MULTI-PLAYER WEB GAMES ON A SHOESTRING

The Story of CyberSkipper™

#3403

James P. Abbott jabbott@longshot.com http://www.longshot.com

CyberSkipper is a multiplayer baseball forecasting game that was designed exclusively for publication on the World Wide Web on a shoestring budget. It has grown, in just over a year, from an unrealized idea to a reality that has now become an important part of the relationship between thousands of fans in the ten markets where Inventure has licensed the game to a Major League team and/or the major interactive newspaper in the market.

What a Game! Cyber Skipper is one of the best sites on the Internet. I have been an Orioles Fan for my entire life and I have never tracked player statistics as I do now. Cyber Skipper makes baseball games more exciting and enjoyable to follow. Jami Casamento Fri, 21 Jun 1996

The growth of CyberSkipper has been accomplished principally with the use of intellectual capital -- and without the use of any venture capital. This paper explains why the Web was chosen as the medium for CyberSkipper, what software was created, and how CyberSkipper was bootstrapped into existence.

WHY THE WEB?

CyberSkipper takes advantage of the Web's great strengths in database management and prompt information delivery. Play on CyberSkipper is almost completely asynchronous, permitting thousands of players to play each day without suffering any disadvantage because others played before them, and without causing delay in play while waiting for others to move.

Thus, CyberSkipper allows competitive multi-player play without ever gathering everyone together at one time. At the same time, because its design does not rely on instantaneous reaction time, it avoids any significant negative impact from the latency period between a player's moves on his client computer and the response from the server.

CyberSkipper also benefits from the ability of the Web to control a relatively complex set of variables, and make the game appear quite simple to players. This is accomplished through the use of an interactive playing grid that guides play in accordance with the rules of CyberSkipper.

MARKETING ON THE WEB:

The Importance of Branding

CyberSkipper was marketed under a plan that placed great reliance on the economies of Web publishing: the low cost of production and the low cost of distribution. The same plan recognized, however, that one of the most important and prohibitive costs associated with game publishing are those connected with marketing and promotion, and that although there are many ways to market though the Web, there is an ever increasing problem with clutter and confusion. The very fact that it is easy and cheap to produce for the Web ensures this. As millions and millions of pages are added each month, Web search engines are hopelessly overmatched. Bottom line: without a powerful marketing campaign there would no way to ensure that even the best of games would not die of loneliness.

Because the Web does not have an effective way to reduce marketing expenses, Inventure addressed the need for promotion and marketing of the game by creating a game that served the goals and objectives of the major brand-holders in the target niche -- Major League Baseball teams. It then presented the game to those brand-holders on terms that were attractive enough to earn the powerful partnership of the Baltimore Orioles, Atlanta Braves, Oakland Athletics, and Los Angeles Dodgers before the game was even six months off the drawing board.

The key to these relationships lay in the decision, early in the design process, to equip CyberSkipper to function as a marketing tool, not just a fun game. Fortunately, the objective of creating a fun game and the goal of creating an effective marketing tool were nicely congruent in this case. CyberSkipper was designed to enhance team loyalties and drive traffic to Web sites. But it could not have been marketed this way had the game not been fun to play.

A Broad Audience Within a Narrow Niche

CyberSkipper was developed in order to serve the fantasy sports game market -- a clearly identified niche audience present in large numbers on the Web. Inventure sought, however, to avoid direct competition with others in this niche by creating a game that was fundamentally different in design from the Rotisserie™ style games that had come to dominate the market. CyberSkipper was designed to allow fans of a particular Major League Baseball team to play a fantasy baseball game -- or, as we prefer to call it, a baseball forecasting game -- based entirely on the play of one Major League team. This had not been done before.

All prior "Rotisserie™" style games required that game players build their fantasy team out of players drawn from all over both Major Leagues. This rule divided the team loyalties of those who did play. Because of this, Rotisserie™ games were not attractive allies of baseball franchises. CyberSkipper focused on enhancing the relationship between fans and their team, and therefore had a powerful story for prospective partners.

