Traffic to a horror walkthrough is not traffic to a horror game you made yourself. Both belong to the same genre, perhaps even the same creator's portfolio, but they answer different questions. Combining them makes a wonderfully flattering and misleading dashboard.
1. “Horror is popular” needs a subject
Someone searching for a walkthrough is usually trying to solve a problem in a game they already know. Someone loading your own game is doing something entirely different.
It is easy to celebrate “my horror project is going viral,” until someone asks which horror project: the article about someone else's horror game or the game you built?
The figures are not necessarily wrong. Their labels have gone missing. The applause came from a different auditorium.
2. Articles, playable games, and prototypes have different doors
An article website receives readers. A game site should be measured for visits and actual starts. A prototype page may be visited just to inspect a new layout.
In this example, a separate prototype had 51 page views spread across several days. Those 51 are not verified players of the live game, and the period is different from the article site's one-day total.
A deployed game program is another milestone again. Opening the cinema does not tell you how many tickets were scanned.
3. Page views, visits, and Worker requests are different quantities
Cloudflare defines a page view in Web Analytics by a successful HTML response, and a visit by a page view arriving from another website or a direct link. A visit may contain several page views.[1]
Worker request counts measure requests processed by the running program, not people playing a game.[2] Assets and repeated fetches can generate requests without representing a new player.
Adding “447 article views + 51 prototype views + game Worker requests” into one audience total is like counting apples, oranges, and shopping bags as pieces of fruit.
If live game measurement is missing, write “not measured”, not “zero players.” Missing data and zero activity mean very different things.
4. French-language pages are not the same as French visitors
Another day in the same anonymized snapshot, French-language article URLs received 56 page views. Of those, 49 were attributed to the United States and seven to France. Many had direct or unknown referrers and were scattered across separate articles.
“French pages were viewed 56 times” is supported. “French readers loved them 56 times” goes beyond the evidence. These counts alone also cannot establish whether every request was human.
For multilingual sites, report content language, visitor country, and incoming source separately. One column cannot do all three jobs.
5. Your dashboard should behave like a group of shops
Rank article traffic on the publication's page. Put the playable horror game on its own line. Give every other game, app, or standalone site another line. For each, show its measurement period and label unavailable data explicitly.
What matters is not producing the biggest combined number. It is showing which property earned which kind of attention.
To learn whether a successful article sends players to a game, track the link clicked in the article and the game session actually started. A popular walkthrough is an opportunity for discovery, not proof of conversion.
6. Three scores: read, played, earned
A guide appearing in search results and earning hundreds of views is a legitimate success. Next, ask whether readers moved elsewhere on the site, discovered the creator's game, played it, or generated revenue. Each deserves its own measurement.
The helpful distinction is read, played, earned.
So yes, celebrate when horror takes off. Just check the venue first.
The walkthrough stand may be packed while the haunted-house turnstile has not even been connected to a counter.
The walkthrough still deserves its applause.
