1999/2000 - Developed by Bullfrog Productions; Produced by Electronic Arts

Dungeon Keeper is considered by many as one of the best real-time strategy games ever made and quickly gained a huge fanbase after its release in 1997. The follow-up, however, did not do as well as expected in the marketplace. When the decision was made to go on with the series, the developers were instructed to make Dungeon Keeper 3 more accessible so that the game would reach a wider audience. With a small team of only three people, the project entered the conceptual stage in November of 1999. Roughly four months later, Electronic Arts pulled the plug in favour of games that promised more profit and were less risky, but they did not let the public know about the cancellation until August 1, 2000.

Source credits: Ernest W. Adams

Written by Ernest W. Adams for pcGTW, August 2006

In August of 1999, I was brought from EA headquarters in Redwood City, California to Bullfrog Productions in Guildford, England, to be the lead designer on the next game in the Populous series. (EA had already purchased Bullfrog at that time, and Peter Molyneux had left to found Lionhead.) Populous: The Beginning had been something of a disappointment (financially) and a Bullfrog team had worked on a sequel for a while without getting very far. I came to start over and provide some new blood. We called the project Genesis: The Hand of God, and did some great prototype work on it for several months.

Unfortunately, Lionhead was hard at work on Black and White at the time, which EA was going to publish. The Bullfrog management became concerned that we were too close to Black and White. I disputed that on various creative grounds, but marketing people see things differently. The upshot was that Genesis got cancelled. They looked around for something else to do with me, and decided on Dungeon Keeper 3. There were only three of us on the project: Nick Goldsworthy, the producer; me, the lead designer; and Nick Ricks, a game designer working for me. At that point I thought, well, this is a franchise that the company has already turned into two products, they’ll be behind this. However, we also knew that Keeper 2 hadn’t performed as well as hoped in the marketplace, so we were under instructions to make the new version more accessible.

We started work around November of 1999 with just the three of us – no programmers or artists – doing high-level concept design. We had a number of key changes in mind. First, in an effort to remove some of the ambiguity about the nature of the game (which would have comforted marketing and the retailers), Keeper 3 would have been more of a straightforward RTS. Not merely a copy of all the others, though; we were still keeping portals and the need to nurture unique individual creatures, as well as retaining the imps and chickens. Another huge change, which I was very much looking forward to, was a move above ground so that instead of building a dungeon by digging you built a castle by construction – not unlike the later Stronghold games. This required some thinking, as an underground dungeon is effectively safe and hidden until the walls are breached, while a castle is just the opposite: it’s both visible and vulnerable until the walls are completed. We didn’t get far enough to resolve this question; it was just something we knew we would have to address.

The third major change we had in mind was to add a new race, and to let you play any of the three: Heroes, Dungeon-Dwellers, or the third race, the Elders. The Dungeon-Dwellers we expected to be very similar to the familiar ones from the earlier games. Their castle would look black and evil, and all the land around it would start to decay and become vile. The Heroes we decided to make very clean and organized – their castles would be white stone and beautiful, and the landscape under their control very orderly and neat, rather like Switzerland. The new race, the Elders, would have represented the spirit of wilderness, neither good nor evil, just wild and untamed. Their castle would have looked very organic, formed of trees and hills, and the land all overgrown with forests and vines. One item on the task list was to devise equivalent creatures to the Horned Reaper for each of these other races.

We also planned to make the economy of the game a little more interesting by including peasant farmers whom you would try to control and exploit, rather than simply digging out gold. We hoped to make it more a game of growth and evolution, moving a little closer to Age of Empires (in economic terms) than simply a question of mass production.

One of our major goals was to improve multiplayer play. Earlier versions of Dungeon Keeper were never very good in multiplayer modes, because the underground layouts were rigidly symmetrical, and both sides always had exactly the same kinds of creatures to play with. It really became a race for resources more than anything else. We wanted to make a more asymmetric game in which players could choose to play any of the races, each of which would have its own strengths and weaknesses. I considered StarCraft a major inspiration, and played a lot of it in those days.

At some point in the project, Nick Goldsworthy and I swapped roles – he had been producer on Dungeon Keeper 2 and wanted to do more design. I was willing to give that a try. However, the difference was minor – we were all working on design.

We got a fair bit of work done, thinking about typical RTS-like issues – supply lines, siege engines that needed creatures to work them, creating unique individuals by mesh warping, interfaces for managing mixed indoor and outdoor combat (the dungeon heart would still have been inside the castle), and so on. However, it all became moot about March of 2000. Electronic Arts foresaw bad times ahead because of the downturn in game sales that preceded the arrival of the PlayStation 2. This was exactly analogous to the slump the industry is in now as it awaits the PlayStation 3 – consumers are waiting for the new machine, and not buying new games for the old one. The company decided to retrench and avoid any risky projects, especially PC projects, which don’t make as much money as console ones do. At the same time – unbeknownst to us – they were negotiating with J.K. Rowling for the rights to Harry Potter, and with New Line Cinema for the rights to The Lord of the Rings. Given the choice between an experimental DK3 and the absolute license-to-print-money that were Harry Potter and The Lord of the Rings, it wasn’t a difficult decision. Dungeon Keeper 3 was cancelled, and we were moved on to other things. In addition, the company began laying off staff because of the financial losses they anticipated (and indeed eventually suffered – their first money-losing quarter in ten years happened because of the downturn between consoles). Finally, they did away with the Bullfrog brand entirely. Harry Potter and Lord of the Rings weren’t going to be Bullfrog-brand games, but they were being made by Bullfrog people, so in effect there was no more Bullfrog.

So DK3 existed as an ongoing project between about November of 1999 and March of 2000. I have no idea who set up the website – there was never much on it, and neither I nor any of my team was involved. It was never more than a teaser, really. [...] That’s about all there is to say. I understand the financial motives for killing it, but I regret it all the same – it would have been fun to work on and, I believe, fun to play as well!

The concept art below was created by Nick Ricks for Electronic Arts.

» more concept art