Why Game Portals Count Plays Differently

You publish a game on two portals and your own site. Portal A says 12,000 plays, Portal B says 7,400, and your analytics shows 18,000 visits. Nobody is lying. They are counting different things.

This article defines the terms, shows why numbers differ, explains the limits of measuring games inside iframes, and uses worked examples so you can compare platforms fairly.

Metric	What it counts	Typical trigger
Page view	Every time the game's page loads	Opening or refreshing the page
Game start (play)	A player actually begins the game	Pressing Play, or the first input
Session	A continuous period of activity	Starts at first activity, ends after inactivity or closing
Unique player	One person or device, counted once in a period	Cookie or device identifier
Play time	How long the game stayed active	Timers or heartbeats

A page view is the largest number and the least meaningful for engagement. A game start is smaller because many visitors leave before pressing Play. Sessions are smaller still when one person plays several times in a row. Unique players are the smallest, because repeat visitors are counted once.

• Different definitions. One portal counts page loads, another counts clicks on Play, another counts sessions longer than a set time.
• Different filters. Bot traffic, very short visits and duplicate clicks may be removed on one platform and kept on another.
• Different identifiers. Unique player counts depend on cookies or device IDs. If a visitor declines cookies, switches device or clears data, they can appear as a new player.
• Different traffic. A portal with search traffic and a portal with a front-page feature send very different audiences.
• Different time zones and reporting delays. Daily totals can shift by a day, and some dashboards update hourly while others update next day.
• Embeds on other sites. Plays on partner websites that embed your game may or may not be included in your dashboard.

Most portals run your game inside an iframe, a page within a page. That creates real limits:

• The portal controls the page around the game, and you control the code inside it. Neither can see everything the other does.
• Browsers separate storage and cookies for embedded content from the parent page, so identifiers may not match between your game's analytics and the portal's.
• Consent choices matter: if a visitor declines optional analytics, scripts that depend on consent do not run, and that visit disappears from analytics-based numbers.
• Ad blockers and privacy tools can stop analytics scripts from loading, so a script-based count under-reports.
• Server-side counts are more complete, but they cannot see what a player did inside the game without the game reporting it.

PlayKovo's own logic shows how these differences appear in practice:

• Play count: the lifetime counter goes up when the player starts the game, once per game page load. It is not tied to your identity, and the privacy policy states that games record general play counts that are not tied to a person.
• Sessions and play time: a separate session record is created with a unique session ID and a visitor ID, then refreshed by a heartbeat every 90 seconds while the game is active. This only runs when the visitor has accepted the cookie notice, so session numbers can be lower than play counts.
• Embeds on partner sites: launches from websites that embed a PlayKovo game are recorded too. A registered partner site is linked to its launches; unregistered sites are recorded by domain.

So on PlayKovo, "plays" and "sessions" can legitimately differ for the same game on the same day. The same pattern appears on every portal.

Suppose on one day a game gets the following traffic:

• 10,000 page loads.
• 6,000 visitors press Play, so 6,000 game starts (60% start rate).
• Among those, 4,800 had accepted cookies and generated session records (80% consent).
• Those sessions come from 3,600 distinct visitor IDs, because some people played more than once.

Metric	Value	Calculation
Page views	10,000	Raw loads
Game starts	6,000	10,000 x 60%
Sessions	4,800	6,000 x 80%
Unique players	3,600	Distinct IDs among sessions
Plays per player	1.33	4,800 / 3,600

Four honest numbers, ranging from 3,600 to 10,000. A portal that reports page views will look almost three times bigger than one reporting unique players, with the same real audience.

1. Ask each platform what exactly its "play" counts. Write the definition next to the number.
2. Compare ratios, not raw totals: start rate (starts / page views), plays per player, average play time.
3. Pick one metric, such as game starts, and track it per platform over weeks to see trends.
4. Keep your own measurement inside the game for events you care about, like level 1 completed, and compare it to portal numbers rather than expecting equality.
5. Treat small gaps between sources as normal, and investigate only when the gap cannot be explained by the definitions above.

Understanding these differences matters when you decide where to publish. Review how games are checked in how PlayKovo reviews HTML5 game submissions, then submit from the developer area.