I think you guys have come up with a great game, and a nifty way to keep busy fans interested in the braves.... Also, do you guys plan on keeping this game going next year when the season starts? hope so!!! DANIEL BALDWIN Sun, 04 Aug 1996

The rules of Rotisserie™ style games also narrowed the target audience to baseball zealots with knowledge of all teams. CyberSkipper sought to broaden its audience by making a game that could be played by game players with only a limited knowledge of baseball. Because a CyberSkipper player only needs to know the players on his or her favorite team, many more fans self-qualify to play. CyberSkipper continued to seek the loyalty of aficionados, and it accomplished this not only because of its underlying complexity, but also because of its prompt feedback and its direct relationship with the play of each separate game during the season.

Finally, Inventure sought to broaden the audience of CyberSkipper by designing play so that the game could be started, and played competitively, at any point in the season. This was a remarkably successful strategy, as almost 4,000 Atlanta Braves fans joined the game even though the Braves version of CyberSkipper did not appear on the Web until late July.

The value of a dynamic daily data input event

For games to become popular, they need to be played repeatedly. For games to be useful as a marketing tool, games need to be played repeatedly. The program that runs CyberSkipper was developed by Inventure in order to permit the company to hinge the game around the statistics that are generated each day by the play of Major League Baseball teams, and that are reported in the form of "boxscores." By identifying the boxscores as a critical dynamic data input event that would recur each day, day after day, through the six month Major League Baseball season, Inventure was able to utilize real time, real world variables in order to determine the outcome of the CyberSkipper games each day. At the same time, Inventure guaranteed that the principal motivation for people to return to the game each day would be created dynamically out in the real world, and would not have to be generated out of the game itself. Every announcement of the baseball games to be played each day also reminded existing players to return to CyberSkipper.

Just a note to say how much I enjoy the Cyberskipper game. I have been playing Oriole ball all season and I must admit, the first thing I check each morning is my results from the night before. Internet Maine Thu, 26 Sep 1996

This meant not only that Inventure could tie its marketing push to a pre-existing promotional engine -- Major League Baseball in local markets -- but also that it could devise a game in which the core game event was already of intense interest to the target audience. CyberSkipper did not have to generate interest in its content out of thin air. Millions of people already cared deeply about the baseball game results that drove the CyberSkipper engine. CyberSkipper merely gave them a chance to feel even more strongly about the games, because now they were interacting with them on an almost real-time basis.

Cyberskipper is a great idea. Have been a Yankees fan since 1960 but like many had become disillusioned with what was going on. I've coached H.S.baseball for 20 years,but other than the Yankees hadn't really been keeping up with baseball even though I get U.S.A. Today Baseball Weekly. Now my 10 year old son and I keep up with everything daily. How many more teams will you be adding? If you get too many my wife is liable to shoot me. Kenny Collins Sun, 18 Aug 1996

CyberSkippers now can watch a baseball game with two objectives that are not inconsistent: they can root for their home team and they can root for themselves to be proven the most knowledgeable and savvy of fans. This has never been true for Rotisserie™ style games, because in such games players always had to divide their loyalty between their home team and the players on the opposing team who were on the game players' fantasy team. Because CyberSkipper is more deeply centered in the play of real teams, the game is more properly called a "forecasting" game, not a fantasy game.

Moreover, a Web game based upon dynamic daily data input from the real world had the appeal of removing pure randomness from the variables that drive play. Historically, games range from those based upon chance, such as casino games, to those of skill, such as chess. Basing a game on a dynamic database that appears almost random, but that is generated in fact by events that are to a large degree predictable, creates a unique game of skill -- and the skill lies not only in the ability to forecast, but also in the ability to maximize by strategic play, the return (in game points) from a correct forecast.

I'm in the middle of writing my dissertation final draft which I defend on October 15. I really shouldn't be paying so much attention to this game, but it is addictive. Just don' t go announcing my name as winner :-) until AFTER I deposit my thesis! My advisor would wonder.... Dave Bott Wed, 11 Sep 199

Baseball players tend to recreate their statistical performances year and year throughout their careers. These statistics, as well as information identifying for each hitter the pitchers against whom he tends to perform well and poorly, are available. This creates a very happy blend. The newcomer can play intuitively and have fun, while the veteran can apply his or her skill and knowledge to make strategic plays based upon calculated probabilities.

