Twitter only mutes 100 keywords
April 2, 2017
By Jeff Johnson, Developer of Underpass
This is not a joke. I wish it were. The Twitter topic du jour has been the new eggless default profile photo, and that change was indeed dumb and pointless, but I want to talk another change that has been rolling out over the past few months: keyword muting. Initially, I was thrilled to finally receive this new feature, which had great potential. However, weeks of usage turned my thrill to disappointment. The last straw was when I ran into this today:
If you haven't used Twitter keyword muting, 100 may seem like plenty to you. To see why it's not enough, you need to know exactly how keyword muting works. To start, ask yourself why @username muting even exists. The ability to mute or block an account was already a Twitter feature. You might think that if you had blocked an account, Twitter would automatically mute any tweets containing mentions of the blocked account. That would be the sane implementation. Yet that is not the actual implementation. Twitter does not automatically mute tweets that mention accounts you've blocked, so you have to waste keyword mutes on those accounts.
The other major problem with Twitter keyword muting is that it only matches full words or phrases. It doesn't match prefixes at all. In other words, Twitter does not automatically mute plural or possessive forms of your keywords. If you want to match a keyword with a suffix (-s, -es, -'s) you have to add additional entries to your keyword muting list. With @username and suffix muting, the number of keywords can easily surpass 100.
Now that I've outlined the problems, let me outline the solutions. In my opinion, here are the ways that Twitter keyword muting could be improved:
- Remove the limit, of course. There should be some limit as a sanity check against attacks from malicious users, but the limit should not present a practical barrier to non-malicious users.
- Automatically mute tweets containing mentions of accounts that you've blocked.
- Allow users to export and import keyword muting lists.
- Allow regex. From my perspective, this would be ideal and give power users the highest level of control. I understand that this would not be ideal to users who don't know regex, though it would be nice as an option you could enable in your settings. Since regex is not for everyone, I'll go on to discuss non-regex solutions below.
- Match prefixes instead of just full words. Or at the very least, automatically match plural and possessive forms. Generalized prefix matching would be preferable, though, because I've noticed that Twitter users are quite "creative" in employing a multitude of variations on words you want to mute.
- Case sensitivity. Twitter keyword matching is case insensitive, but I feel that case insensitivity was a mistake. I understand that this is controversial, and there are decent arguments to be made on either side. In my own usage, however, a lot of my keywords are proper names. I would guess that the most commonly muted keyword on Twitter is one that serves as both the last name of a person and a common verb and noun having its origin in card play. Out of discretion I won't mention the word here. Capitalized, the word almost always refers to the person, unless it comes at the start of a sentence, but that would be highly unusual. Lowercase, the word has innocuous uses that you wouldn't want to mute. Although admittedly, people are probably avoiding those innocuous uses more often lately.
Perhaps each keyword could have a checkbox for case sensitivity. And prefix matching or regex for that matter. These options may complicate the user interface, but for people who rely on keyword muting, the details are important.