If 50% of your respondants are answering a specific way, the last thing you want to do is drop that option. Even if it is a disingenuous response, you want to alter it to remove the motivation for lying (Give it a less cute name), rather than remove the response and create a weird influence on your other choices (ex: If lots of players play 'jank' 90% of the time but goes to aces as a fallback because its the best or they do it at serious tournaments, it may result in aces over-preforming because they are 'the strongest' when most players DON'T play it and want more support for combo or control or whatever).
And I don't think it IS purely disengenous., if you peruse metawing you find some true gems like "Intimidation VCX+Moldy Crow Jan+Wedge, which I think is fair to call jank. Like I guess you could call it an 'aces' list but it clearly doesn't fly at all like an aces list, or a beef list, or whatever. A LOT of archetypes are very strange and there clearly are a large group of people so comfortable with 'jank' that they go to tournaments. Sometimes enough to get a jank list to the top spot for a underused tournament faction like Rebels or Republic! And because these players are likely under-represented at tournaments, it isn't actually hard to believe 50% of the playerbase 100% sincerely plays jank.
But still, I think this comes down to your 'other' choice being poorly worded. I literally thought it meant 'I could not decide between these two options' rather than 'I play something else. You tried to, I think, cover too much ground with one option. If it was more that was the intent of the option (or even better, was split up for a few different reasons like 'A combination of two of these' 'A different archetype' and 'I don't play just one' ect) I think it would be fine.
Also I get this is a fun fan project and your not seriously conducting market research. I am just a borderline parody, because I teach and grade this stuff and it looks like I won't be this year and have nowhere to put this educational energy. I am in full 'will teach for the thrill of one more informative hit' mode!
Also: Yes, write ins are a nightmare and if I recall google doesn't do them very well. Like if I recall correctly you aren't able to create categories to sort write ins so you can collect different variants of the same thing as the same response so you can compare the correlations between different responses, which makes them a total nightmare. No shame in not bothering for a casual survey.
