Editorial Guidelines

This page explains, concretely, how SpinWave builds its tools and writes its content. We would rather describe the actual process, warts included, than make a general claim like “expertly reviewed” that tells a visitor nothing checkable. Everything below is something you could verify yourself by using the tools or reading the site’s own code.

How our random-picker tools actually work

Every tool on this site, the spin wheel, the coin flip, the dice roller, the number generator, draws its result from your browser’s own crypto.getRandomValues(). That is the same cryptographically secure random source used for generating security tokens, not the weaker Math.random() many similar sites rely on. We also apply rejection sampling wherever a result needs to fit into an uneven range, picking a number from 1 to 7, for instance. That step stops a random byte squeezed into an odd-sized range from quietly favouring the low end. A spin that a room full of people watch has to actually be fair, not just look fair.

How results are generated, specifically

A plain spin draws one item from your list with equal probability for every entry. Turning on Elimination mode removes each winner from the pool once drawn. A full pass through a list then never repeats a name before everyone else has had a turn. Turning on Enable Weights lets you assign different probabilities to different entries deliberately. At that point the wheel is no longer an equal split by design, and we say so directly on the templates that use it. We do not let a weighted result masquerade as a plain fair draw. Nothing on this site fakes an outcome behind the scenes to match a marketing claim or a sponsor’s preference. If a tool says every entry has an equal chance, that is the actual mechanism, checkable in the browser’s own developer console.

How we test a tool before it ships

Every long-form article and every template on this site runs through the same automated check before it goes live. A deterministic content analyzer checks average sentence length and the proportion of unusually long sentences. It also flags em dashes specifically, since that punctuation pattern is one of the more reliable tells of unedited AI writing. This is not a subjective “quality score”. It is a fixed set of rules, and a piece of content either passes them or it does not.

Separately, a regression test suite checks structural facts across the whole site: no two templates share a slug, no canonical link points at a page that does not exist, no template both defers its own SEO signal to a tool page and also carries unreachable long-form content of its own. As of this page’s last update, that suite runs 250 automated tests. Both the analyzer and the test suite run before every single deploy, on the actual production database, not just against a local copy.

What "verified live" means in practice

After every content or code change, we check the actual rendered page on the live site, not just the source file. We use the same tools a visitor’s browser would use: the real HTTP status code, the real page title and meta description, the real canonical tag, and the real visible content. Source code that looks correct but renders incorrectly in production counts as a failure, not a success. We fix it before we consider the change finished.

How blog content gets written and reviewed

Some articles on the blog are written directly by a person on our team. Others start as an AI-assisted draft, generated through the admin tools we built for exactly that. Both paths lead to the same place. Our content database has a genuine review flag. Any article submitted through an automated or agent-driven route is marked pending review automatically. It cannot go live until a person on our team looks at it and explicitly approves it. A human-written article follows the same publish workflow. We do not publish AI output unreviewed. We do not treat a human-written draft as automatically exempt from the same check either.

Our content quality statement

We would rather publish a shorter page that fully answers a real question than a long one padded to hit a word count. A page earns its length by genuinely covering the topic: distinct sections that each say something the others do not, not restating the same point in different words. We do not keyword-stuff. We do not fabricate statistics. We do not claim a tool does something it does not actually do. Where a claim about a tool’s behaviour appears on a page, it reflects the tool’s real, current behaviour, checked against the running code rather than assumed from memory.

Our accuracy statement, and what we do about mistakes

Content on this site covers things like party game rules, classroom activity ideas, sports team rosters and general trivia. We make a genuine effort to get names, rules and facts right at the time of writing, but rosters change, rules get house-modified by different groups, and errors do slip through. If you find one, the contact form reaches us directly, and we correct genuine errors rather than leaving them live once we know about them. We do not silently edit a correction into a page without acknowledging it if the change is substantive.

How your data is handled

The short version: names, questions and options you type into a tool stay in your browser and are never sent to us unless you explicitly press “Save & get link”, and even then we store only the wheel’s labels and colours, never anything that identifies the people those labels might refer to. The full details, including our AdSense and Analytics disclosures and your rights under GDPR and CCPA, are in our Privacy Policy. This page is about how content and tools get built; that one is the authoritative source on data handling.

Where the randomness disclaimer lives

We are upfront that our tools are built for classroom selection, informal giveaways and casual games, not for regulated gambling, official lotteries, legal proceedings or security key generation. The full disclaimer, including the specific limitation of liability, is in section 4 of our Terms of Service.

Questions about this page

If something here is unclear, or you want to know more about how a specific tool or article was built, email admin@spinwave.com or use the contact form. We answer within about 48 hours.