A chart of people online is a chart of who was let in

Four reasons the measurement itself favours free

None of these are about which games are better. They are about what a concurrent count is counting.

1. A price filters before anything else happens. Every person who would have played, enjoyed it and left is counted by a free game and not counted by a paid one. The gap this opens is not a few per cent. It is the difference between a game’s audience and its customers.
2. Free games are built to be opened today. Daily resets, weekly quests, a season with an end date. Those exist because a free game’s revenue depends on you being present, so presence is what gets designed for — and presence is exactly what the chart measures.
3. The peaks land in different places. A premium game peaks in its launch week, because everyone who bought it plays it at once, and then settles into a fraction of that. A live game peaks whenever it ships an update, several times a year, on a schedule someone chose. One shape decays; the other keeps getting re-lit.
4. Free games can buy population directly. A giveaway, a free weekend, an event that hands out something for logging in. These are cheap for a game with no unit cost and they move a concurrent number immediately, which is why a spike on this kind of chart is often a marketing calendar rather than a change in taste.

Put those together and the top of a concurrent chart stops looking like a ranking of games. It looks like a ranking of business models, with the games attached.

The three doors

A game with no price still has to be paid for. There are only three things it can sell you, and a fourth arrangement that bundles two of them behind a chance.

1. How you look. Skins, colours, emotes, a different set of hands. Nothing here changes the game, which is what makes it the cleanest of the three: a player who never spends is not playing a worse version. It works because other people can see it, which means the product being sold is really the audience.
2. How fast you get there. Experience boosts, skipping a grind, a currency that buys back the evening you would have spent. The design pressure here is unpleasant and worth naming: the slower the free path, the more the shortcut is worth, so the thing being sold is partly the tedium the game itself installed.
3. How strong you are. Power sold directly. It is the most profitable door and the one that empties a game fastest, because a competitive match where the outcome was bought stops being worth turning up for — and turning up is the whole asset.
4. Randomised draws, which bundle the first and the third. A sealed box with stated odds. This is the arrangement that attracted actual regulation: several jurisdictions now require the odds to be published, and a few have restricted the mechanic outright. The law here differs by country and keeps moving, so treat any specific claim about it as dated the moment it is written.

Three years of one player, as arithmetic

Five ways to charge for the same three years of someone’s time. Every figure below is this page’s own assumption multiplied out — a sixty-dollar game, a ten-dollar pass four times a year, a thirteen-dollar month. None of it is measured revenue, and the last column is the one that does the real work.

The row worth reading twice is the battle pass. Forty dollars a year passes the one-time purchase somewhere in year two and keeps going, from a player who was never asked for anything at the door. That is the whole commercial argument for giving the game away, and it is why the shape at the top of a concurrent chart has been stable for years.

Why the average is the wrong number

Divide a free game’s revenue by its players and you get a figure that describes almost nobody. Spending in these games is not spread around an average — the large majority pay nothing at all, a modest group buys a pass now and then, and a small group spends amounts that would look like a mistake in any other product. The total is real. The average player it implies does not exist.

This matters for reading any chart of players, including the ones on this site. A game with a million people in it might be carried by a few thousand of them. A game with a hundred thousand people in it might be carried more evenly and be the healthier business. The count on this board is a headcount, and a headcount cannot distinguish those two cases.

What a concurrent count cannot tell you about money

1. A body is not a buyer. The chart counts people in the game at one second. Whether any of them has ever spent anything is not in the response, has never been in the response, and is not published anywhere per-game.
2. Population can be bought. A free weekend, a giveaway, a crossover event. These move the number on purpose, cost the publisher very little, and look identical on a chart to a game that earned its players.
3. Price is regional. The same game costs very different amounts in different countries, which means the same headcount converts to wildly different revenue depending on where the heads are — a dimension a global chart flattens completely.
4. Smaller can out-earn larger. A premium game at sixty dollars with a tenth of the players can be worth more than a free game with ten times the crowd. Neither number on this site would show it.

That is not an argument against the board. It is the boundary of what it is for. It answers one question honestly — how many people are in this game right now — and every question about money is somebody else’s data.