TypeTango, a site even older and less active than this one, remains alive. Though less technically sophisticated, it has a few features that I like. I suspect that it used to do more, but perhaps had to rip out some functions than ran afoul of Myers-Briggs and/or Keirsey copyrights.
TypeTango offers no tests these days. It is apparent that very many of its early members were mis-typed – a problem, perhaps, with whatever test they were using at the time and amplified by the fact that most of their users knew nothing about MBTI. Since then, its users have probably relied on whatever free online tests and instruction that PEVO users find, which produce false results at about the same, still alarming, rate. The posted types are, as often as not, a distraction.
TypeTango has no compatibility algorithm. Search results can be filtered by member age, type, location and recent activity. As an approximation of compatibility, TypeTango allows users to provide dozens of open-ended “keywords” – positive and negative. Users can assign a weight to each keyword, but almost no one does. The open-endedness is both a strength and a weakness. Lists of the 100 most popular keywords are provided. This is far too few. Near misses dominate: synonyms; changes in spelling, parts-of-speech, tense or hyphenation; attempts to cram multiple words into one “keyword.”
A few things I like.... TypeTango allows only one new contact per day and tells you to make the most of it. (Ongoing conversations do not count toward this limit.) Users cannot see the date and time that others last signed-on – only the month and year. There is no “Viewed Me.” All of these features seem to make the site less attractive to those who might abuse it. It does not ask about educational attainment or profession. (Curiously, it does ask optionally about weight and seems to get mostly ballpark-accurate answers.) That's it. Pretty bare-bones, really. Nice logo. The site seems to have attracted about 25,000 total users 19 years-of-age and older since about 2002.
There are a few things that an MBTI / Enneagram site really needs to get right, in my opinion:
#1) It needs to attract a large number of users. This usually means that it needs to offer something more than match-based searching to attract users. PEVO's Chatrooms, Groups, Forums, Polls, Meetups and Psychology pages were a valiant and sophisticated (for its time) attempt to add value. PEVO's bulletin-board style home page is also fairly unique. None of this was enough. The original OkCupid offered hundreds of games, including a pretty good MBTI test and another one that assigned a fairly comical but strangely accurate “dating persona.” It was great fun. It probably did attract many users but was probably too costly to maintain. The feature was dropped ages ago.
#2) It needs to provide some sort of typing algorithm or a tie-in to a specific other site. The results have to be right and have to be explained in a way that is understandable to an unsophisticated user. Many people are obviously content to hear that their “rare” type is the likely reason that they haven't yet met the person of their dreams. Which we know is not true. Let me stress: the results have to be right. In my opinion, all of the historic free online dating sites fall far short, notwithstanding their creators' credentials and/or professed expertise. One way to make the results more valid is to refrain from all the story-telling that usually accompanies the results, describing everything else that is supposedly going on in the lives of typical test-takers of a particular type. That's just astrology. A dating-site test either helps someone find a compatible romantic partner or it fails.
#3) The site needs at least a theory of compatibility. Several of today's best-capitalized online dating sites are using machine learning – a form of artificial intelligence – to adjust their algorithms to do “whatever works” (based on still questionable criteria) to improve the predictability of a match. This is far outside the capabilities of most start-ups, and it is not apparent that it helps, anyway. I have posted elsewhere my belief that most of the MBTI-based compatibility algorithms are very weak, purely hypothetical, not objectively tested and contradictory of one another. An enhanced scheme is needed. PEVO's use of Enneagram Instincts was a valuable addition, in my opinion – one that was never fully exploited here or anywhere else. I also like values-based schemas described above. An algorithm that can adapt to its own successes and failures (based on something more than a swipe) would be better still.
#4) It needs to protect its users from threats – both institutional and individual – to their privacy, security and safety. Too many users still give away far too much information. The site has to be better at protecting them than they are themselves – without losing their interest.
A web site or app that does all these things hasn't yet been built.