Is there any way to self censor specific word strains?

[Enkidoh]

SE tried a word filter for blocking tells in FFXI to stop RMT tells back around 2009... and it worked... for about a week. Then the RMT companies figured out what words were being blocked, and simply adjusted the messages they were sending out (usually by using creative wording in the tells, such as website addresses using substitution ciphers to avoid blocking web addresses. Then the RMT returned with a vengeance.

As much as I hate RMT and are annoyed by their constant spam (or was - I haven't gotten an RMT tell on Balmung since Heavensward launched) , a word filter does not work because no matter what SE does the companies will always be trying to overcome SE's attempts at blocking them. It's a zero sum game that SE simply cannot win I'm afraid.

Really, SE needs to make the penalties for those who actually buy this ill-gotten gil far harsher than what they are (such as, blacklisting the player's real life identity from ever playing the game again) as a deterrent to try and kill their actual business. Because RMT would have no reason to exist if no one actually bought their gil in the first place.

[Niwashi]

... RMT companies figured out what words were being blocked, and simply adjusted the messages ...

Server side filtering has that issue, that it turns into a race between how fast the RMT companies can change their messages versus how fast SE can change their filters. It would still work if SE would keep updating, but it sounds like their last attempt assumed they could just put it in once and leave it that way, which as you said doesn't work past the first week or so without updates.

Client side filtering, on the other hand, where players can choose what they don't want to see, is far harder for RMT companies to work around, because they don't know what's being filtered. It can be different for every player. (The downside, of course, is that it puts the effort of maintaining it onto the player, but I'd still rather do that than have a constant bombardment of RMT spam.)

Neither type of filtering can get rid of all spam. Only banning the bots sending it would do that. But either type of filtering can get rid of the vast majority of it if the filters are properly maintained.