How to block betting sites on iPhone

Why the sites are the harder half

An app is a single object with a single seal. A website is an address, and an address can be reached from Safari, from Chrome, from a browser living inside a messaging app, from a search result, from a link somebody sent you, and from a shortcut you saved to the home screen a year ago. Several of the largest sportsbooks run perfectly well without an app at all.

So the question is not really how to block a betting site. It is which of the phone's layers answers the address, and which of them a browser can step around.

Apple's own website filter, which is real and small

iOS ships a website filter inside Screen Time, and it does what it says: a domain on the list is refused, and what the reader sees is Apple's own restricted site page rather than anybody else's block screen.

Two limits come with it. The first is coverage. A curated list is a list, regional bookmakers are missing from most of them, and new addresses appear faster than any list is edited. The second is the one that matters more: the filter takes effect inside Safari and nowhere else. Chrome, Firefox, Edge, Brave and every browser embedded in another app are untouched by it, which is a large share of where betting actually happens on a phone.

There is also the question of who can lift it, and the honest answer is usually you. The self built version of this is a Screen Time passcode that somebody else typed in, and Apple's own page on a forgotten Screen Time passcode explains that you can reset it with the Apple Account used to set it up. If the Apple Account is yours, so is the reset, and the friend never hears about it. Why blockers fail quotes that page directly.

A DNS or VPN profile you install yourself

The next step up is a filtering service you point the phone at. Some do it with a DNS profile, which changes where the phone sends its name lookups; some do it with a local VPN profile, which routes traffic through a filter on the device. Both are real mechanisms, both stop real pages, and both reach every browser rather than Safari alone. If breadth is what you are missing, this is the layer that supplies it.

Two holes come with the class, and neither is a fault of any particular service. A profile lives in Settings on a device you own, so removing it takes about as long as adding it did; a filter of this kind is exactly as strong as your willingness to leave the setting alone. And name based filtering cannot see an app that ships its own addresses, pins its own resolver inside itself, or reaches its servers through a shared host, so a native app may keep working while every browser is covered.

The same layer, with something holding it on

Fold'em's web shield is the DNS mechanism above. Every name this phone looks up goes to Fold'em's own encrypted resolver, and the resolver answers a known gambling domain with 0.0.0.0, the address for nowhere, so the page never loads. That reaches every browser on the phone, browsers inside other apps included, and it needs no tunnel.

The difference from the paragraph above it is not the blocking, because the blocking is the same idea. It is what happens when you decide to end it. Switching the setting off leaves the app seal standing and leaves the term running, so it weakens one layer rather than ending the block, and the term's end date is on the server where the phone cannot move it.

One thing is worth saying plainly rather than leaving you to discover it. Switching the DNS setting off is the one weakening Fold'em does not report: it sends no email, and on a phone whose apps are sealed it changes nothing on a keyholder's pass either. Only a phone that sealed the web shield on its own shows it there. Anyone telling you a keyholder is always told when protection drops has not read the product.

Three site lists, and why they are never added together

Reading a product page you would think there is one number. There are three lists in Fold'em, they belong to different layers, and adding them up would be arithmetic on things that are not the same kind.

The web shield's resolver answers for about 356,500 gambling addresses and reaches every browser on the phone. Screen Time's own website filter carries a much smaller curated list, ten domains today, and reaches Safari alone. The third list is yours: any website you tick in Apple's picker, up to fifty, shielded the way an app is, so it lands on Fold'em's own block screen rather than on Apple's page. Which sites are on the block list keeps the three apart.

What still gets through, and what each one means

Something getting through is usually one of four things, and they need different responses.

The address is not on the list. A resolver answers for names it knows. A regional bookmaker, a new domain or a site that moved is a gap in coverage rather than a failure of the mechanism, and it is fixed by adding the address rather than by changing products.

Something else is answering first. If a second VPN or a private relay is running, it takes the answer before the shield does. Fold'em reports this rather than pretending: the web tile reads Standing by, which means the shield is installed and another service is in front of it.

The app is reaching its servers directly. Name filtering is blind to an app that carries its own addresses. This is why an app seal and an address shield are two layers rather than one, and why a product that only sells you one of them is selling you half.

The shield is not on. The web tile reads Off or Unknown, and turning it back on takes about a minute. A betting site still loads walks each of the four scenarios in order and says which are the block working at its edge.

Belt and braces is a legitimate answer

Nothing here argues for one layer. A broad address filter for reach, plus a sealed set of apps for the ones that matter, plus something holding both on, is a better arrangement than any single piece of it, and the reason is that they fail in different directions. The layer to be most careful about is the one you can undo alone, at the hour you will want to.

The pillar covers the app half of the same question, and the mechanism table sets the four classes side by side with what each one can and cannot reach.