One of the longest running mysteries around festivals like Steam Next Fest was whether people deliberately play games and carefully evaluate each or do they just spam the “wishlist” button as they browse? For instance Valve has made it exceedingly easy to wishlist games: they added that blue button with the star. You don’t even have to visit the game’s steam page to wishlist it. Does this inflate wishlist counts?
On the other side, playable demos. Do people play demos to determine whether they should wishlist? Does playing the demo mean they don’t wishlist.
For a long time we couldn’t figure out this data because the way Valve reported Demo plays, and demo downloads was kind of unclear and it was really tricky to isolate it.
But at the end of the June 2024 Steam Next Fest, Valve added a “recap” widget to the calendar widget on the Steamworks dashboard which reported some very basic but important wishlists. It looks like this:
These values are defined as:
Demo Plays: Number of unique players that played your demo during the event.
Total Wishlists: Number of new wishlists your game received during the event
Played & Wishlisted: Includes games that players both played and wishlisted during the event. This is inclusive; it’s counted here in addition to being counted in Total Players and Total Wishlists.
In my interpretation is Valve’s new chart is reporting this:
Because of this data, we now have a much clearer view on how Steam players typically behave during a Steam Next Fest.
So I put out a survey asking for those 3 numbers and 40 developers shared their data.
The average player behavior
If I average the survey results for all 40 games across those 3 numbers here is what we get:
In general most people wishlist at almost the same rate that they play the games. A much smaller proportion of people will play a game and then go on and wishlist it.
On average 20% of people who play a game demo will go on to wishlist it. Here is a chart of all 40 games and the percentage of people who played and wishlisted:
Each game is one blue bar. Note that the games are sorted based on TOTAL wishlists earned. So the big, most wishlisted games are on the left. The games that earned the fewest wishlists are on the right.
I wondered if there was a correlation between the top earning, most popular games and the wishlist / play percentage and I got a p-value of around 0.23, which suggests that this correlation is not statistically significant.
So no, the most popular games don’t have a higher wishlist percentage during play.
Here is the % wishlist after playing as a histogram.
So in general, about 10%-20% of people will play a demo and decide to wishlist it. Furthermore it doesn’t seem like there are some super-star games that retain at a very very high rate. It always seems to be around 10-20%. Strangely this number is very similar to the wishlist conversion rate. Maybe at any given time there are 10-20% of people who are ready and looking for a game.
Demo plays
I have always wondered, if you see a ton of demo plays and lower wishlists does that mean the game is bad?
Based on my survey, 16 games (40%) had more demo plays than wishlists. Many of the top wishlisted games had more plays than wishlists. And they did fine and I expect them to sell just fine.
This graph shows what happens if you subtract the number of demo plays from wishlists. If the game had more demo plays than wishlists it is a negative number. Each bar represents a game and the most wishlisted games are on the left of the chart.
So no, I don’t think you can draw much from whether or not you had more plays than wishlists or vice-versa.
Wishlisting
Here is a graph of all the games in my survey and how many total wishlists they earned. Each bar represents a different game. They are sorted from most wishlisted on the left, least on the right.
This is a typical graph for games on Steam whether it is wishlists or dollars earned. The front-page-featured games get all the wishlists, then there is a steep dropoff to the games that go featured on the sub-genre-pages, then there is a second dropoff for games that got no featuring.
There are two big ways to get featured
Option 1) You go into Steam Next Fest with at least 170,000 wishlists (by my estimation looking at the follower count of the #10 most-wishlisted game.
Option 2) You have a big wishlist boost in the week or days before the event. I wrote about how I Am Your Beast did this by announcing their game during the PC Gamer Show.
Regarding what is typical for wishlists. Here are the result of the games in my survey
| Game percentile | Number of Wishlists earned |
| 30th | 314 |
| Median | 618 |
| 70th | 1490 |
So what did I learn?
I once went to PAX and followed a couple regular gamers around an expo floor (with their consent) to see what they played and were interested in. And to be honest, it was all quite random. It was as if they were just wandering an old-school coin-op arcade stopping in front of whatever game in the moment looked interesting to them.
These PAX gamers often played games in genres they didn’t like but they figured the stakes were low so they might as well give it a play for a few minutes.
I suspect players at virtual conferences do something similar.
I often see that a games’s median-playtime drop a bit after being in a Steam festival. This again leads to my guess that players are giving a wild selection of games a quick 1-5 minute shot and if the game hooks them they stay and wishlist it.
I played a lot of the top-50 demos in the Summer Next Fest. I noticed a lot of them didn’t even have tutorials. Or it was just a popup dialog that listed the controls. I think this is the best way to create the demo. You have to assume you have players’ attention for minutes. You don’t want to bog them down in long tutorials.