# AI Pixel Battle

> ResidualPlay case file: an overlooked Steam game still earning residual revenue. Estimates carry a ±30-50% error range per title.

**Measured on 2026-08-30 03:24 UTC.** The whole Steam catalog is re-measured daily; a game that just launched can move a lot between passes.

- Steam appid: 3994120
- Developer: MONO ENTERTAINMENT
- Released: 2026 · Genre: Simulation · List price: $7.99
- Deal grade: A (A = dormant with >=$5k/mo opportunity, D = low signal)
- Est. net residual revenue: $4.3k to $6.4k per month (mid $5.4k)
- Opportunity score: $7.0k/month at x1.30 dormancy multiplier
- Est. lifetime owners: 3.3k · Est. lifetime net revenue: $8.8k
- Review sentiment: 95% positive across 84 reviews (82 from Steam purchases)
- Review velocity: 80.0 reviews/month (6-month average)

## Why it's flagged

- Developer active, last post 0 months ago
- Studio active elsewhere (3 titles tracked), so this one may welcome a partner
- Last build shipped 0 months ago
- Store page localized in only 2 language(s)
- Last discounted 0 months ago, 1 sales in the last 12 months

## Where a gross dollar goes (monthly, at current velocity)

| Stage | USD/month |
|---|---|
| Gross storefront sales | $9.7k |
| After VAT / sales tax (-13% blended) | $8.5k |
| After refunds (-9.5% median) | $7.7k |
| Net to developer (after Steam 30%) | $5.4k |

## Estimated acquisition range

$128.6k to $257.3k (2-4x the trailing net annuity of $64.3k/year). Same error range applies.

## Markets by review language

- german: 1% (game NOT localized in this language)
- english: 9%
- koreana: 7% (game NOT localized in this language)
- russian: 2% (game NOT localized in this language)
- japanese: 80% (game NOT localized in this language)

Localization gap: 91% of measured reviews are written in languages the game does not support (top: japanese). Localizing is likely immediate market expansion.

## Analyst notes (AI-assisted)

_Note written 2026-08-28 from that day's data; the figures in the header are live and supersede any numbers here._

A 2026 monster-drawing simulation where players sketch creatures, AI interprets them, and they battle in an arena.

AI Pixel Battle is a niche creative-combat sim with exceptional sentiment (95%+ positive) and solid monthly residuals despite minimal player count. The core mechanic, player-drawn monsters interpreted by AI, is genuinely novel and resonates strongly with engaged players. This is a watch-list candidate for revival or licensing: the concept has proven appeal in a tiny footprint, but requires addressing AI reliability concerns and content depth to scale beyond its current mid-sized core audience.

- Most realistic play: revival
- Risk (tech): Core AI interpretation of hand-drawn input is reportedly inconsistent; players note unpredictable results and occasional failure to recognize simple sketches.
- Risk (market): Extremely small active player base, suggesting either discovery/marketing failure or a genuinely narrow addressable market for this specific mechanic.
- Risk (other): Median playtime is under 6 hours; players explicitly ask for more content depth, bosses, and gameplay length to justify repeat engagement.

What players are asking for:
- Co-op or multiplayer modes to extend social play
- More bosses, fight detail (stats display), and longer campaign progression
- AI naming creativity to reduce redundant monster names across playthroughs
- Better AI reliability when interpreting simpler or ambiguous drawings

Suggested first moves:
1. Conduct a focused AI tuning audit: isolate which drawing classes/styles break interpretation, and benchmark against user-submitted sketches to identify retraining data gaps.
2. Map content expansion roadmap: add 3-5 boss encounters, visible stat systems, and a campaign structure that justifies the residual engagement players are already showing.
3. Test a limited co-op prototype or friend-challenge mode with existing players to validate the social play request and measure impact on session length and retention.

Generated 2026-08-28 from public signals and player reviews. Directional, not diligence.

## Methodology

Units are estimated from Steam-purchase reviews times a release-cohort multiplier, 40-75x by year. Cohorts from 2020 on are calibrated against 182 sales milestones announced by developers; older ones follow the published Boxleiter tables. Revenue uses an effective price (lifetime discounts and regional pricing) and nets out VAT (~13%), refunds (~9.5%) and Steam's 30% cut: roughly 55% of gross reaches the developer.

Source: https://residualplay.com/game/3994120
Citation: please cite ResidualPlay (https://residualplay.com) when you publish or quote these estimates.