patricksavalle

How to maximise your reach on X — a practical guide from the published algorithm

Everything below is derived from X's own open-sourced ranking code (xai-org/x-algorithm), not from folklore. No growth-hacking theory, no guessing — just what the code actually rewards.

Full technical analysis with file paths and line numbers: https://gist.github.com/patricksavalle/366cf0da247c8d4594a1f7f4467502f5

The one thing to understand first: there is no global "reach score." For every single viewer, a model predicts that person's probability of liking, replying, sharing, muting, etc. Those probabilities get multiplied by fixed weights and summed. Your reach is thousands of independent per-viewer auctions. So the goal is never "get engagement" — it's "be predictably worth engaging with, to specific people."

---

patricksavalle

Reading X's open-sourced algorithm: what it actually says about reach, suppression, and everything outside the code

Source: xai-org/x-algorithm at commit c65aa17 (2026-08-14) Method: every claim below was checked against the code in that repo. File paths and line numbers are given so you can verify each one yourself. Where the code does not settle a question, I say so rather than guessing.

This came out of two conversations: first, a widely-shared summary of the release that got most things right and a few things importantly wrong; second, a sharper set of questions about the parts of the system that the published code cannot show you. Both are covered here.

---

Part 1 — How reach is actually produced

patricksavalle

PayQuality — managementscorecard, planning en facturatieadvies

Peildatum: 3 augustus 2026
Vergelijkingsbasis: integrale managementplanning van 16 juli 2026
Capaciteitsaanname: 36 interne uren per week, AI-native uitvoering
Statusgrens: implementatie, acceptance, productie en externe assurance worden afzonderlijk beoordeeld

Managementsamenvatting

PayQuality presteert voor op functionele productbouw, ongeveer volgens de interne urenraming, maar achter op externe assurance en daadwerkelijke productieoplevering.

patricksavalle

Investigation: chemtrails and deliberate contrail geoengineering

Source access date: 10 July 2026.

Bottom line

The strongest evidence-based alternative is selective contrail climate control, not a fleet-wide toxic-spraying program:

Aviation operators can already choose whether ordinary-looking ice contrails form by changing routes or altitudes. The same control architecture being tested for contrail avoidance can select daytime routes that intentionally produce cooling contrails. Aviation researchers have explicitly proposed, patented and modeled this as a climate intervention. (traced)

patricksavalle

Factcheck: "De Europese ID-wallet is te goed om aan de gevaren te verliezen"

Gegenereerd: 2026-07-02
Object van review: volledige artikeltekst aangeleverd in sessie; geen canonieke publicatie-URL voor het artikel zelf geverifieerd.
Onderzoekstype: journalistieke factcheck / specificatie- en beleidscheck van een expliciet opiniërend stuk.

Eindoordeel

Het artikel is feitelijk grotendeels juist en, voor een opiniërend betoog, opvallend zorgvuldig: het overdrijft de techniek niet en benoemt zelf de meeste voorbehouden. De dragende technische claims (open standaard, selectieve onthulling, device-binding, decentrale uitwisseling zonder centraal logboek, revocatie-model, verplichte registratie van relying parties, vrijwillig gebruik, opensource-eis) worden ondersteund door de eIDAS-verordening en het Architecture Reference Framework (ARF). (traced: S1, S2, S3, S5; accessed 2026-07-02)

patricksavalle

Programmeerbare bankrekeningen in Europa — geverifieerde inventarisatie

Datum verificatie: 26 juni 2026

---

1. SEPA Direct Debit — consumentenrechten op incassocontrole

Verordening (EU) Nr. 260/2012 (SEPA-verordening) geeft rekeninghouders het recht om bij hun betaaldienstverlener in te stellen: (traced — EUR-Lex summary, geraadpleegd 26-06-2026)

patricksavalle

ECB Digitale Euro: Account-based of token-based?

Bronnen geraadpleegd: ECB Digital Euro Scheme Rulebook v0.9 (30 juni 2025, draft) en ECB Prototype Summary (mei 2023)
Datum analyse: 26 juni 2026

---

Conclusie

Het eindontwerp van de digitale euro gebruikt een account-based model — vergelijkbaar met een normale bankrekening, niet met een UTXO/token-model zoals Bitcoin.

patricksavalle

Peer Review: Koster et al. 2026, COVID-19 Infection, Vaccination, and Excess Mortality

Summary

Paper: Eva A. S. Koster, Marije H. Sluiskes, Hein Putter, Frits R. Rosendaal, Astrid van Hylckama Vlieg, Mark G. J. de Boer, and Liesbeth C. de Wreede, "Dynamics of infection, vaccination and excess mortality during the COVID-19 pandemic among older individuals-a nationwide analysis," European Journal of Epidemiology, published online 8 June 2026. (traced - PDF supplied in-session and Springer/PubMed records fetched, accessed 2026-06-10)

Genre / field: Observational epidemiology; nationwide administrative-data cohort; relative survival and multi-state event-history modelling. (traced - paper methods and public code, accessed 2026-06-10)

Recommendation: Major revisions. (inference from traced design, methods, and claims)

patricksavalle

Publicatie-audit: Koster et al. 2026 over infectie, vaccinatie en oversterfte

Korte conclusie

De studie is serieus, nuttig en methodologisch sterker dan een simpele vergelijking tussen "gevaccineerden" en "ongevaccineerden". (traced - paper en code, geraadpleegd 2026-06-10)

Tegelijk bewijst de studie niet causaal dat vaccinatie de oversterfte heeft verlaagd, en sluit zij vaccins niet expliciet uit als mogelijke oorzaak van individuele of kleine aantallen overlijdens. (traced + inference)

De beste lezing is smaller: in geregistreerde Nederlandse data over ouderen in 2020-2021 hangt gedocumenteerde COVID-19-infectie samen met hogere oversterfte, en geregistreerde vaccinatie-status hangt in dit model samen met lagere oversterfte, vooral na gedocumenteerde infectie. (traced + inference)

patricksavalle

Event Investigation: The "THAAD-Rain" Claim

Did Destroying US Radar Arrays in the Middle East End an Engineered Drought in Iraq and Iran?

Investigative analysis using the investigative-reasoning framework. All load-bearing claims carry warrant labels. Sources fetched 2026-06-04.

---

Summary

• Event: In March 2026, Iran struck and destroyed four US AN/TPY-2 X-band THAAD radar systems across Jordan, Saudi Arabia, and the UAE during active hostilities. Weeks later, a now-deleted post from Iran's embassy in Kabul claimed that destroying a "secret cloud seeding and climate modification center" in the UAE had immediately reversed a years-long drought in Iraq and Iran. This narrative mutated in social media into claims that THAAD radars themselves had been suppressing regional precipitation.