Series Info...Trials, Triumphs & Trivialities #181:

Anatomy of a Game: The Amazing Race

by Shannon Appelcline

One of the interesting aspects of our time is for us game designers is being able to see game designs writ large on the television screen. I've written in the past about Survivor (TT&T #74), Big Brother (TT&T #133), and, tracking the evolution of the game, Survivor again (TT&T #135). Now, thanks to the magic of DVDs and Netflix, I've gotten to dip into another popular reality show: The Amazing Race.

This week I'd like to offer an overview of that TV game show, based on seasons 1 & 7, examine their game design, examine the evolution of the game, and see what lessons that offers to us in the computer gaming world.

A Competitive Game

Unlike Survivor and Big Brother, The Amazing Race is not a voting game. Instead, it offers up the other gaming element that's very common in reality-based TV shows: straight-up competition.

The Amazing Race is, at its core, a race around the world between eleven teams of two people each. It's divided into a number of legs. During each leg the players start from a pit stop, visit a number of other specified points, engage in one or two specific tasks, then arrive at another pit stop. In some cases the mode of transport between specific points is noted (e.g., "take a cab" or "take a train") while in others its freeform. At each pit stop, the last-place team to arrive there tends to be eliminated (with some exceptions, as noted below).

The specific tasks are sometimes cultural and sometimes geographical. Examples include milking goats, searching a cave for a clue, rowing a boat down a stream, and climbing castle walls. They're broadly divided into two types: "road blocks" (where one team member must decide to do the task) and "detours" (where a team decides between two tasks).

More strategic elements introduced into the game include "fast forwards" and "yields", each of which I'll discuss.

Managing a Real-Time Game

One of the most interesting elements of The Amazing Race is that it's a true real-time game. Players continuously move forward, heading on to the next "clue" as soon as they discover the last one. There's only exception, the pit stops. At these locations each team has a mandatory 12-hour (or more) stop--enough time to rest and relax and give the production team time to set up the next day's worth of clues and make sure that they're staying ahead of the teams. However each team sets off from the pitstop exactly 12 hours (or more) after they left, so exact ordering and leads are maintained throughout the whole race.

On the face this offers the opportunity for disaster, as teams could end up spread across the entire world, destroying any potential drama and requiring a much larger production crew with much better communication. Instead The Amazing Race carefully manages its course in such a way as to keep the players together without looking grossly unfair in doing so.

This is done through what I call "funnel points", which I've seen other folks refer to as "equalizers". These tend to be locations that are only open at certain times of the day, or required services which are only available at certain times of the day. Most common is a store, building, or resort which only has daytime hours. If players arrive at night (and the course is clearly designed to generally make this happen), then everyone queues up at the funnel and will leave it at about the same time. Airplane flights often provide a similar service in The Amazing Race since there tend to be several hour gaps between them, allowing teams to catch up.

Overall these funnels are a very good design for a real-time race. They often immediately follow pit stops, starting players out approximately equal at the start of an episode. There's still a big impetus to get to that pit stop as fast as you can (since first-place players sometimes win a prize, at least in later seasons of The Amazing Race, and last-place players are often eliminated). But after this set of rewards and punishments, everyone is then evened out.

There are some problems with this design too.

First of all, the funnels sometimes seem unfair, because players aren't just brought closer together, but are usually totally equalized. A few funnels which involved a one-team-at-a-time task right afterward seemed fairer, because you could close up gaps in a real-time race, but still maintain a fair order afterward.

Second of all, the funnels have the potential of totally putting a team out of contention. In the first Amazing Race one of the teams entirely missed the timeframe for a funnel, and so ended up a full day behind everyone else. It was late enough in the race that we had to watch them stumble a full day back through the end of the show. This sort of disaster is exactly the type of thing that you have to be careful of in real-time gaming (and why The Amazing Race is quite impressive in its design for usually avoiding it).

Managing a Made-For-Prime-Time Game

You need to design any game for its audience. For most of us that means entertaining the players who directly interact with our offering. For The Amazing Race, however, that instead means entertaining the viewers who will eventually watch the competition. This may not immediately apply to the world of computer games, but I'm sure we're just a year or two away from a computer game opening itself up to a secondary viewership in a pretty major way.

