Why Live-Service Games Sell Battle Passes Instead of Loot Boxes

The standard account of how games monetisation changed goes like this: loot boxes were everywhere, regulators called them gambling, and the industry retreated to the battle pass. It is a tidy story and the sequence is wrong.

Valve shipped the Dota 2 Compendium in 2013. Fortnite launched its pass in 2018. Belgium’s gaming commission published its loot box finding in April 2018, and most of the regulatory action that followed came later still. The dominant fixed-price seasonal product was already in market and already working before the legal pressure arrived in any serious form.

What actually happened is a convergence. One model was getting more attractive on its own commercial merits at the same time as the other was getting more expensive to keep selling. Understanding the difference matters, because it tells you what happens next — and the answer is not that publishers found their conscience.

What a Battle Pass Is Actually Selling

A loot box sells a chance. A battle pass sells a schedule.

That difference does more commercial work than it appears to. A randomised purchase has to be re-sold every single time; the player decides again, transaction by transaction, and the publisher’s revenue is a function of how many of those decisions go their way this week. A pass is sold once per season, at a known price, and everything after that is the player working through content they have already paid for.

From a finance perspective those are not variations on a theme. One produces revenue that is difficult to forecast and concentrated in a small number of heavy spenders. The other produces something much closer to a subscription — a per-season figure that can be modelled, reported to a board, and multiplied by an expected retention curve.

The Spend Concentration Problem

The uncomfortable arithmetic of randomised monetisation is that most of the money comes from very few players. The industry analytics published through the 2010s — the Swrve and deltaDNA-era studies that shaped how the sector talked about itself — consistently found that a low single-digit percentage of paying players generated a large majority of in-app purchase revenue.

Those are historical figures from a particular period and a particular set of titles, and they should not be read as a current measurement of anything. But the structural point survives the specifics: a revenue line concentrated in a tiny fraction of your users is fragile in ways a spreadsheet does not show you. Those users can leave. They can be legislated away from. And an increasingly loud part of the conversation about them is about whether monetising that group at that intensity was ever defensible.

A pass flattens the curve deliberately. It converts a small number of very large payments into a large number of small, identical ones, and in doing so removes the single most criticised feature of the previous model.

Then the Other Side Got Expensive

Running alongside that commercial drift was a slow rise in the cost of continuing to sell randomised items.

Belgium moved first in 2018. The ESRB introduced an “In-Game Purchases (Includes Random Items)” label in 2020, which put the disclosure on the box rather than in the terms. The UK spent 2022 pressing the industry toward self-regulation rather than legislating. And in Australia, classification amendments that came into force on 22 September 2024 attached a minimum M rating to games with paid randomised in-game purchases, and an R 18+ rating to games with simulated gambling.

None of those individually killed anything. Cumulatively they changed the maths. A rating floor is not a ban, but it is a marketing constraint, a retail constraint and in some markets a platform constraint. If a mechanic pushes your rating up a band, the audience you can legally advertise to shrinks — and for a live-service title that depends on continuously recruiting new players, that is a direct hit to the top of the funnel.

Why This Distinction Is Worth Getting Right

If you believe regulation caused the shift, you expect the pass to be the endpoint — the compliant model the industry was forced into.

If you believe the pass won commercially and regulation merely raised the cost of the alternative, you expect something different: continued iteration toward whatever produces the most predictable revenue per retained player, with the regulatory line treated as a constraint to be routed around rather than a destination.

The evidence favours the second reading. Note what the Australian rules actually catch. The trigger is paid randomised purchases. A fixed-price pass with a published, deterministic reward track does not engage it at all. The industry did not move to a model regulators mandated; it moved to a model that happened to sit outside the line regulators drew, and which it preferred anyway.

What the Pass Costs the Publisher

The model is not free money, which is the part player-facing coverage usually misses.

A pass is a promise of new content on a schedule, and that schedule does not care whether the team is ready. The content treadmill it creates is a permanent operating cost — a live-service title needs people producing cosmetics, events and seasonal structure continuously, for as long as the game earns. That is a very different studio shape from one that ships a game and moves on, and it is a large part of why live-service teams are simultaneously the most valuable and the most exposed to a downturn.

It also compresses pricing power. Once players learn a pass costs roughly the same each season, that becomes the reference point. Raising it is visible in a way that raising the effective cost of a randomised item never was.

What to Watch Instead

The useful signal is not whether a game has a pass. Almost all of them do. It is which parts of the monetisation sit inside the deterministic, published track and which sit outside it in a separately-priced store — because that second category is where the concentration and the pricing power quietly went, and it is far less examined than the pass on the front page.