They put it to the test by gathering data on 32 halvings across 24 assets and evaluated the performance of the coins six months before and after and then compared the results with cryptocurrencies not going through halving events in the same timeframe.
The coins were: BTC, LTC, XVG, FTC, MONA, NMC, FLO, POT, ABY, CURE, NYC, MOON, VTC, EMC2, IOP, MEME, COLX, ONION, DIME, LINDA, UNO, TRC, ANC, and SXC.
Well, yeah. Include shitcoins in your calculation and you get misled.