It was no coincidence that last year the CyberSkipper short-term contests (in which players all started from scratch, again and again throughout the season) repeatedly saw the same players scoring in the top ten (among thousand of players). This evidence of learned skills promises a bright future for ventures that generate games with an educational purpose based upon other dynamic databases of relatively predictable information -- especially if the information is more important (if not more engaging) than the morning boxscores. People can and do learn from games. Moreover, these games can be created and marketed at low cost in partnership with brand holders who already have a following, and who have an interest in intensifying the involvement of those followers and expanding the size of that group.

It should be clear that the CyberSkipper model is not relevant to the baseball market alone. The same principles that drove the success of CyberSkipper in the baseball market are equally applicable in other sports markets. Nor is the concept relegated only to sports. Dynamic data events are being generated in virtually every area of human endeavor, and are now being reported with a consistency that they never were before. Inventure already has projects underway with respect to the financial markets and the weather, and there are many other sets of dynamic data, and many ways to use such data in a variety of game contexts.

WHAT WAS DONE:

Innovations In Game Design

Prior to CyberSkipper, baseball fantasy/forecast games (Rotisserie™ style games) always have required game players to make their strategic plays by selecting individual baseball players from among a large set of possible baseball player choices (generally all of the players in both Major Leagues or all the players in one of the two leagues). In these prior games, once each game player had formed his or her fantasy "team," each game player automatically earned the same value in game points as other game players who selected the same baseball player -- based upon the baseball player's actual real life on-field performance.

Different prior fantasy/forecast games have counted different statistical categories to determine the value of player performance, but they have always counted and weighed the selected categories the same for every game player who has selected a particular baseball player. In all prior baseball fantasy/forecast games, therefore, the key game objective was to select the right group of (20 to 30) players out of the very large pool (300 to 600) of available baseball player selections . The variation within game play was created by the fact that game players inevitably end up selecting different groups of players who generate different statistics and therefore different game results.

CyberSkipper turned the Rotisserie™ style fantasy/forecast game model on its head.

Instead of the past practice of treating the on-field results of each baseball player the same for each game player and generating the variation necessary to make the game competitive by having each game player "own" different baseball players, CyberSkipper created varied outcomes for many game players by permitting them to make a wide variety of different strategic forecasts with respect to each of a small number of players.

The first critical innovation in CyberSkipper is that all game players must play with the same very limited number of baseball players (10-15), all of whom play for the same team. Game players cannot choose broadly among all Major League baseball players, but must restrict their choices to the players of a single Major League team.

This is an imposing game constraint, because if the same rules as in all prior Rotisserie™ style games were applied in such a game environment, there would not be enough variation to create competition among large numbers of players, because many game players would choose the same baseball players for their "team" and would end up with the same game score. On the other hand, it is this very constraint that makes CyberSkipper attractive to so many players. Because of this constraint, players do not need a working knowledge of baseball players beyond those who play for their favorite team. This opens up the game to a much wider group of players, but how is it made into a competitive game?

The second critical innovation in CyberSkipper is that, in distinction from all prior baseball fantasy/forecast games, a single baseball player can be utilized in CyberSkipper in many different ways. Despite its apparent simplicity, CyberSkipper players are offered the opportunity to select among 51 different strategic plays for each of the limited number of available baseball players. This means that in CyberSkipper, the limited number of available baseball players actually can be selected in many millions of different strategic combinations.

The different "strategic selections" for each baseball player are generated by selecting a limited number of players for zero, one or two categories out of five common baseball statistical categories (runs, hits, rbi, Hrs, and multiple-hit games), and CyberSkipper adds a third critical innovation. A game player may select categories for a baseball player as "risk" categories, in which the score from a particular category is doubled (at the risk of losing points if the baseball players fails to generate an entry above zero for that category). Through this mechanism, the amount of variation available in each game player's use of a particular baseball player is dramatically increased.

Variation in these game player moves is increased by the fourth critical innovation in CyberSkipper. Game player selections can be instantaneously changed at any time during the baseball season, allowing each player to constantly completely revise and update his or her selections. In all prior fantasy/forecast baseball games, changes were rigorously constrained by rules limiting "roster" moves.

