In October 1984, as I envisioned what the new version of The Oregon Trail would be like, I saw many crucial ways in which the product should be quite different from the original OREGON. One of the most important differences would be the opportunity to interact with characters in the game. In the original game, this concept was completely absent. Other than shooting at bandits and “hostile riders”, the player never meets another human being on the entire 2000-mile journey. I saw so many ways in which the game could be enriched by human interactions – or more precisely, by interactions with human characters in the game. At the very least, I wanted the game to be full of reminders that there are other people on the trail – some of them travelers like you, and some who are not going to Oregon at all.
However, it was much easier to imagine these interactions than to design and program them. As a result, perhaps the biggest struggle throughout the entire 10-month project was to precisely define the nature of the character interactions, and then to create the stories, data, and algorithms to make these interactions happen. In several of my documents during the early stages of the project, I described highly ambitious goals for the character interactions. But eventually reality took its toll, and I had to dramatically scale back these plans. That said, I did succeed in including several kinds of character interactions in The Oregon Trail – none of which were present in the earlier versions of the product.
As I thought about the concept of interacting with people on the Oregon Trail, I saw a huge distinction between 1) meeting and interacting with strangers along the way, and 2) interacting with your family or other people traveling with you. These two categories provided very different challenges.
In the original timeshare version of OREGON, the very first paragraph of text includes the following: “Your family of five will cover the 2040 mile Oregon Trail in 5-6 months – if you make it alive.” However, your family is never mentioned again, except for an occasional random event that says that a child got sick or broke a bone. So it is easy to forget that you are supposedly traveling with four other people. Also, in the original, it is always the player – never a member of the family – that gets sick or that dies. To me this presented an intriguing opportunity. What if the game made it clear that you were responsible not only for your own health and safety, but also for the well-being of your entire family?
On the real Oregon Trail, in addition to interacting with other people in the same wagon, you would probably interact with emigrants in other wagons traveling together with you. Multiple wagons would often band together for safety and to assist one another through difficult times. These alliances were seldom permanent. If the various parties in a “wagon train” ended up disagreeing about some major issue, then some of the wagons might split off to go their own way. Some wagons might break down or otherwise be lost along the way. So if a group felt it had gotten too small, it might join up with another group it encountered.
I felt it was really important for the product to communicate the idea that you were not traveling alone. Your wagon carried more people than just yourself, and your wagon was probably traveling in the company of other wagons. But how could I communicate this? I thought about this for a long time, and I brainstormed with my colleagues on the team to see what ideas they had. I wrote down several different ideas and thought through the consequences of each approach. Eventually I reached the following conclusions:
I felt that the above decisions represented an appropriate balance and tradeoff. I was excited that the player would travel with a “family” in my new version of the game. However, I was disappointed that I had not found a good approach to convey the concept of wagons traveling in groups.
Many months later, when the game was programmed and we began to test it with children, I was amazed to see just how effective the “family” concept turned out to be. Some kids, especially those playing it for the first time, were emotionally affected when their family members died along the way. Most kids realized that these deaths might have been caused, at least in part, by bad decisions that they had made as the leader of the wagon. This effect was augmented by the fact that many of the kids used the names of real family members in the game. Most of these kids resolved to try again, to prove that they could safely lead their families all the way to Oregon.
On the other hand, in the initial user testing with kids, we found that it sometimes caused confusion if the player’s character died while one or more family members were still alive. The kids would say, “We haven’t all died yet, so why can’t we keep playing?” The solution to this issue was fairly simple. I changed the game design so that the player’s character would not get sick or die until after the other four family members had died. Not only did this remove the confusion, it actually improved the gameplay quite a bit.
In my earliest concept documents, I was already proposing various ways that the player might meet people along the trail. In a document dated October 24, 1984, I mentioned nine different categories of people that the player might meet along the trail:
A particularly interesting detail in the above list is that I divided “emigrants” into four different categories, depending upon their current situation.
In a document dated November 16, 1984, I presented a detailed model of how the interactions with these various people might be structured. My document contained a complex mix of ideas – some of which actually made it into the product, and some of which were later abandoned for one reason or another. It begins by defining three broad methods of meeting people:
The document then goes on to present each of these three types of interactions in some detail. Keep in mind that this was an early concept document, not a final design, so I was well aware that the concepts would evolve over the next few weeks and months. These three interaction models are described in some detail later in this chapter.
