I didn't say or imply it's just a you thing. I said the overwhelming majority seem to like them. Every single objective measure we have shows they are very popular. Top 3 species. And that that's relevant regardless of whatever /tg/ little wiki says.
What measures do we have though? We have 5E's character numbers, which we know for a fact is inherantly biased due to the half elf's inherant power. We don't have anything that comprehensive for any other edition.
I decided to do an experiment on this, incidentally, and ways we can determine how popular something is, because there's a lot of "Yeah this thing is popular" being said but the only numbers we have are the inherantly biased ones that don't have any data on older editions.
Because they are extremely popular.
Even if you ask 10 times, it will be that same answer.
Are they extremely popular, though?
The only edition we have any sort of numbers on for rolled characters is 5E, where half elves specifically are one of the stronger races from a mechanical standpoint with a lot of websites, if you google 'dnd
class', telling you just how to roll one and set them up easy. We have no data for 3E, none for 1 or 2E (unless someone can find an old Dragon or Dungeon Magazine article, but that's undoubtable going to be very limited to who was reading the book), and if there's any 4E data, I haven't found it
I decided to have some fun on this one actually, I fished out Google Trends to try and use that to determine how much interest there is in them over a time period. You can replicate this experiment yourself! I limited it to the year range due to Google's search result thing changing, as the information before hand wasn't moving too much
Over this year period, half elves are not the most popular. They're not even in the top ones, the most popular one I managed to find are Elves by a long shot. Humans, Dragonborn, Dwarves, Orcs, Goblins, Aasimar and Tieflings fit a second tier, and Half Elves were all the way down in the third batch with Genasi, Goliath, Gnomes, Halflings and Tabaxi, with Half Orcs just under that batch but close enough they're statistically in the same zone. There's also a fourth tier which is the complete unknowns, Yuan-ti, giff, thri-kreen, firbolg, the obscurees. Githyanki was in that bottom-most tier until BG3 released, at which point its jumped up to just under Half Orc
Half elves have their peaks of popularity but, using this as an indication of overall popularity? These numbers indicate they're on par with halflings, genasi, gnomes and tabaxi, just above half orcs, and I don't see anyone saying genasi or tabaxi should be making it into the PHB on their popularity, unless you can provide an explanation that explains why they're being looked into less but are somehow more popular than the races being looked at more. Elves are clearly the ones the most popular looking at this
Half Elves are more popular than Greyhawk over the same time period though, so that's something. Greyhawk has basically no interest on it going through
Being a top 3 race that people like is reason enough to be in the PHB. Thri-Keen didn't even rank top 10. We have so many different ways we've measured popularity. Half-elf always scores very very high. Thri-keen never do.
Humans, Halfling, and Dwarves all ALSO rank very high in every objective measurement we have of popularity. They're not "hanging around" they're actually directly measured repeatedly for decades as popular by fans of this game.
Yes, being very popular is in fact a fantastic reason to include it. It's a way better reason than mechanics, given something can be very unique mechanically and yet be unpopular. Giving people what they want for their entertainment is pretty crucial to entertainment. It's kind of a primary point of it.
This is just a massive speil about ensuring data collection and usage is appropriate to the task at hand so if you also get twice annual training courses on that at your job, feel free to skip this but just note its "The D&D Beyond data is suspect with noted biases towards free items, and therefore cannot be used to confirm anything except that 'This is what D&D Beyond users roll', making it difficult to use as a true measure of anything individual popuarlity point in the wider D&D fan space."
Looking at the information I have found, they're not a Top 3 race except for this edition. We have one measurement of popularity (Characters rolled in a single character builder, which was not Roll20 and not the market leader at the time) and that seems fairly inaccurate,
this goes into some of the raw data and you can see why human fighters with grappler (which the numbers suggests is THE most popular combination of all time) is an artifact of how D&D Beyond works and not necesarily true popularity, for example. Even this is only 1.2 million out of 35 million so, once again, data is skeptical and we don't have true numbers, to say nothing of the fact that it wouldn't have any of the games that use anything other than D&D Beyond.
We only have one survey, from years ago, based on a dataset we can see is easily influenced by definitive product factors and does not necessarily show the true number of rolled characters, along with missing an unknown number of characters from home games and other online tabletops. Without being able to analyse WotC's data they used for that old survey or remove customer biases (IE: D&D Beyond's audience preferring non-paid options leading to below average numbers of the various custom lineages that may not match up with other games), we cannot confirm the data cleansing is sufficient to confirm these are actually rolled and played characters, or indicative of the wider whole. So, I'll look at the data I can justify (IE: Google loves its trends and using Google Trends as a rough indicator) to throw further doubt on top of it. If D&D Beyond's trends were accurate, we'd expect to see similar in the Google trends. And yet, we don't. The two don't even closely match. So there is reason to be skeptical that one truly reflects overall popularity. This is the point we'd go and analyse other data at hand to make a more complete one, but, well. We don't have that data.
