Aviator rating spread across three cards

Checked: star marks and review samples for the three Aviator cards. Method: linear comparison on a 0–5 scale with sample sizes beside each mark. Result: a 0.4-star spread from 4.4 to 4.0 with a steep drop in review volume.

Raw marks on the sheet

Air Defense publishes 4.4 from 16.8K ratings. Sky Attack publishes 4.2 from 3.55K. Galaxy Air Fighter publishes 4.0 from 2.19K. All three figures were captured on 29 October 2026 from Google Play. The homepage ratings chart maps those marks to bar widths at 88%, 84%, and 80% of the track (rating ÷ 5).

The absolute gap between first and third is four tenths of a star. That is narrow on a five-point scale, which is why sample size matters in the gunsight order. Without the secondary key, a noisy 4.0 with a tiny sample could overtake a steadier 4.2.

Sample size cliff

16.8K → 3.55K → 2.19K is not a gentle slope. The lead card carries roughly 4.7× the reviews of the second card and about 7.7× the third. Protocol reading: confidence intervals on the lead mark are tighter. The desk does not compute formal statistics; it only records that the secondary sort key is decisive when stars are close.

Download bands reinforce the story without overriding it. Air Defense sits at 1M+ while the other two share 100K+. When stars and samples already agree on order, downloads act as a confirming tertiary signal rather than a rewrite.

What the spread does not mean

A higher star mark is not a promise of longer sessions, better netcode, or fewer ads. Sky Attack and Galaxy Air Fighter both declare ads; Air Defense’s pass sheet does not. That monetisation difference is orthogonal to the rating spread and is covered in the ads guide.

Nor does the spread justify “best Aviator” marketing language. Kakob forbids best/top phrasing. The gunsight states order under a named rule set, then stops.

Chart hygiene

Bars use CSS widths from real ratings only. No decorative inflation. Labels carry the app name and the numeric mark. Colour comes from the Electric Violet palette; length is the data. If a future re-check moves a mark, the chart and this note update together.

Reader use

Use the spread to calibrate expectations: all three cluster in the low-to-mid 4s, so differences are incremental. Use the samples to calibrate confidence: the lead card’s mark is the most heavily reviewed. Then open the individual Aviator passes for role fit (tank AA, ship-upgrade loops, roguelike arcade).

Worked arithmetic

4.4 ÷ 5 = 0.88 → 88% bar. 4.2 ÷ 5 = 0.84 → 84% bar. 4.0 ÷ 5 = 0.80 → 80% bar. The chart uses those percentages as CSS custom properties. No easing curve, no decorative minimum width that would lie about a 2-star title if one ever appeared. On this Aviator board the bars look similar because the marks are similar; that visual honesty is intentional.

Sample ratios: 16800 / 3550 ≈ 4.73; 16800 / 2190 ≈ 7.67; 3550 / 2190 ≈ 1.62. The steepest cliff is between first and second, not between second and third. That is why Air Defense’s lead feels decisive even though the star gap to Sky Attack is only 0.2. Protocol readers should quote both star gap and sample ratio when summarising the board.

Download bands cannot be divided the same way because Play publishes ranges. 1M+ versus 100K+ is an order-of-magnitude statement, not a precise multiple. The tertiary key therefore only breaks true ties; it does not invent precision.

Communication guidance for the desk

When summarising the Aviator board aloud or in a column lede, prefer “4.4 from 16.8K, then 4.2 from 3.55K, then 4.0 from 2.19K” over “three similar fours.” The second phrase erases the sample cliff. Prefer “0.4-star total spread” over “basically tied,” because 0.4 plus sample ratios is how the lock was justified.

Avoid implying that Galaxy Air Fighter is “bad.” 4.0 with 2.19K reviews on a 100K+ band is a passing inclusion, not a reject. Avoid implying Air Defense is unbeatable forever. Play ratings move. The note is a snapshot tied to 29 October 2026.

Chart colours use the Electric Violet primary and cyan accent on the Sunset Golden stylistic surface. Colour is branding; length is data. If someone screenshots the chart without labels, the percentages still encode the marks for anyone who knows the 0–5 mapping.

Appendix: numbers in one place

Air Defense 4.4 / 16.8K / 1M+ / bar 88%. Sky Attack 4.2 / 3.55K / 100K+ / bar 84%. Galaxy Air Fighter 4.0 / 2.19K / 100K+ / bar 80%. Spread 0.4 stars. Sample ratios ≈4.73× and ≈7.67× from the lead. Check day 29 October 2026. Chart block id on homepage inherits these widths. Note slug: /notes/aviator-rating-spread-three-cards. Parent index: /columns. Related method: /guides/aviator-arcade-gunsight-method.

Keep this appendix when quoting on social or in email so the Aviator numbers travel with their samples. Stars without samples are incomplete gunsight data.

Why a note instead of only a chart

Charts show shape; notes show argument. The Aviator rating spread needs both because a 0.4-star visual looks mild until the sample ratios arrive. Writing the ratios in prose also helps screen-reader users who may not infer percentage widths from CSS alone. The note therefore duplicates the numeric payload intentionally.

Columns index lists this note after the three review passes and the method guide so readers meet the cards before the arithmetic. That order mirrors how the gunsight itself works: identify survivors, then explain why the lock looks the way it does.

If the chart and this note ever disagree, treat that as a bug. They share one data source: the 29 October 2026 capture rows. Fix both in the same commit.

Re-check triggers for the spread

The gunsight re-opens the Aviator rating sheet when any of these land: a star mark moves by 0.1 or more on a card, a review sample jumps a full thousand, a download band changes its published tier, or a package vanishes from Play. Small day-to-day wobble inside the same tenth does not force a rewrite. Protocol prefers quiet boards over thrashing the homepage chart every morning.

When a re-check does fire, update order: capture rows first, ranking cards second, homepage chart third, this note fourth, methodology page fifth. That sequence keeps the Aviator desk from publishing a chart that still quotes yesterday’s samples. The checked date string changes only when the capture rows change; cosmetic copy edits do not move the stamp.

Readers who bookmarked this note should still trust the appendix even after a re-check, because the appendix always restates the three marks with their samples. If the appendix and the chart disagree for more than a deploy cycle, file a correction through Contact with the URL and the two conflicting figures.

India-facing reading of the same marks

SITE_LOCALE en-IN means the desk writes for an India-facing English audience, not that Google Play India necessarily shows identical review totals to every other storefront view. The 29 October 2026 pass used hl=en&gl=in when the listings were opened. If your account locale surfaces a different sample, treat that as a storefront variance, not as Kakob inventing a private counter.

Download bands are likewise Play-published ranges for that view. 1M+ and 100K+ are the bands on the pass sheet. The Aviator gunsight does not convert them into daily active users, retention curves, or revenue guesses. Those metrics are absent from the public card, so they stay absent here.