Most notably, The Amazing Race needs to fill out a half-season of television, which usually means 11-14 episodes. However at the same time they need to keep their number of contestants to a reasonable number, because too many would make early episodes too long, as they chronicled everyone's story, and also confuse the viewers. Most reality TV shows have settled on 13-18 players as a fair starting number. The Amazing Race had to offer a slightly different compromise because they have teams of at least two players, so instead they've tended to have 11 teams of 2 players each, for 11 or 22 players, depending on how you count it.

The goal in The Amazing Race is to eliminate one team each week, then have a finale starring the last three teams racing for victory. Unfortunately that only sums to 9 episodes, so from the first Amazing Race there were a few "non-elimination" pit stops. This was a clear compromise required by the medium which The Amazing Race was broadcast on, but it's the same sort of compromise that we all eventually have to make in bringing our games to market.

In the first season these non-elimination pit stops totally killed the tension. There were no first-place prizes in that race, there were clearly no eliminations in those episodes, and funnels on the other side usually put everyone right back into the race. It was a week of wasted TV. (The credits for the non-elimination rounds also totally spoiled the show by saying, "Who will come in last?" rather than "Who will be eliminated?" By the seventh season they'd change their patter to instead always say, "One team may be eliminated.")

In season 7 the producers of The Amazing Race tried something different: they dramatically penalized the last-place finishers. They took away any money they'd saved from previous legs of the race and all of their supplies. They also didn't give them money for the next leg of the race. (Usually the players get a small amount of cash, one or more hundred dollars, for each leg of the race to pay for cabs, entrance fees, food, lodging, and other necessities. Airplane tickets are the one thing that the show always covers.)

Having no money and nothing else but the clothes on their backs and their passport seemed to me like an absolute recipe for disaster. But it worked out. Every one of the three teams penalized during non-elimination rounds in season 7 managed to survive the next leg where they started with no money. (They begged from other players and from random people on the streets to good effect.) At the same time they were clearly at a severe disadvantage, and you never wanted your favorite team to end up in that position, even when you were pretty sure that you were watching a non-elimination week of television.

I've written before that game design is a constant iterative process. You see a problem, you solve it, then you create a smaller problem which you must solve. The penalty non-elimination rounds were an excellent answer to the earlier problem of non-dramatic weeks of television.

The Evolution of Fast Forwards

Moreso than most of the other reality TV shows, The Amazing Race contains actual strategy: thoughtful decisions which can impact your gameplay now and in the future. One of the strategic elements that was present from season one was the "Fast Forward".

The idea here is that you can choose to follow the fast forward clue instead of the main clue path. For the fast forward you just have to complete one task or reach one destination, and you're then fast forwarded to the Pit Stop, immediately ending that leg of the race for you. This could easily give a one to several hour jump on the other teams, but more importantly can help a last-place team climb out of last. The catch: only the team that finishes a fast forward first gets to complete it; everyone else must return to the regular race course. As an additional catch: each team can only fast forward once during the race.

However in the first season of The Amazing Race the fast forwards were largely a failure. They were available on every race leg, and each time (with one exception) one last-place team inevitably did the fast forward, and only one team. The producers were clearly hoping for some tension as several teams strived for the fast forward, one of whom was rewarded with a multi-hour jump, and the others of which would then have wasted all their time and be even further back. But instead there was only one leg where there was actually competition because the fast forward opportunities were so spread out.

By season seven the producers were using a dramatically different tactic: there were only a few fast forwards in the entire race. Since the teams didn't want to "waste" their opportunities to get ahead, there was a bit more jostling to see who'd get the fast forward. In one tense situation we saw one team waiting outside a fast forward task, only to learn that the team who had beat them there had successfully completed it. In another we saw a team lose their nerve at the last minute because they were sure that another team was going to the fast forward ahead of them. (They were right, but the other team got lost a few times on the way.)

Inevitably players are competing for resources in most games, and by limiting those resources the producers of The Amazing Race created a much tighter and more tense game.

