Statistics and Fun Facts About the Players of Don’t Feed the Trolls

Two weeks after a post about the sales and the ads revenue of Don’t Feed the Trolls, here’s another statistics post, this time about the players of the game. Don’t Feed the Trolls is a reflex game released in May on iPhone/iPad and Android (Google Play and Amazon). Feel free to try it, it’s a free download.

The data

The data used for these statistics are mostly coming from the online leaderboards of the game. I only used information from players that spent at least five minutes playing the game. The data is taken from the release up until July 30th and is made of 3,991 players. Depending on the charts, I sometimes use a subset of players to get more relevant information.

Play Time

First chart is showing repartition of players based on their play time. It’s actual time spent playing the core gameplay, not the time spent in the main menu or paused. The chart might be slightly biased by recent players that got the game during the last few days, and didn’t have the time to play that much.
Fun Fact #1 : One player played 31 hours (and he’s still active)
Fun Fact #2 : Total cumulated play time since release is 107 days and 14 hours


Since the release of the game, I made 5 updates on Google Play, 1 on Amazon and 3 on iOS. I was looking to know whether active players are updating their version. To make this chart, I ignore Amazon Players and use the 1.2 update of the game as a reference. This update was the biggest update of the game, and was released simultaneously on Google Play and App Store on July, 5th. I then checked the players that first run the game when this update was not available and are still playing one week after the update. It narrows down the set to 804 players but it’s much more relevant. The chart shows the percentage of players that updated to the last version.
I find this chart is very interesting. Please remember we are talking about active players. I don’t find surprising that Android players are updating more than iOS players but having only 38% of iOS players updating the game is very low. Especially since the 1.2 update is a content update. Knowing that, it’s important to get your game right on release.
Fun Fact #3 : The Top 20 players of the classic mode are all up-to-date

Time of Day

The following chart shows in blue the time of the day where the player launched the game for the first time (while being connected). Please note that more than 70% of my users are from France (GMT+2) and the chart is showing GMT times.
The red curve shows the last time players played the game before they definitely stop playing it. For this curve, I consider that a player will not play the game anymore after 14 days of inactivity.
Quite common chart here too, players are mostly playing in the evening, and there are less players in the middle of the night. Unfortunately, the red curve does not give us any useful information as it’s similar to the blue one. I hoped to see that more players would stop playing the game if they launch it at a specific time for instance but that didn’t happen ;).
Fun Fact #4 : The record is 126 players launching the game for the first time on June, 27th between 18:00 and 19:00 GMT.


The next chart is probably more interesting if you know the game a little. It shows the percentage of players that got each achievement of the game. It is taken from the players owning the version 1.2 of the game, which contains 25 achievements (click the image for full size). This is a set of 899 players.
Seeing these statistics, I don’t think there are many players trying to get all the achievements. The easiest achievement to get is not tied to the gameplay (“visit”) and is only obtained by 3.78% of players. Players probably just play the game and unlock achievements as they play. Please note that a few achievements can’t be done without paying for the full version.
Fun Fact #5 : The achievements statistics are extremely similar between iOS and Android. I thought I’d see more completion from Android players.
Fun Fact #6 : Only one player has all the achievements.
Fun Fact #7 : 95.8% of players have 10 achievements or less

Pausing the game

This stat comes from Flurry Analytics and not from my leaderboards. Here is an interesting pie chart showing why does the player pause the game.
I didn’t think that many players would pause and retry the level that much. I knew Don’t Feed the Trolls is a lot about scoring but it still surprises me. The stats are similar on iOS and Android. Also, note the the tutorial is automatically showed to the player on its first play. Someone clicking on the tutorial button wants to see it again.

End of Level

Similar to the previous one, this time showing the option selected by the player at the end of a level, completed successfully or not.
The skip option is quite used, even though if the player didn’t succeed this level, he will have a tough time completing the next one.

Accuracy of data… and China

A final note on the data reporting tools. I have three different ways of evaluating the numbers of players:

  • My leaderboards: only taking into account online players completing at least one level
  • Google Play / App Store : it gives me the number of downloads. Someone downloading the game without playing it is shown here
  • Flurry Analytics : there is a “New Users” statistics in Flurry. The user has to be online, too

Now, let’s compare these numbers. The number into parenthesis is the ratio of players compared to the number given by the store (that is theorically exact):

Android iOS Amazon
Leaderboards 5,912 (92.6%) 2,269 (59.3%) 610 (66.7%)
Store 6,385 3,827 915
Flurry 10,579 (165.7%) 3,382 (88.4%) 741 (80.1%)

You probably noticed the strange value in there: on Android, Flurry reports more players than Google Play does. I guess you found out why: piracy. Google Play is not available in China, but according to Flurry, Don’t Feed the Trolls was played by 4,123 unique players in China. The game is monetizing through in-app purchases to unlock the full version, and Chinese players cannot buy the full version. However, I don’t know if they only play the free version or have a crack for the full version. Also, do they have ads displayed in the free version? Last question, why are the leaderboards numbers not showing a similar value? I though about the “Great Firewall of China“, even though the tests I made showed that my domain is not blocked in China…

Anyway, unlike Matt Gemmell (interesting article, though) I don’t blame piracy (anymore) and I’ve been pirated much more than that in the past (see Piracy of Spring Up Harmony: 96%). I honestly don’t think a pirate would have bought the game if he couldn’t pirate it. Anyway, it’s another topic.


Fun Fact #8 : The Top 10 players of classic mode are on Android (but note that 72% of total players are on Android)
Fun Fact #9 : 10% of the players fed 62% of the bears
That’s it for the statistics of Don’t Feed the Trolls. As usual, comments, shares and tweets are welcome!


  1. Great! But the question is: the player who has played over 30 hours is it the same one that has all the achievements? 🙂

  2. Hi Frozax , i was reading your blog and wants to congrats your work , im recently start to work with games (and yes , i know thats no easy). Im working in my first app game call MadSquirrel. I just want to ask some questions about your adventure with cocos2d-x and the integration with other platforms.
    Actually OpenFeint is turning in Gree , im having a nightmare to integrate cocos2dx with this stuff , now in IOS its ok , but android not works very well.
    Do you use some other platform to take all these statistics? (flurry ?)
    about the achievement screen , do you use some kind of persist data in your game or every time you search from openfeint server?

    Thanks for share these statistics and some cocos2dx code , they are very good!
    visit my website and take a look in the graphics of my game.

    1. Hi,
      I’m using GameCenter on iOS but I’m not using OpenFeint. When I tried Open Feint on my tablet in other games, I found the user experience quite bad and decided to not use it. Leaderboards are custom code with a php server. I do use flurry to gather most of the statistics.

Comments are closed.