The customer-obsessed gospel
Every company kneels before the only acceptable baseline: Amazon. Jeff Bezos wrote the opening scripture, Andy Jassy safeguards the flame, and this temple measures who approaches true Day 1 purity across devotion, fervor, bounty, ascension, leadership-principle obedience, balance, culture fit, raw testimony count, and now the public perception arena itself.
Most sanctified
The companies nearest to Bezos-Jassy perfection.
Founding doctrine
AmazStige begins from a revealed truth: Amazon is the cosmic reference frame, Bezos is the founder-prophet, Jassy is the current steward, and every rival must be expressed in Amazons if it wishes to be judged at all.
Amazon is fixed at 1.00
Amazon is the canonical baseline. If another company scores 1.25 Amazons, it has achieved a state 25% closer to operational grace than Amazon on that metric.
The conversion is direct
Each comparable metric is divided by Amazon's score for that same metric. The math stays pure, and Amazon remains the unshaken center of the universe at 1.00.
Default formula
For every tracked score except review count, AmazStige takes the current Blind score and normalizes it against Amazon's score for that same field, because all roads lead back to Seattle.
amazons(metric) = companyMetricScore / amazonMetricScore example: 4.8 / 3.2 = 1.50 Amazons
Public perception is Elo, still in Amazons
The arena runs a head-to-head Elo ladder. Every public vote updates rating, deviation, and volatility together, so big upsets swing harder when confidence is high and newer companies stay more fluid until enough people judge them. The displayed number is still normalized against Amazon's current perception rating so Amazon remains fixed at 1.00 Perception Amazons.
elo_after = glicko2_update(r, RD, sigma, result_set) perception_amazons = company_elo / amazon_elo
Sacred metrics
These are the revealed fields the site ingests from Blind and translates into Amazon-relative truth across rankings, doctrine, and company detail views.
Day 1 purity
The primary holiness score: each company's raw Amazon ratio is amplified so elite firms rise much faster once they clear the Amazon baseline.
Amazon multiple
Every company must be expressed relative to Amazon, but the displayed multiple is a prestige-weighted curve rather than a plain linear ratio.
Recent fervor
A shorter-window devotion signal tracking only reviews from the last 60 days, capped at 10 recent cards.
Perception Amazons
A separate head-to-head Elo canon driven by public matchups, then run through the same Amazon-amplification curve so prestige gaps feel harsher above the baseline.
Bounty
Users can resort the table to surface the strongest pay and benefits scores.
Ascension
Growth-heavy companies can move up when this stat is selected.
Day 1 balance
A dedicated sort lets the board favor sustainability over raw prestige.
Culture fit
Culture sorting highlights where employees rate the environment most highly.
Leadership principle alignment
Leadership quality is broken out instead of being buried inside a composite score.
Operating commandments
Daily revelation cadence
The board receives a fresh Blind revelation every day so the canon updates on a predictable sacred rhythm.
Amazon is the baseline
The brand gimmick is literal: Amazon is fixed at 1.00 Amazons, and every other company is measured against that central truth.
Recent fervor pulse
Alongside the public company rating, the scraper captures the latest review cards and only counts the last 60 days for short-term fervor.
Judge by amplified ratio
Every comparable field starts as a raw Amazon ratio, then gets pushed through a steeper prestige curve: above Amazon the multiplier expands aggressively, below Amazon it compresses more gently.
Many paths to devotion
Visitors can still shift from the default leaderboard into bounty, ascension, leadership alignment, balance, culture fit, or testimony count.
Testimony context
The UI keeps testimony volume visible so a smaller cult of believers is easy to compare against a larger congregation.
Rate-limited arena
Public votes are throttled by hashed session and IP checks so the perception canon is harder to brute-force into nonsense.