How indie games discount on Steam

Measured over 6295 games that had been on sale for at least a year, the deepest discount taken inside a twelve month window has a median of 65 percent, with a quartile range from 45 to 80.

Every game here is observed over the same months, from 2025-08-06 to 2026-08-06, so this figure carries no age bias. That is what separates it from the all time low that price trackers usually quote, and it is why the numbers in this section are the ones worth citing.

Two tails are worth naming. 8.3 percent of the pool did not discount at all in those twelve months, and 39.5 percent went to three quarters off or deeper.

Across the 95 subgenres that clear the sample floor, the median deepest cut runs from 37.5 percent in Extraction Shooter to 80 percent in Cinematic Platformer. That is a wider spread than anything price alone produces.

The pattern behind it is not subtle. The shallow end is where competitive and live service games sit, the kind that sell on an active player base rather than on a catalogue price. The deep end is short, narrative, finishable games, which lose nothing by being cheap a year later because a buyer only ever needs one copy.

Over the first 24 months of life, measured on 2006 releases that have completed that window, the price bands are hard to tell apart on depth and on speed. Three of the four share the same median deepest cut, and the time to half price moves by less than two months across the whole range.

What does move, and moves in one direction, is how many games get there at all: 53.5 percent of the Under $10 band reaches half price inside two years against 73.8 percent of the Over $30 band. Expensive games are not discounted deeper or sooner. They are simply more likely to be discounted that far.

The median game runs 8 separate discounts a year and the mean is 7.2. An episode here is one move from full price to any discount, so a sale that deepens midway counts once.

That cadence is close to the seasonal Steam sales plus a couple of the store events in between, which suggests most of the pool is following the calendar rather than running a plan of its own.

60.6 percent of releases open with a launch discount. That single fact ruins the obvious question: measured naively, the median time to a first discount is zero days, which is true and tells you nothing.

Asked properly, excluding the first 14 days, the first discount arrives at a median of 55 days and half price at 325 days. The gap between those two numbers is the interesting part: the first cut comes fast and shallow, and going to half takes most of the first year.

Each release year is measured over its own first 24 months, so a game from the start of the window gets no more time to reach a deep cut than a recent one. On that footing the cohorts are flat: every one of them lands on the same median deepest cut.

This is a null result and it is reported as one. Whatever else has changed on Steam over these years, how far a new release discounts in its first two years is not it.

Across 6863 games the median all time low is 65 percent and the median game takes 1112 days to reach it. That number climbs on its own with age: a game released a decade ago has had a decade of seasonal sales in which to touch its floor.

It is also, and this is the point, the same median that a single twelve month window already produces. Waiting years adds spread at the edges rather than depth in the middle, which is why every claim in this study rests on the fixed window instead.