantasy team? Can they trade players among teams without regard to salary issues? And so on.

You can also implement special modes of play. A practice mode can let you practice the same activity over and over without actually playing a game. A situation mode would allow you to set up any situation you like and see if you can win from it. In football, for example, with the ball on your own 8 yard line and less than four minutes remaining, can you duplicate the 49ers’ spectacular 92-yard drive to win Super Bowl XXIII?

Licensing

In the early days of computer gaming, most sports games weren’t licensed. The sports leagues didn’t treat interactive entertainment seriously, and the game companies didn’t have the business experience to negotiate licenses anyway. Athletes were used to seeing their names in the newspapers and record books without being paid for them, and didn’t expect, or know to ask, game companies to pay them. The whole business was very relaxed.

All that has changed. The leagues and players have gotten savvy about the value of their names, photographs, and trademarks. It’s still possible to develop an unlicensed game based on a real professional sport, but you have to be extremely careful. The team names and logos cannot be used, nor can player names.

Nowadays, it’s not unusual to have quite a number of licenses associated with a single product. A license with the league will entitle you to use the league’s own logos, plus the team names, logos, and colors — assuming the league owns all those rights; different leagues have different rules. However, the league doesn’t generally have the right to license player names. To get them, you’ll have to negotiate with the players’ union, if they have one, or with the players individually if they don’t. And even if you do have the right to use a player’s name and image, that doesn’t mean you can just put them on the front of the box. A license from the player’s union will let you use the player in the game, but you can’t create the impression that the player has endorsed the product — that would require a special license directly with the player himself.

Old-time players who aren’t current members of the union are another problem. There’s no organization that can sell you the right to use their names as a group, so you have to negotiate with them one at a time. For player photographs, old or new, you’ll obviously have to have the right to use the player’s likeness, and then you’ll have to get an actual photograph from somebody — possibly the league’s photo office or a trading card company — and pay the photographer as well. Even the stadiums have started getting into the act, especially since many of them are now sponsored by big companies. The stadiums are all owned by different groups, some public, some private, so like the old-time players, there’s no one organization that can sell you the rights to use their names and images.

Finally, there are referees, broadcasting personalities, TV networks, and team sponsors to consider. American sports teams don’t usually have a company name on their uniforms, but European and Japanese teams routinely do. Any and all of these can be licensed to increase the verisimilitude of your game — and every license costs money, either in cash up front or royalties afterwards. You’ll have to decide whether the benefits gained are worth the price.

Conclusion

This column has only scratched the surface of everything is to know about sports game development. Sports games are a complex and challenging genre of the interactive entertainment business, with a long list of special problems and considerations not present in most other games. I didn’t address artificial intelligence (tactical and strategic), statistics, motion capture, physics, or any of the myriad other programming, art, and sound issues that make them different from other games. Still, I hope it has served as a useful introduction.

My final piece of advice to a potential sports game developer is this: get to know the psychology of the sport—what makes the players and fans tick. For example, golf is a slow-paced game with a lot of pretty scenery, in which the competitors rarely even see one another. Australian rules football is something between a sport and a barroom brawl. In theory, basketball and soccer are non-contact sports, but the practice differs considerably from the theory. In Britain, soccer is always accompanied by the fans singing drunkenly in unison, and so on. If you’re serious about it, these details will inform every development decision you make, from menu fonts to motion capture. And paying attention to them makes the difference between a good sports game and a great one.