In the original OREGON game, Native Americans were barely mentioned. Only one of the 22 random events specifically mentioned Indians – and that is when “helpful Indians show you where to find food”. But there were also frequent attacks by “hostile riders”, which allowed the player to shoot at the attackers. (In the original timeshare prototype, the phrase used in this situation was “hostile Indians”.) In my opinion, neither one of these interactions seemed appropriate for the new product – but for different reasons:
However, now I had a dilemma. If “hostile” attacks and help in finding food were both going to be extremely rare events, then would I include instead? I knew that interactions with American Indians would be very important, but I also knew that such interactions would be subject to intense scrutiny and critique – so I needed to do it right. Therefore one of the first notes I wrote to myself was a set of seven interaction rules:
The rules listed above are somewhat cryptic, so here are the same seven rules with a bit more detail:
Now that I had a set of guiding principles, I needed to provide some examples of what I meant by positive, non-violent interactions (rule #1). So in another note to myself, I created my first two examples, based on realistic interactions that I found in my research:
With the guidelines and examples in place, I could now plan the human interactions that would occur in the game – including a variety of interactions with Native Americans. In my design documents, I focused on four principal types of interactions with Native Americans:
In addition to these four principal types of interactions with Native Americans, I envisioned that Native Americans might be mentioned in certain random events. I was also intrigued with the idea of incorporating the Whitman Massacre into the product – a real historical event that occurred along the Oregon Trail in 1847. (For more information on this event, see Chapter 9.) If all of these ideas had made it into the product, then there would have been many mentions of Native Americans – and the portrayal would have had a good balance.
For various reasons, I had to scale back these plans; some of the ideas made it into the product, and some did not. In the finished version of The Oregon Trail, there are several opportunities to speak with Native Americans at landmarks. At several of the landmarks you can also see Native Americans in the full-screen graphic that depicts the scene. And you can hire a Shoshone guide to help in crossing the Snake River. However, these interactions with Native Americans are quite limited compared to what I had originally hoped to include in the game.
The first of my three main categories of character interactions was to “talk to someone” – that is, you strike up a conversation with someone you meet, without trying to conduct business at the same time. I described it this way:
The document then went on to provide examples of the kinds of people you might meet at each of the 16 landmarks that I had planned so far. At every landmark, at least one of the people would be a fellow emigrant taking the Oregon Trail westward. However, not all of the emigrants were actually heading to Oregon. On the first half of the trail, as far as Fort Bridger, you would also meet emigrants who were heading to Utah or to California. The conversations with these people would reveal a range of motivations for the westward journey, as well as differences in the mode of travel and other details. At all of the forts you could meet local traders, some of whom did a lot of traveling back and forth in order to acquire the goods to offer for sale. Local traders also tended to be quite familiar with the details of the trail and terrain in the immediate vicinity of the fort. Traders and emigrants would sometimes share rumors that they had heard. At Fort Kearney, which was a U.S. government operation, you could also meet soldiers. At many of the landmarks you could meet local Native Americans, some of whom would be quite interested in conducting trade. At some of the more western landmarks, you could meet fur trappers. At the rivers that had ferries, you could talk to the ferry operators. And at Whitman Mission, you could talk to the missionaries there.
This feature made it into the final product, almost exactly as described in this early document. However, unlike Matt in the general store, we don’t show a picture of the person that you are talking to. Unfortunately, the limited disk space made it impractical to include an illustration for each of the several dozen people that you can talk to along the trail.
In early December 1984, I prepared the first draft of the conversations at the landmarks. By this point I was calling these interactions “monologues”, in contrast to the more elaborate character interactions that I still hoped to include elsewhere in the product. Furthermore, after I discussed with John Krenz the best method for storing and retrieving the monologues, we decided to limit each one to a maximum of 255 characters. Based on the historical research I had been doing, and incorporating some bits of monologue text that Shirley provided, I composed and submitted the following initial set of 16 monologues – one for each landmark – with the idea that eventually there would be three different monologues for each landmark:
I attempted to accomplish several things with the above set of sample monologues:
I felt that this sample clearly accomplished the first three goals. However, it was still an open question as to whether goals #4 and #5 would be fully achieved.
