Dovetail Games on future-proofing simulators

Many retail titles are released with a fire-and-forget mentality. Granted, they may receive a Day One patch and subsequent updates to fix any bugs that emerge, but by and large the game has to be complete when it arrives on shelves.

The Train Simulator series is slightly different. While the game is complete to a point when it is made available for sale, the team at Dovetail Games – formerly known as RailSimulator.com – is constantly working on updates, new content and technology upgrades, much of which is made available to players for free.

It’s a different kind of business model, as chief technology office Tim Gatland explains. 

What specific technical challenges are involved with making simulation games like this? How do they differ from those of other genres?

There are two broad challenges we face. One is that our players desire highly realistic environments that closely match the real world, but trains move through the game world at high speed, so we have to be able to seamlessly stream highly detailed scenery. Very few game engines can do this on the scale that we need. The second challenge is not so specific to us, but our users demand highly realistic models. If a rivet is modelled in the wrong place, our players will certainly be quick to tell us.

A couple of differentiations are that we provide tools for others (and ourselves) to make additional content. These tools need to be powerful enough to create the high quality content that our players demand, while still being accessible to novice users. They also need to allow us to create content in a cost-effective manner.

We also recognise that our audience has a wide range of interests. Train fans are interested in specific types of trains or even individual models and so we have developed a business model that is based around offering a set of DLC that is designed to meet these different interests. We never expect players to buy all of our DLC, in the same way that no one would be expected to buy all of the trains in a model train store – they would just buy the locomotives that they wanted to collect.

Does the nature of the simulation affect these challenges? Are the expectations different? Is it trickier to develop a train simulator than a more mainstream title like Gran Turismo or Forza, for example?

It is a real challenge to provide a realistic experience without making it a tedious experience. We are very nervous of some of the detailed and unforgiving checklist following that is required for the extreme end of the flight simulation world – this approach can make some of these products virtually inaccessible to many players. Also, we have to work quite hard at identifying elements of gameplay drawn from the real world that will hold the players’ attention. 

Trains are not naturally used for racing and so we need to set challenges that are demanding, but which are still fundamentally based in the real world. Creating tension is quite hard.

I think that all genres of games present challenges, but we also have a number of problems that are probably remarkably similar. One that immediately comes to mind is the challenge of getting reference material; often the trains we want to model are still in operation and getting access to the material required to model the driving cab and get accurate performance data is a real challenge.

It is a real challenge to provide a realistic experience without making it a tedious experience. We have to work quite hard at identifying elements of gameplay drawn from the real world that will hold the players’ attention.

How do you balance between realism and accessibility/playability?

We have always tried to ensure that our locomotives behave and drive in a realistic way – our core players demand that, and rightly so. However we recognise that for many new or more casual players this can be daunting, so over the past six months we have spent a lot of time thinking about, designing and play-testing our tutorials. These are intended to get anyone, regardless of ability or experience, into the cab, driving and having fun quickly.

How do you lay the foundations for continuous updates when developing your titles? How can you predict what you’ll need to add or change in the future?
We have a rather unusual business model in that, using Steam, we provide a free update to our technology on an annual basis. We have been doing this for the last five years and we believe that this is a key differentiator from other games that charge for their annual iterations.

This approach means that all of our current customers are on the same technology base providing a common platform for all of our DLC. To manage the product’s technical evolution, we continually listen to our customers and look to combine their feedback with our own ideas. These ideas are then sifted and the ones that are practical and desirable are slotted into an annual release. 

Predicting the future is hard, but knowing that we have a yearly shot at improving the technology means that if something is too ambitious for this year, it can always be considered for next year.

What roadmap do you work to? How many years ahead do you plan for?

For our trains business, we usually operate to a two to three-year plan for the core technology and a one-year plan for content such as locomotives and routes.

How can you make additions and changes without altering the balance of the original game?
Because we have a large stock of DLC to sell and then support, we need to be careful not to break any content when we issue technical updates. This need for backwards compatibility could restrict our ability to improve the product, so we embed ‘version number’ elements in some of the content so that legacy content will continue to work alongside updated software. Because our game is rooted in the ‘real’ world, maintaining the realism is key.

Where do you draw the line between bundling everything into the original game and holding too much back for future DLC?

Because we offer free core technology upgrades to our current users we view our annual product release – which offers the technology, plus three routes and a number of locomotives – primarily as an entry point for new players. It’s not a matter of holding content or features back, it’s more about offering the best suite of content that we can at a single point in time and then upgrading with new features and offering a wide range of DLC that will allow players to customise their experience and build their digital hobby.

Predicting the future is hard, but knowing that we have a yearly shot at improving the technology means that if something is too ambitious for this year, it can always be considered for next year.

What are the advantages of a model based on continuous updates? What are the disadvantages?

Our model helps us to manage our business because it keeps our community on the same technology and it means that from a compatibility point of view we are only testing our DLC against one piece of core software. From a consumer point of view, it means that I can buy into Train Simulator knowing that the developers are committed to making improvements to the experience over time. 

Is there a danger you are alienating those who have lower spec PCs, slower connections or no connection at all? 
We keep a careful eye on the range of hardware our players are using and we try to avoid making changes that will leave players behind. Although we don’t employ an ‘always-online’ policy when playing the game itself, players are required to register with Steam when they first install the game. This allows them access to our extensive range of DLC, which in many ways sits at the heart of the hobby.