Please let me know of any weird bugs or issues or ideas for improvement in my discord. Join Our Discord!
Every week hundreds of games release on Steam. We can determine rough estimates of how well they perform by using things like the review/sales ratio. This ratio changes over time, and changes based on how popular a game is, but we can use it to estimate approximately how well a game has sold (source [1]).
This website lets you predict how many reviews you think a game will have at the end of its first four weeks on sale and then measures and displays the result once those four weeks have passed. By gamifying it and simplifying the work necessary to keep track I hope to make it more fun and easier to keep an eye on the Steam releases and the state of the store! I'm also adding more review count analysis and (extremely) rough sales estimates that you can view for released games. Hopefully you'll be able to use what you learn here to inform your decisions and help you reach your game development goals.
Every week on Saturday the list of games to predict will be updated with the next week's releases (sometimes games that are delayed or have uncertain release dates will still show up, I'm working on filtering those out!). At any point in the week you can go over the list of games and predict how many reviews a game will get. You can predict what range the review count will fall into or predict the exact count (or both!) as well as record how certain you are about the prediction and any additional thoughts you may have. As soon as you click the radio button or out of the textbox your prediction is saved. This is currently a grabbag of options since this site is still under development and may change!
The median review count/game sale ratio is 30x, meaning for each review a game has sold approximately 30 copies. From the current publicly available statistics this ratio is generally lower for lower performing games (20x for games selling under 1k copies, source [1]) and higher for higher performing games. It's totally okay to start out having no clue how many reviews a game will sell, but you can also look at the results page to get a sense of how games perform even if you don't have any previous predictions yet.
You don't need an account to make predictions and have it evaluate your predictions four weeks later, but if you don't make an account it can only save your predictions on the computer you make them on. Making an account is totally free and also gives you access to other options and features so I recommend it! If you decide to make an account later it's able to import all your old predictions.
This site is a recreation (with permission) of the concepts behind the site Steam Prophet by Lars Ducet (explained here). I enjoyed using the website while it was up, although I didn't use it as much as I should have! Steam Prophet shut down because Steam made it harder to determine how many sales a game recieved, but I think it's still useful even if the estimations are less accurate. I would never have made this site without using and enjoying Steam Prophet so thank you!
Currently we're in alpha so the site may have bugs, the privacy policy and terms of service are uncertain and will change, and the prediction ranges are estimates of what will be vaguely useful!
I picked the ranges based on useful divides for very small indies but I may be adjusting them or make a way for you to adjust them. Currently 0-10 reviews means a game hasn't yet reached Steam's threshold for additional promotion, and presumably several of the reviews are by friends and family members so the review ratio is going to be less accurate. The next steps are based on rough estimates of United States minimum wage salaries. Here are my assumptions: Please let me know if I made a silly math error somewhere!
For many developers (myself included) I think it's useful to consider the development time necessary to create each of these games and how they stack up to the apparent sales they're receiving. This isn't to say give up on dreams of full time indie development, it's to help you determine the best way to accomplish them! Look at what genres are selling well, at what aesthetics are selling well, and try to project those trends into the future. Hopefully as this website gets more developed I'll be adding more graphs and tools to help you visualize trends in the market and better figure out ways for us all to be sucessful!