Skip to content
indielist (beta)

Celeste vs SANABI

Reviews on Steam

Celeste

3.1× more reviews

Critical reception

Celeste

0.1pp gap

Cheaper at launch

SANABI

$5.00 cheaper

Estimated sales (median)

Celeste

3.9× more units (white-box estimate)

More recent

SANABI

5 years apart

A
C

Celeste

2018 · Action

Maddy Makes Games Inc. · Maddy Makes Games Inc.

Reviews
129,534
Positive
97%
Launch price
$19.99
Engine
Unknown
Sales (median)
9.7M
Net rev
$110.9M
B
S

SANABI

2023 · Action

WONDER POTION · NEOWIZ

Reviews
41,177
Positive
97%
Launch price
$14.99
Engine
Unknown
Sales (median)
2.5M
Net rev
$21.2M

Tag overlap

75% shared , 15 shared tags, 5 only on Celeste, 5 only on SANABI.

Sales estimates

Paid-game estimates use the same transparent multi-factor Boxleiter method. Free-to-play games never receive paid unit or revenue estimates.

Sales estimate

v1.0

5.8M to 13.6M

units (median: 9.7M)

≈ $66.6M to $155.3M modeled net revenue (fixed store-cut, refund, and regional-price assumptions applied)

5.8M median 13.6M
+ How is this calculated?
129,534 reviews × NB(75) = 9,715,050 median units
Factor Adjustment
base +50
year_2018 +15
price_$20 0
positive_97% +10
final NB 75

Heuristic display range: median × [0.6, 1.4]. Boundary protection: NB ∈ [15, 150].

Methodology & limitations

Sales estimates use an additive Boxleiter-derived heuristic against public Steam review counts. They are not authoritative sales data, and the fixed display range is not a statistically validated confidence interval. Actual results may fall outside it, especially for deeply discounted or bundle-heavy titles. Algorithm version v1.0. Read the exact method and validation status.

Sales estimate

v1.0

1.5M to 3.5M

units (median: 2.5M)

≈ $12.7M to $29.6M modeled net revenue (fixed store-cut, refund, and regional-price assumptions applied)

1.5M median 3.5M
+ How is this calculated?
41,177 reviews × NB(60) = 2,470,620 median units
Factor Adjustment
base +50
year_2023 0
price_$15 0
positive_97% +10
final NB 60

Heuristic display range: median × [0.6, 1.4]. Boundary protection: NB ∈ [15, 150].

Methodology & limitations

Sales estimates use an additive Boxleiter-derived heuristic against public Steam review counts. They are not authoritative sales data, and the fixed display range is not a statistically validated confidence interval. Actual results may fall outside it, especially for deeply discounted or bundle-heavy titles. Algorithm version v1.0. Read the exact method and validation status.

How comparison pages work

Verdicts are objective, they call out which game scores higher on a single measurable axis (reviews, price, sales estimate, recency). They don't claim "better." Tag overlap is computed from the public Steam tag list. Sales estimates use a multi-factor Boxleiter method whose formula is shown in full on each game's page.

Want a comparison that doesn't exist? Browse all games, find two slugs, and concatenate them: /compare/games/<slug-a>/<slug-b>. Both orderings render the same content; we canonicalise to alphabetical for SEO.