Tuesday, September 11, 2012

Open letter to the players of Railroad Empire

I have thoroughly enjoyed working on Railroad Empire for the past 2 years. We launched Railroad Empire to the public two years ago. There have been some ups and downs as I have worked to try and make Railroad Empire better, but unfortunately Railroad Empire is costing me a lot of money and I am losing money every month. It is not fair for my family to continue to bear this burden, so I am writing this letter.

Right now unless we change something I will have to shutdown Railroad Empire on September 30th. I don't want to do this, but right now the game brings in an average of $260 a month and the total cost right now is about $560 a month. I am currently paying the extra $300 a month out of my own pocket.

Many of you may be wondering how I ended up in this state. I attribute it to two mistakes on my part. The first is that when I designed the game I chose to use the virtual currency monetization method. At the beginning of the game's history this proved to be a very effective method of monetizing the game, however as the percentage of users changed from new users to veteran players the revenues dropped off and the cost increased slightly as veteran players have more trains which costs more to run the game.

The second potential mistake I made is that I chose to build Railroad Empire on Google App Engine. Last year Google made a major change to its billing practices for App Engine. At the time these changes happened, I was hoping I would be able to work through the increases and control costs, however I have been unable to reduce the costs on Google App Engine. A while back I started on a project to change the backend for the game from Google App Engine to a traditional server architecture. However I substantially underestimated the amount of work that would be involved in the project and I also have realized that it may not have a substantial impact on the cost. In addition there are some features of Google App Engine that would be lost in the transition. There are also some issues with user accounts from users logging in through Google, Yahoo, and other OpenID providers.

Right now the game gets 7863 unique players a month. So if people are interested we can explore moving the game to a subscription model. This creates a potential issue in that Google Maps requires an app that is not freely available to pay for a Premier license which is $17500 a year. We may be able to switch the map software to something else but I am not sure how difficult that will be.

I would like to hear everyone's opinion on this matter. If anyone has any ideas, please feel free to let me know what you think. If you want to contact me privately, feel free to e-mail me at mjr@mytechmaster.com


  1. This comment has been removed by the author.

  2. Cost for conventional architecture all depends on what you want available. I don't know what you're requirements are, but some sites like Dreamhost do shared hosting for a standard low cost, and their private server rate is quite reasonable.

    I've worked on various projects in the past where shared hosting was used, and other then a small security vulnerability if your host isn't on the up-and-up, it's pretty good. Far better then it was 3 years ago anyway.

    I can think of a couple ways you could get the same functionality with java, but it's just as you said - It wouldn't be a quick fix, even if it did save cost. The bottom line is, if you still want the game to be a reality, there are cheaper ways to take it - My last project cost the host a total of 400$, domain costs and extras included for a year on his end. I've run a personal page on there for the last five years for about $117 a year. Just something to think about.

    Another thing to think about is that some of your players may know a thing or two about coding, and as silly as it sounds some of the best ideas can come from your playerbase. A great example is Kingdom of Loathing, where official devster CDMoyer was picked out of the group.

    Anyways, I hope some of this helps. it always sucks when a project goes south, and I wish you the best of luck! Keep trucking, and I'm sure you'll figure it out!

  3. Why exactly are you using the google app engine again? Is it just for the google map?

  4. When I started the project I chose Google App Engine because it has some nice features like a built in queue system and not having to worry about the scaling of the application. Unfortunately they changed their pricing model significantly on me and a bigger problem is that the App Engine Datastore is not relational and the games data is surprisingly relational, and a traditional database would have worked better. At the time they did not have MySQL as an option. Now that all the code is written and working, it is a major project to rewrite the code to get off the App Engine Datastore.

  5. Hey mike, I have just sent you an email.

  6. Wow, Mike, enjoyed this game a while back and while I don't really play it now, I check in from time to time to see how it and you are doing. You've put so much effort and work into it and it saddens me that it's starting to flounder. Any thoughts on how many people are willing to subscribe and how long they will subscribe for?

    When do people typically stop playing? After all stations are built out? Just kinda curious :-)

    1. As far as how many will subscribe and for how long? I don't know.

      The average is a couple of months. The extremes are 1 login and never again to playing since the game launched.