Editorial policy

Pages built from data

Card pages, player hubs, set checklists and the value guides at /articles are produced from the catalogue itself. The price ladder, the rankings, the year spans and the price history are queries, not prose. Where those pages once carried written commentary, it has been removed: the same handful of sentences spread across tens of thousands of pages says nothing about any individual card, and we would rather show the number than a sentence wrapped around it.

Where those numbers come from, and what they mean, is set out in the pricing methodology.

Articles drafted by AI

The blog and the card guides are written by AI and published automatically. No person reads an individual post before it goes live. That is the same disclosure that appears on the blog index and on every post, and it is the reason those pages exist at all: they are background reading around the data, not a price source.

The rules the drafting runs under:

• No invented numbers. Prices, sales figures and statistics must come from Ripbin's own catalogue or from a real sale that was verified during the run. A topic that cannot be answered with real figures is written qualitatively instead — by tier, not by dollar amount.
• Written against real demand. Topics come from a queue of searches people actually run that we have no page for, ranked by measured impressions, rather than from a list of keywords someone liked the look of.
• A gate decides indexability, not the publisher. Every article is scored when it is published: length, whether it duplicates an existing page, and whether it competes with a page we already rank. Articles that fail are still published and readable, but they are marked noindex so they never enter search results. Failing the gate is common and is meant to be.
• Outbound links are not vetted. We do not editorially endorse links inside an AI-drafted article.

If you find something wrong in one of these, tell us on the contact page. We would rather delete an article than leave a bad one up.

How the investing picks are made

The monthly report at /invest is 40 cards across four price tiers. It is assembled on the first of the month, in this order:

1. A screen runs over our own catalogue and a daily eBay watcher, producing candidates in each price band with a recorded comp attached.
2. The most expensive tier is researched separately, because thin comps at that level cannot be screened reliably. A pick with no verifiable recent sold price is dropped rather than estimated.
3. Each pick is written up with a thesis, two catalysts and one risk. The risk line is mandatory. A pick without an honest downside does not get published.
4. The report is published automatically. Comps are stated with the month they came from, so you can see how fresh the number behind a pick is.

We do not accept payment to include a card in a report, and no marketplace, grader or seller has any input into which cards appear. The affiliate relationship that funds the site is described in the affiliate disclosure, and it does not touch the picks.

Corrections and updates

Errors get fixed rather than quietly overwritten. A correction that changes a published value or a stated method is logged in the corrections section of the methodology page. Reports come in through the contact form or help@ripbin.com, and we reply either way — including when we decide the original stands.

Dates on articles are real. We do not refresh a date to make an old page look new.

Who is responsible

Editorial responsibility for everything published on Ripbin, including the automated output, sits with Saad Ahmed. Automation does not move that responsibility anywhere. Complaints, corrections and right-of-reply requests go to help@ripbin.com or through the contact page, and reach a person.