How do I estimate revenue from Steam review count?

The Boxleiter method estimates lifetime sales as approximately 30-70x review count (the multiplier varies by genre and recency of launch), then multiplies by price for revenue. The original 1990s-2010s formulation used roughly 50 sales per review; modern variants land in the 30-70 range. A calibrated probability cone (P10/P50/P90) is more honest about the uncertainty than any single multiplier.

What's the right Boxleiter multiplier in 2026?

30-70 sales per Steam review, depending on genre, age of the game (older games have a higher review-to-sales ratio because reviews accumulate post-launch), and platform mix. Use 30 as a conservative floor, 50 as a typical mid, 70 as the optimistic ceiling. The cone overlay shows why even the 30-70 range understates real variance.

Why is a calibrated cone better than the Boxleiter mid number?

A single-number estimate (Boxleiter mid) is right ~50% of the time and wrong on either side the other half. A calibrated probability cone gives you P10 (90% chance of landing above this) and P90 (90% chance of landing below this), so you can plan budget against the floor instead of betting on the median.

Is this calculator the same as the full forecast tool?

No. This calculator takes post-launch inputs (review count + price) for already-released games. The full forecast tool at /forecast takes pre-launch inputs (Steam app ID, wishlists, days to launch) and runs a calibrated CQR model trained on a 77K-app corpus to output a real probability cone with comp-set evidence and Boxleiter cross-check.

How accurate is the cone overlay shown here?

The overlay on this page is illustrative — it shows the SHAPE of a calibrated cone (typically 5-10x wider than a single Boxleiter number reflects) so you can see why probability ranges matter. For an actual calibrated cone with empirically-validated coverage on your specific game, use /forecast with a Steam app ID. Steam Launch Forecaster's current model is verified at 82% empirical coverage on n=1,560 historical launches.

Steam Revenue Calculator

Free · Boxleiter + calibrated cone overlay · No signup required

Estimate a Steam game’s lifetime revenue from review count + price. Outputs Boxleiter low/mid/high (the indie industry rule of thumb) plus a calibrated P10/P50/P90 probability cone overlay showing why the realistic range is wider than any single multiplier suggests.

Calculator

Boxleiter estimate (single-point)

Low (conservative)

$300K

Mid (typical)

$500K

High (optimistic)

$700K

Calibrated cone overlay (illustrative)

Real launches at the same review count + price land in a much wider range than the Boxleiter band suggests. The cone below shows what an empirically-calibrated P10/P50/P90 envelope typically looks like — roughly 5× wider than the Boxleiter low/high band — so you can plan budget against the floor instead of betting on the median.

P10 (downside)

$120K

P50 (median)

$500K

P90 (upside)

$2.1M

Bar = P10-P90 cone (full color). Dark band overlay = Boxleiter low-high. Dark line = P50 median. The cone is illustrative; for a calibrated cone on a specific Steam game (with empirically-validated coverage), use the full forecast tool with a Steam app ID.

Want a real calibrated cone for your specific game?

The free single-game forecast at /forecast takes a Steam app ID, pulls wishlist + comp-set data, and runs a CQR-calibrated model trained on a 77K-app corpus. Output is a P10/P50/P90 cone with empirically-validated coverage (currently 82% on n=1,560 held-out launches).

Free single-game forecast →

Want the full pre-launch report with marketing-lever causal estimates and Total Lift Attribution? $299 launch report →

Why use a cone instead of a single number?

A single-number revenue estimate (Boxleiter mid, single multiplier) is right roughly half the time and badly wrong the other half. The structural problem isn’t that the multiplier is wrong — it’s that a single number is the wrong shape for an uncertain forecast.

Real Steam launches at the same review count + price routinely land 5-10× apart. A game with 1,000 reviews at $20 might earn $300K (P10 floor) or $4M (P90 ceiling) with the median around $1M — and you cannot tell before launch which side of that distribution you’ll land on. A calibrated cone tells you the realistic floor and ceiling so you can:

Plan against the P10 floor for runway, breakeven, and worst-case decisions.
Communicate the median (P50) to investors and stakeholders as the central case.
Treat the P90 ceiling as upside, not a target.

Where the Boxleiter method came from

The "approximately 50 sales per Steam review" rule of thumb came from Jake Birkett (Grey Alien Games) and was popularized in Chris Zukowski’s How To Market A Game. Modern variants put the multiplier in a 30-70 range depending on launch year, genre, and platform mix. It’s a useful first-pass estimator and a sanity check on more sophisticated forecasts — but it’s a single number, which is the structural issue.

The Boxleiter method explainer guide walks through the full history, the multiplier table by genre, and where the rule breaks down (mega-hits and novel genres).

Built by Greg C. — senior software engineer with production ML experience in calibrated prediction. Steam Launch Forecaster trains a CQR-calibrated model on a 77K-app Steam corpus. See the methodology →