How Far Do Players Get?
A small script that answers the most important question for an idle game – where do players stop? Plus what an AdSense review taught me about forgotten subdomains, and how the web version handles ads while waiting.
Balancing with a simulation tells me how the game should feel. Real players tell me how it does feel. With the first testers playing on Android and in the browser, I wanted one simple answer: where do people stop?
Two sources, one table
There are two sources of information, and neither is complete:
- Cloud saves exist only for players who signed in, but they show exact progress: highest region, current phase, project phase and stars.
- Analytics events exist only for players who agreed to analytics, and they show milestones over time: phases opened, regions unlocked, projects started.
The two cannot be linked to each other – there is deliberately no shared id between them. So a small script reads both and lists them separately: per save the highest region and phase, per analytics user the furthest milestone. It leaves out my own accounts and a list of friends and family, who play more than a typical player and would distort the picture. The output is a plain table I can read in a minute.
The numbers are small, since this is still a test phase, but they already show a shape: where people put the game down, and which stretches are long without anything new to look forward to. Several of the following changes (the earlier trophy, the story cards, the bonus events) were made with this table open next to the balancing simulation.
Requirements by reputation
The same day the requirement for the second special project changed. It used to need a region with a certain number of rating stars, which tied it to one specific playstyle. Now it needs reputation level 5, which can be reached in several ways: completed regions, rating stars or the first project.
What an AdSense review taught me
The website version applied for AdSense, Google's ad network for websites, to show rewarded videos in the browser. The review took a long time, so I checked my domain for anything that might be a problem.
It found something I had completely forgotten: an old subdomain from a project I stopped maintaining years ago, with leftovers of experiments with other ad providers. Nothing on my main site linked to it, but it was on the same domain. I removed the subdomain and its files, cleaned up the search console entries and requested a new review.
The lesson for anyone with an old domain: reviewers look at the whole domain, not just the page you submit. Old experiments do not disappear because you stopped thinking about them.
The web version while waiting
Without approved ads, every "watch a video" button in the browser showed "no video available". That is a bad experience for something that is not the player's fault. So while the review runs, the web version grants ad rewards without a video for everyone, just as for supporter pack owners. When the ads are approved, the videos come back.
The website also got a proper description and a noscript text, so search engines and reviewers see what the page is about even without running the game.
And this is also why this devlog exists: a site that consists of one game and a legal notice is thin content in the eyes of a reviewer. A development diary is honest content about the game, and hopefully interesting for anyone curious how a small game is made.