I'd settle for a single one that wasn't deeply flawed, mistaken, or outright dishonest.
The now notorious, but well intentioned Rivera, Thomson, Rivera paper, inadvertently compared nice middle class helmeted kids cycling in parks and supervised against inner city kids cycling unhelmetted in busy roads. Astonishingly the pro helmet evidence showed and even stronger correlation with reduced lower leg jnjuries. Despite the flaws in the paper, self serving organisations like RoSPA and others used the bogus results in propaganda because it seems the ends justified the means.
Then more recently I was close to eating
my words and reverting to helmet advocacy after reading a paper which compared head injury vs other injury stats pre and post compulsion in australia - for me a good and convincing approach, but on reading the words they'd taken a different time slice on the evidence than that for compulsion. It wasn't mere cherry picking data but looked liked deliberate dishonesty, which frankly shocked me as Riverat T R did at least mean well.
If helmets really did help, surely there'd be overwhelming evidence rather than shonky papers and outright lies. Every instinct tells
me I'm safer in a helmet, but the evidence doesn't back that up. I have to back the evidence against my instinct