Finally with respect to game design, the fifth critical innovation in CyberSkipper is that all play takes place on bright, colorful graphical interfaces that invite participation and immerse you in the baseball experience. Major competitors that have taken their Rotisserie style games to the Web have presented them in a way that has been singularly unattractive, even to the most devoted of "Roto" fans. Their games look like spread sheets, and it's no wonder that their families wonder why Dad doesn't "get a life." CyberSkipper draws the family into the game.

Innovations In The Player Interface

The CyberSkipper cgi program runs two small server-based playing grids on which a limited number of playing squares create a very large number of options and opportunities for play. Each square in the grid is assigned a particular value in the game, and the player interface program does the following:

1) it keeps track of the value of each square -- a value which varies according to the type of baseball event that is being tracked (such as hits or runs scored), and according to the position of the batter in the batting order. This means that a square can be worth as little as 1 and as much as 8 (times the number of times an event recurs) -- under normal game conditions;

2) it reports to players the scoring rules by way of a help function accessible through the grid;

3) it changes the "values" of the squares when a game player decides to place particular emphasis on a specific square (through the selection of the square as a "risk " square), so that individual squares range in values from as little as 1 to 16 (times the number of times that an event recurs);

4) it enforces a relatively complex set of CyberSkipper playing rules by automatically closing off those portions of the grid that can no longer be played, once certain moves have been made. For example, the playing grid automatically closes off a five column row as soon as two columns have been selected, and closes off a nine row column as soon as three rows have been selected. This guides the game player to compliance with the rules, while leaving the game fun to play;

5) it records and sends to the database, on a completely flexible timetable that permits asynchronous play by competitors, the selections made by thousands of remote players who access the program through the Web;

6) it permits the development of "friends leagues" in which game players easily can organize and track competition at personal, family, or community levels;

7) it allows players to respond each day to subtle changes to the lineup made by the Major League team's real life skipper; and

8) it provides direct links to additional information that educates and informs players about strategic moves within the game. For example, players can obtain very specific statistical information about the performance of hitters against the pitchers that they will face each day.

Innovations In Database Construction And Management

Inventure has created a robust database designed specifically to serve the need of a multi-player Web game that is based upon a daily dynamic input event. The specially programmed database performs the following functions:

1) it holds all asynchronously posted selections in its database, and permits players to make any changes desired, right up until game time each day during the season;

2) it closes off entries automatically at game time;

3) it records in its database the outcome of the real life Major League games on which CyberSkipper is based as soon as that information is provided to it, automatically incorporating the data sent to it from a third party without human intervention;

4) it compares the real life results with the plays of all competitors in a process called "closing;"

5) it calculates and records all game results according to CyberSkipper rules and automatically returns those results to the individual players pages, to the central tally of overall, contest, league and daily results, and to selected subsets (such as Friends Leagues) of the overall database. All this occurs shortly after the end of each Major League game;

6) it publishes on the Web in standard format all results set forth above, adding the new scores to the running totals in each type of contest: daily, short-term, league and overall;

7) it awards, and places on the home page of each winning player, graphical "trophies" symbolizing the victory; and

8) it resets itself and begins gathering selections for the next game.

HOW CYBERSKIPPER WAS BOOTSTRAPPED

When one begins the process of designing a multiplayer game for publication on the Web, it is critical that he or she concede that the design should conform to the audience, rather than the other way around.

If you want to make a profitable game, you must be ready build it for an audience. It is romantic to think: "If you build it, they will come." This sentiment is not, however, realistic in a world where there are hundreds of competitors seeking to win the attention of your potential players. You must think of the people who are going to play your game as an audience, and must consider that while it will be relatively easy to present your game out there on the Web, it will be hard for you to reach your audience, no matter how good your product might be, if you do not design with that objective in mind.

The Appropriate Platform

Before construction of CyberSkipper began, Inventure identified a specific audience that it could serve with a very specifically designed game, and which it was convinced was present on the Web. It then designed a game specifically for this audience. It also designed the game so that it could reach the broadest possible audience within the identified niche. This meant that it did not seek to incorporate at the heart of the game any new Web technology that was not necessarily available to those who were on the Web.

