You block the number, and the next day the same betting text arrives from a new sender. Blocking silences a sender, not a campaign. Turkcaller reads the text instead of the sender and pulls its list of known betting names from the server, so a new name does not wait for an app update.
Nobody types these one by one. Bulk messaging services push the same text to tens of thousands of numbers at once.
The sender does not know you and does not care whether you read it. The plan rests on a few taps per hundred thousand.
Your number usually reaches those lists one of these ways:
The sender side is not fixed either. Sometimes a short name made of letters and digits, sometimes an ordinary mobile number, sometimes a line registered abroad. All three can be the same campaign.
The answer is economic. One message costs very little to send, so a handful of taps covers a run that reached a hundred thousand people.
Every closed channel is replaced at once. Your block is not a cost, it is one row marked on a list.
There is a second reason. These runs are not built on permission. A compliant brand keeps a permission list, removes you when you opt out and can show its records.
A sender outside the rules keeps no record, so opt-out routes carry no obligation.
Blocking silences a sender, not a campaign. The same text arrives tomorrow from somewhere else. These parts get rotated:
The one thing that does not change is the offer. The sender name is new every week but the text is not, so the filter has to read the text, not the sender. That only works on one condition: the list of betting names it knows has to refresh as fast as the sender renames itself.
When the list changes, Turkcaller rescans the messages that already arrived. A betting name nobody recognised today moves into the spam folder by itself once it reaches the list tomorrow. You do not have to install an update or press anything.
Turkcaller pulls the gambling words from the server instead of carrying them inside the app. The list refreshes at most once a day, when you open Messages. If the phone cannot reach the server it keeps using the list it has and tries again next time.
A late refresh costs you nothing, because the earlier messages are rescanned once the list arrives.
A word list baked into the app is the opposite. Add the word, build a version, wait for store review, wait for people to update. By then the sender is on their third name.
An invented but familiar example: a text from 0555 000 00 00 using the name ornekbet and an address like join.example.com. The next day the same text arrives from another number, with dots between the letters of the name and a digit added to the address.
To a fixed list those are two separate messages. To a filter that flattens spelling and refreshes from the server, they are one campaign.
Key words are broken on purpose, so that a plain word match misses them while your eye still reads them normally.
| Trick | How it looks | What it is for |
|---|---|---|
| Spaces between letters | b e t t i n g | Stopping the word being matched whole |
| Dots or dashes inserted | b.e.t.t.i.n.g | Making the word look like another string |
| Digits standing in for letters | b3tting, b0nus | Slipping past the spelling on the list |
| Look-alike foreign letters | Cyrillic or Greek letters that look identical | Same on screen, different in a comparison |
| Invisible characters | Hidden separators wedged between letters | Splitting the word in the middle |
| Fancy fonts | Bold or boxed characters that still read as letters | Read normally, counted as something else |
Turkcaller undoes this before comparing anything. The text is reduced to one spelling first:
Only then is the decision made. That is why dozens of variants of one word do not each need an entry.
Official permission systems work for senders who follow the rules. If you no longer want a brand's marketing, withdraw permission through the relevant registry or use the opt-out route in the message. For a brand you recognise, that works.
A sender working outside the rules is not registered there at all, so there is nothing to remove you from.
Replying can also backfire: it confirms that your line is live and read. Use opt-out replies only with brands you know.
Your carrier takes complaints about unwanted runs on its network and can act on the sending account. That cuts off a source but will not catch every campaign.
The filter on your phone is the practical layer: it works on your side, without anyone else's cooperation.
These take seconds each:
One caution: the word list also applies to messages carrying verification codes. A word that appears in bank or sign-in codes can silence your own codes.
Most filters ask one question: spam or not. Turkcaller asks a second one: what type. Turkcaller keeps gambling and betting as a type of its own, next to advertising and fraud.
That means:
The order is: your own decisions on top (block, report, chosen type), then the on-device checks on the text. If you turned server review on, the server's answer sits in between, and it still cannot override what you decided.
With server review off, the decision is made on the phone. No connection is needed for it, and the text is not sent anywhere. The only thing coming down is the word list.
The only thing going up is the sender's number, so the server can count how many people a number is texting. Never the content, never your own number. You can switch that off on the Protect screen too.
Server review itself is a separate setting on the Protect screen, so you can see whether it is on and turn it off whenever you want. It arrives off on a fresh install, where you are asked; on devices that updated to this version it arrives on.
Two different jobs for artificial intelligence, worth separating. With server review off, no AI sees your messages: the AI works on the server side, gathering new betting patterns to keep the word list current, and that list is what reaches your phone.
Turn server review on and the message text does go to the server, where hard cases are put to an AI model. That is a separate choice, on the Protect screen, reversible at any time.
A message marked as spam is not deleted. It moves to the spam folder and its notification is silenced. You can open the folder and read it whenever you like.
The one exception is the automatic clean-up you switch on yourself. With it on, spam threads untouched for 30 days go to the bin and can be restored for 24 hours. It never switches itself on.
If a message lands in the wrong folder, mark it as not spam. The conversation moves back to the Inbox and that sender is no longer treated as spam automatically. A trust this sender option makes that permanent.
Verification codes have their own protection: a code is not silenced by the automatic checks on the sender. The exception is your own decisions.
Block a sender yourself, or add a word to your list, and that sender's codes go quiet too. The app says so on screen when you block.
Not at the source, because the sender is not on your phone. What you control is whether they reach your inbox: the aim is that they land quietly in the spam folder.
It is most likely not the same number: the same campaign is arriving under a new sender name. That is why filtering on words and type matters. Even an unknown sender can be recognised by the text.
Your number most likely ended up on a list. The common routes are details you left on a site being sold on, a leak at a service you joined years ago, and automatic lists that simply count through number ranges. Because of the last one, never sharing your number is not on its own enough to keep you off those lists.
No. It stays in the folder and only the notification is silenced. If you switch on the automatic clean-up yourself, spam threads untouched for 30 days go to the bin and can be restored for 24 hours.
Messages carrying verification codes are not silenced by the automatic checks. They can be silenced if you block that sender yourself, or add a word from the code text to your blocked words list.
In the base setup the decision is made on the phone and no message text leaves the device. Server review is a separate setting on the Protect screen: it arrives off on a fresh install, where you are asked, and on devices that updated to this version it arrives on. While it is on, the text goes to the server and hard cases are put to an AI model. The Protect screen shows you which you are in, and one tap turns it off.
Because most of them carry the word list inside the app, so it only changes with a new version. Turkcaller keeps the gambling word list on the server and refreshes it at most once a day when Messages is opened. When it changes, the app rescans messages that already arrived, so folders correct themselves.
Yes, names made of letters and digits can be blocked, not just numbers. For a whole family of numbers, put a star at the end of a prefix and block them all at once.
It is not deleted, it waits in the spam folder. Open it and mark it as not spam: it returns to the Inbox and that sender stops being treated as spam automatically. Use trust this sender to make that permanent.
Turkcaller files gambling and betting messages as their own type, keeps the word list fresh from the server, and makes the decision on your phone. A filed message waits quietly in the spam folder, and nothing is deleted unless you switch the automatic clean-up on.
App StoreGoogle Play
TürkCaller