Other Strategic Elements

Various other strategic elements in The Amazing Race are interesting for the game design ideas that they highlight. These include: alliances, road blocks, and yields.

Alliances. There is no official mechanism for "alliances" in The Amazing Race. There wasn't in Survivor either, but as soon as Richard Hatch realized the power of an alliance in a voting game they became a required part of Survivor gameplay.

The impetus isn't as great in The Amazing Race because you don't asbsolutely need the help of other contestants to win. Despite that, some small and fluid alliances developed in both seasons 1 and 7 of the show. The strength of an alliance is that 4 (or 6 or 8) heads are better than two. You can solve clues faster and find locations quicker. The lack of access to cell phones severely limits these possibilities, but they still exist. In season 7 we saw several different teams hurt because they didn't find an early airplane flight, and conversely we saw a few teams brought along by their allies who had discovered the flight.

The flipside is that the game is ultimately a singular race, and so you'll need to eventually cut out your allies. As a result alliances only last until you and your allies are fighting for last place.

I've been surprised that alliances haven't been a stronger influence in at least the two seasons of The Amazing Race that I've seen. Nonetheless they're an important pointer to the types of player culture that will develop in a game whether you like it or not.

Road Blocks. From the first season of The Amazing Race we've had road blocks, which are tasks that only one member of a team can do. In season one anyone could do each roadblock, but by season seven there was a limit: each team member could only do 6--about half--of them.

As far as I could tell the teams didn't use this very strategically. By the end several teams were sending up their less qualified member to do tasks because they had to. (Fortunately the teams were more uniformly strong by that point.) However there is a strategic basis that I suspect some of the teams missed: if they'd sent weaker team members into road blocks where they were already well ahead, it might cost them a second or a third place, but not the game. They'd probably catch up at the next funnel, then have their strength available for later use.

As game designers, the new road block rule offers an interesting example of how balance can evolve in a game. Clearly the producers felt that it was "unfair" that a single strong member could carry their team, and so they began to reward teams that were more uniformly strong by season seven.

Yields. Yields were introduced somewhere between seasons 1 and 7 of the Amazing Race. There were just a few of these locations on the race, and you could use one to hold up any team behind you for an hour. But you could only yield once during the race.

In season 7 the yields never got used well. No one yielded on the first two yields, doubtless because they wanted to save them. On the third yield the lead team did yield, but did it in such a way that I think it cost them the game. They had three teams behind them: one strong team who was a little bit behind them and a weak and a strong team who were considerably behind them. They even knew the approximate locations of all the teams because they'd caught a flight with their close runner and they knew the other two teams had caught a later flight. (On one of the earlier yields, it was very hard for players to assess who was behind them, dramatically limiting the usage of the yield.)

The lead team choose to yield their strong close-runner. This assured them of victory in that leg of the race, and they won a nice entertainment system as a result. In the rear of the race the strong team beat out the weak team, leaving the three strong teams in the final two legs of the race.

If the leaders had conversely yielded the back-running strong team, they would have fully knocked them out of the race, keeping the weaker team in to the end, and upping their chance of winning from 33% to closer to 50%.

This all tells me that sometimes no matter how strategic of elements that you offer in a game, players will instead make knee-jerk choices. I think this happens more in a real-time high-pressure game like The Amazing Race, but it can happen anywhere, and ultimately as designers you need to decide whether that's a good thing or not. Do you offer players the opportunity, accepting that they'll waste them? But kudos to The Amazing Race for adding more strategic elements as the seasons have gone on.


The Amazing Race, yet another reality TV show, offers up some interesting game design ideas for our medium as well. In particular I find it interesting to look at early seasons and later seasons and see how the show has evolved, because that sort of fast-forwarded evolution is exactly the sort of thing we need to mimic in online, always-on games.

I've written a number of other times on the television medium, and how its lessons learned can inform our own:

We'll be back to collective choice in the next column; Christopher & I are working on a piece about reputation systems.

[ <— #180: Collective Choice: Competitive Rankings | #182: Anatomy of a Game: The MTV Challenges —> ]

Miniseries ...