For example, Inventure chose to present the game in a simple client-server cgi interface. It also provided the option of a java-based front end interface that people who had java enable browsers found quite attractive and useful. But had Inventure chosen to provide CyberSkipper only in that particular platform, it would have eliminated perhaps two-thirds of its potential audience. The truth is that strategy game designers have not even begun to utilize effectively the most standard of available tools in html programming in order to present a compelling game interface on the Web, and they should focus on mastering these wonderful tools before leaping to each new platform -- where the game may be cooler, but will also certainly be more lonely.

Dependable Traffic

Inventure designed CyberSkipper so that anyone who became interested would be likely to play on a regular basis, returning to the site again and again. There are three main reason for this, two of which we discussed above:

1) the dynamic data event input; and

2) the incredibly detailed results reported by an agile and robust database engine.

The third, and apparently the least important (judging from Inventure's experience), incentive that drives people to return is the offer of prizes and recognition on the Web. With respect to CyberSkipper, Inventure built enormous traffic and player loyalty long before any prizes were announced.

Honestly I have had such a great time playing all teams in Cyberskipper that I have not even paid attention to what I win or how to receive whatever it is that I win. If you could let me know I would really appreciate it. "The Dees Family" Mon, 30 Sep 1997

Inventure isn't the only firm that calculates that it is the fun, not the prize, that drives the traffic. ESPN offered a nation of avid college basketball fans only one trip and 64 pizzas in its Web-based contest that asked players to forecast the outcome of the NCAA basketball tournament.

Low Cost Production

One of the principal considerations for Inventure at the outset was that it had to keep its costs down. At the starting point, it was not in a position to approach venture capitalists and ask for large sums of money to fund its efforts. Inventure simply had an idea for a game. It had virtually no proof of concept, no proof of a revenue stream. In light of this, the option of significant funding up front really wasn't even there. Instead, if Inventure wanted to go forward, it had to resolve that it could afford to work on "spec" for an extended period of time and that it could devote its intellectual capital and a limited amount of financial capital to the enterprise. In addition, Inventure had to find individuals who were willing to begin working on the project under approximately the same terms. The company was very fortunate to have available, from the start, a brilliant programmer who was capable of realizing the vision for the game, and was willing to "roll the dice" with the other principals, at least in part.

In order to be able to construct this kind of enterprise, Inventure had to be a "virtual" corporation, which meant that although it was turning out a professional product and presenting a unified corporate front, it actually was composed of a series of independent individuals or entities that had joined in a strategic alliance to produce CyberSkipper. This provided great efficiencies, as production capabilities could expand to meet demand, and contract during slow periods.

Ready, Fire, Aim

Inventure's decision to keep costs down, and to bootstrap production, would have been futile had it not been for the company's successful pursuit of relationships with the major brand holders. This eliminated from the CyberSkipper budget the need for marketing and promotion costs, and permitted Inventure to present its product to a large number of perfectly targeted players the minute that it was ready -- at virtually no incremental cost. This was perhaps the most important and successful business decision in the CyberSkipper story, but production began before even one of these relationships was in place.

The dice were being rolled, and a second message from the CyberSkipper story is that it is all right, sometimes, to move forward with actions before the target is fully within sight.

Without question, Inventure operated under a plan when it commenced CyberSkipper. But it did not draft a formal business plan document. The underlying game design principles discussed above were carefully followed, but Inventure's efforts were focused centrally at first on producing the product, so that the company could have something concrete to present and from which it could build. Thus, rather than spending an extended period of time and effort developing business plans and perfecting a corporate presentation in order to obtain funding from outside the corporation, the CyberSkipper team just went to work on CyberSkipper. Had Inventure chosen to focus on fund-raising at the start, we might still be talking about a interesting idea, rather than an exciting reality.

The bottom line is that if you want to build a Web game it is possible to do so if you have a nucleus of talented, dedicated people who are willing to devote their time to the project. This is your "intellectual capital." I like this term better than "sweat equity," but in truth I am describing the same thing. If you have this form of capital, then you are going to be able to begin your project, no matter what. The success of your project, on the other hand, will depend on your talent -- and on the care with which you have selected your subject matter and your audience.