According to my framework design, at certain landmarks the player must make a decision that is specifically tied to that landmark. (In one document, excerpted below, I said that all landmarks would have such a decision.) For example, at river crossings the player must decide how to cross the river, and at forts the player must decide whether to make any purchases. I called these decision points “regular events”, to distinguish them from the random events that would pop up between landmarks. On the real Oregon Trail, many of these decisions would have involved interacting with people. Therefore I figured that some of these decisions ought to be presented using human characters, if feasible:
In fact, my final design of the product did require the player to make important decisions at river crossings, forts, and major trail forks. Furthermore, the player does have the opportunity to seek additional advice at all of these places – but we were not able to personalize it as we had with Matt at the general store. No character appears on-screen when you make purchases at forts, and the ferry operators do not appear on-screen either. Again, it was an issue of priorities. It was very important to allow the player to make these decisions, but less important (although highly desirable) that these interactions be personalized with on-screen characters.
Of the three methods I outlined for meeting people along the trail – 1) talk to people at landmarks, 2) regular events at landmarks, and 3) random events between landmarks – I was especially eager to pursue the third category:
The key idea above is that when important but unpredictable events occur between landmarks, then sometimes you will have important decisions to make – and these decision points could be structured as conversations with people. The original OREGON had exactly one such decision point – without any conversation system – in which you meet “riders” that might or might not appear to be “hostile”, and then you had to decide whether to shoot at them or to take some other action. In my design I was determined to include a much richer and more balanced set of interactions with people. And while I wanted the majority of these interactions to be positive, I really liked the idea that some of the encounters along the trail might be rather tense.
In fact, in my early design concepts, I thought that human interactions triggered by random events might be the single most important aspect of the new product – even more important than hunting, river crossings, or resource management. I envisioned three principal categories of such interactions:
Furthermore, I assumed that many of the interactions between landmarks would be with Native Americans. In many cases, the mood and intentions of the Indians are not immediately clear. Therefore the beginnings of these interactions would often be tense.
All of these categories of interactions are explained in more detail below.
Early in the product cycle, in a document dated November 16, 1984, I presented a robust first-draft model for simulating encounters with Native Americans on the trail between landmarks. I specifically chose interactions that were realistic, and that were reasonably common on the actual Oregon Trail. In the first screen of such an interaction, you do not know yet whether the meeting is going to be friendly or unfriendly. All you know is that the people approaching you are Native Americans of some specific ethnicity:
(The notation “PSB” means “Press Space Bar to Continue”.)
After pressing the space bar, the player learns why the Indians have approached. For example, they may want to engage in trade – which would be a friendly encounter. On the other hand, they might demand payment for crossing their lands – a potentially tense situation. These two distinct kinds of encounters are summarized below (as “Branch A” and “Branch B”):
Branch A takes the player into the Trading Goods module, described later in this chapter. Branch B takes the player into the Toll Station module, described immediately below. (Keep in mind that these were very early concepts. Later in this book we will examine which ideas actually made it into the product.)
Right from the beginning of my research, the concept of tolls kept coming up. This concept took three forms:
The first two types of tolls are both associated with specific landmarks, and both of these were eventually included in the game. The third type of toll was unique, because it was not associated with any major landmark, and the overlanders never knew when they might encounter such a toll station. Therefore this type of toll was ideal for incorporating into the random event system.
On the real Oregon Trail, some overlanders refused to pay any tolls to Native Americans, and forced their way across the log bridges or other toll points without paying anything. This led to hard feelings. In one case, when overlanders tried to push aside a group of Pawnee who were guarding a log bridge, a small battle erupted, resulting in the deaths of nine Indians. No overlanders died in this battle. In some other less dramatic cases, the Indians retaliated for unpaid tolls by stealing livestock during the night. Therefore my idea was that the player would have the choice of whether or not to pay the toll – but if you did not pay the toll, then you might suffer certain consequences. For example, you might find the next morning that some of your livestock was missing.
In reality, the tolls for crossing Indian bridges were usually rather small, on the order of fifty cents. However, the prices fluctuated. In my first attempt to design this interaction, I decided that the Indians at a toll station would not actually announce the toll rate. Instead, they would simply ask you to give them something. Any reasonable amount would normally satisfy them – most of the time. But sometimes they might say “That’s not enough”, expecting you to give more.
