The Quality Premium: Why the Best-Reviewed Steam Games Discount Less (and How Long to Wait for Them)

There is a rule of thumb that gets repeated in every Steam sale thread: never pay more than 75% off, everything gets there eventually. For a lot of games that is true. For the games most people actually want, it is false, and the gap between the two is one of the cleaner patterns in the price history ATLGames has recorded.

I took every game in our history with at least 500 Steam reviews, grouped them by review score, and asked a simple question of each group: what is the deepest discount the typical game in this band has ever had? The answer slopes hard, and in the direction you would guess if you thought about it for a second but probably never checked.

The numbers

878 games had both a recorded discount history and enough reviews to put them in a band. "Deepest discount" is the largest percentage off ever recorded for that game on Steam US; the table shows the median within each band so a handful of extreme titles cannot drag it around.

Read the two columns together, because the contrast is the point. The right-hand column barely moves: every band goes on sale about every six weeks. The middle column moves a lot. The best-reviewed games stop at two thirds off; the worst-reviewed ones get cleared out at 90%. A high score does not delay the next sale. It caps how deep the sale will ever be.

Why the floor rises with the score

None of this requires a conspiracy. A game that people keep recommending keeps selling at 40% and 50% off, year after year, so its publisher never has a reason to go deeper. The discount is a tool for converting hesitant buyers, and a game with an Overwhelmingly Positive score has fewer hesitant buyers to convert. Meanwhile a game with mixed reviews has exhausted its full-price audience early, and every subsequent sale is a search for whoever is left, which means price has to do more and more of the work.

There is a second effect that pushes the same direction. Deep-discount sales tend to come late in a game's life, and the best-reviewed games tend to get sequels, remasters, and expansions that keep them commercially alive. Their publishers keep managing the price instead of liquidating it. Baldur's Gate 3 is the textbook case in our data: 23 recorded sales, a six-week cadence, and a deepest cut of exactly 30%, reached three separate times at $41.99. Anyone who has been waiting for it to hit 60% has been waiting since 2023 and will keep waiting.

Genre changes the depth, not the timing

I expected genre to matter more than it does for timing. It does not matter at all. Every genre with enough games to measure has a median gap between sales of 40 to 44 days, which tells you the cadence is set by Steam's event calendar, not by what kind of game it is. Depth is where genre shows up.

Racing and sports games discount the deepest, which fits a category that ships on an annual cycle: last year's edition has to be cleared before this year's arrives. Roguelikes sit at the shallow end, and I suspect that is the review effect again wearing a genre costume, since the genre's breakout titles are overwhelmingly well reviewed. Everything in between clusters at 80%, which is roughly where the whole data set sits.

What this changes about waiting

The rule of thumb needs a second clause. Everything goes on sale; not everything goes deep. So the waiting decision should start from the review score, because that tells you what "deep" is going to mean for this particular game.

• 95% or higher: treat 60% to 66% off as the floor neighborhood. If the game is tied with its all-time low at anything in that range, there is very little left to wait for. Many never go further.
• 80% to 94%: 75% off is a realistic eventual floor, but the game also goes on sale every six weeks, so you can afford to skip a 50% sale and wait for the next seasonal event.
• Under 80%: the game will very likely reach 90% off at some point. Unless you want it now, the patient move is to wait for a near-giveaway price, and the only argument against waiting is whether you still care by then.

The flip side is a warning for the other direction. A 90% discount on a mixed-review game is not a signal that you found something special. It is the normal end state for that category. The headline percentage is doing exactly what the publisher wants it to do, which is make the game feel like a bargain rather than like a game nobody finished.

A worked example from each end

Top-rated: Baldur's Gate 3 at $41.99 (30% off) shows the "Tied ATL" badge on ATLGames. That is the floor, reached three times, with every sale between them stopping at 25% or less. If you want it, that price is the price; there is no 60% version coming on any evidence we have.

Mixed: for a game with a 65% score and a $29.99 list price, the data says expect a $2.99 or $4.49 sale eventually, probably in a seasonal event, probably within a year. A current 50% discount is the middle of that game's price ladder, not the bottom.

Both answers come from the same two numbers: the game's review score and its recorded sale history. The second one is on every ATLGames game page, under the chart, including the deepest cut we have seen and how often the title discounts.

Method: games with at least 500 Steam reviews and a recorded discount history on Steam US, grouped by percentage of positive reviews; medians shown. The broader frequency and sale-length figures are in How Often Do Steam Games Actually Go on Sale?. Browse current prices on the deals page, where the Steam Rating filter lets you apply the review-score cut directly.