Building on Main Street

The selection of your audience -- and of your possible brand-holding partners -- will in large part answer the question of whether you will be building your product out in the Web wilderness, or will be presenting it on a Web "main street." And, believe me, you want to build on the main streets. The key difference for Inventure was that, at no incremental cost, it was able to construct CyberSkipper (and later CYBERCOACH) on central thoroughfares through which its audience was bound to pass. This was because Inventure designed a game that was of interest to the Major League Baseball teams. Because they own the key brands, these baseball teams are the "main street" as far as baseball games are concerned.

So, what kind of deal should you consider with the potential branded partners that you might have? Of course, you can approach it as if the branded partner should pay you large sums of money up front for the privilege of having your terrific game on their Web site. But I can tell you that is an approach that is unlikely to work -- at least before you are established, and probably never. Instead, you are much more likely to have success with an approach that presents the brand holder with a business opportunity that looks so attractive that it it is almost impossible to say no. What this means is that the deal has to involve little or no up-front, out-of-pocket cost to the brand holder. But low cost -- and low risk to the brand holder -- is not nearly enough. The deal has to provide attractive positive incentives.

The incentive that makes the deal happen can be the addition of a powerful interactive element to an otherwise unattractive or uninspired Web site, but it also should be potential monetary profit. If you look at the transaction from the perspective of the brand holder, you want them to be thinking:

"Why should I say no? I'd be turning away interactive content that I can have for no cost and that might one day earn me a profit. If I say yes, I will add this powerful new element to my Web site without adding any cost to my budget, and these guys might make me a hero on the bottom line as well."

If you cannot afford to place the brand holder in this mind set, you are unlikely to be able to make the first deal. If you can, you will probably have a deal. But when you make such a deal, make sure that you are only leasing your game through a license, and not selling it outright -- but that is the subject of another paper altogether, as is the following discussion of the revenue models, which we will address briefly in order to whet your appetite.

Revenue Models

When you are discussing the possibility of profit for your game on the web with potential licensees -- the brand holders -- there are three different models that you should consider. These are the advertising model, the tiered services subscription model, and the straight pay-to-play model. Inventure has chosen the advertising model at this stage in the growth of CyberSkipper and the growth of the Web.

When you build a dynamic, fun, attractive, and free game that creates the incentive for people to come back repeatedly, you build Web traffic. And when you build Web traffic, you build the opportunity to sell advertising, allowing others to use your game interface to market their goods and services to the people who are coming to play your game. This advertising can be sold at prices that range from $10 to $100 per thousand page views. You can do the math, but let's pursue one example:

Let's assume that you can attract 1000 people to come and play your game every day. If that group returns every day, and if each member of the group on average views three separate Web pages in order to play your game, you will generate 3000 daily page views. At the rate of $30 per thousand page views -- a fair average rate -- you would be generating $90 per day from your game, even if you only put one ad on each page. This total may not sound like much -- and it certainly wouldn't be in the traditional game business -- but on the Web, your production costs are very low, and your potential for scaling up is quite high.

For example, CyberSkipper grew to a player base of 7,000 in only a few months time. Only half of these people played on a regular basis, but Inventure generated over 3000 daily page views on both the Atlanta and the Baltimore games. This meant that Inventure had page views that were worth in the range of $100 per day in each of these games. Even if the use does not increase -- which is unlikely -- and Inventure scales out to provide games for all 28 Major League teams, it can move from a situation where it is generating page views worth $100 per day, to where it is generating page views worth $2800 each day.

Of course, this example discusses page views that CAN be sold, not page views that HAVE been sold. The problem with the advertising model is that advertising on the Web is still in its infancy -- and it is the largest actors in the marketplace who have thus far controlled the vast majority of the advertising. This is why Inventure has turned to partnerships with major media companies -- interactive newspapers -- in each of the baseball markets in order to earn the commitment of a partner that can and will sell the page views effectively through the use of existing staff.

Inventure is planning to introduce a CyberSkipper PRO game in mid-summer that will represent an application of the classic "tiered services" model. In the Fall, CYBERCOACH is likely to be be presented purely as pay-to-play. Come back next year for the next installment of the CyberSkipper story.

Conclusion

It is possible to create and to successfully market a well-designed game on the Web today on a shoestring. But this is only so if you plan a wise strategy that enables you to earn the partnership of major brand holders, and only if you are willing to make the compromises necessary to close such deals. Good luck.