However, even though toll stations of this sort would be a fascinating window into some real history, I was somewhat afraid of how this kind of interaction might be interpreted. Instead of seeing the demand for tolls as justified, some players might see it as a form of extortion – which is exactly how some of the overlanders saw it. In a school market product, this kind of ambiguity would be a perfect setup for a classroom discussion. But in a home market product, this kind of interaction was risky, potentially painting Native Americans in a negative light.
Due to budget, timeline, and disk space limitations, we eventually had to cut several modules from the project plan. The concept of Native American toll locations was a top candidate for removal, because of the sensitive nature of the topic – and therefore this module was cut from the plan fairly early in the project cycle.
In the original versions of OREGON, you could purchase goods for cash in Independence or at any of the forts. But you could not engage in any other sort of trading. In other words, the game did not allow you to trade clothing for food, or ammunition for miscellaneous supplies. But on the real Oregon Trail, trading was a very common experience. If you were an overlander traveling to Oregon, then you might trade with another overlander that you met. Or you might trade with people who were part-time or full-time residents in the local area – Native Americans and traders – who offered goods for trade but who often had little interest in cash. Furthermore, trading of this sort was seldom based on fixed prices. Instead, the people trading would haggle over a fair exchange rate that satisfied both parties. I very much wanted to include an activity that would reflect these aspects of real life – and would involve interacting with the game characters.
In my earliest visions for the trading module, I oscillated between the two different emphases – trading one type of good for another, or bargaining for a good price. For example, in one early document, I mentioned trading 200 bullets for 100 pounds of meat. But in another early document, I described a trading interaction that puts the emphasis on haggling for a good price, while paying in cash.
In December, after I had done more historical research, I had a new perspective on the nature of trade with Native Americans along the Oregon Trail:
However, it was not until February 1985 – as we moved into the “reality phase” of the project, that I finally fleshed out the trading concept – and I actually designed two distinct trading modules at that time. (See Chapter 17: “Restoring Abandoned Features”.)
In a document dated November 16, 1984, I described a model for a kind of interaction that I called a “hardship case”, which would present the player with a dilemma:
I really liked the idea that the player might encounter “hardship cases” along the way – that is, other overlanders who were in dire straits and needed help. This would be realistic and somewhat poignant. Helping these people would seldom help you in your quest to reach Oregon – in fact, it might hurt your chances of arriving safely. But, through the point system, you would be rewarded “in the end” – a subtle analogy to a heavenly reward.
“Good Samaritan Acts” were still part of the plan until March, 1985 – when I had to cut a lot of ideas that were no longer practical to include.
It was fairly common for a traveler on the Oregon Trail to become a victim of theft. Early in the game design process, I described my initial thoughts about how to include this in the product:
This approach solved two historical problems in the original game. First of all, most thefts were non-violent – they occurred quietly in the night while people were sleeping. Secondly, when armed robbery did occur, it seldom led to a gunfight. This was consistent with my decision not to include gunfights in the game.
As I continued my research, I became fascinated with the details about the frequent nighttime thievery. Needless to say, a successful thief would usually get away without being seen. Livestock were particularly vulnerable to theft. It is likely that much of the theft – but not all of it – was conducted by Native Americans. In some cases, it was a form of retaliation – in other words, the Indians felt that they should be compensated for the degradation to their hunting and grazing lands. In other cases the theft was the equivalent of a teenage prank. Young men would sneak into an emigrant camp and steal something in order to show off to their friends – to prove their bravery and skills. And indeed, it was quite dangerous to sneak into an armed camp of overlanders and steal something.
So I definitely wanted to include theft – especially theft during the night – as part of the game experience. But this raised a lot of questions. Would the player ever see or catch a thief, or would the thief always get away successfully? If you discovered that a serious theft (such as livestock) had occurred during the night, and you could see what direction the footprints went, then could you pursue the thieves? Would the game ever reveal the ethnicity of the thieves, and if so, then would it explain the motives of the thieves?
It was several months – and well into the reality phase of the project – before I had a complete set of answers to these questions. Ultimately, and by necessity, I decided to keep the theft interactions as simple as possible. A thief who steals things during the night always gets away. You cannot pursue the thief, and you never learn who the thief was.