I'm ashamed to say that I didn't. You're right. Medallion prices look pretty close to average for the dataset, which only covers 18 months. The article also men

I'm ashamed to say that I didn't. You're right. Medallion prices look pretty close to average for the dataset, which only covers 18 months. The article also mentions problems with the data. Low data, small dataset, questionable outlier removal. "There was only one medallion sale in September, followed by nine in October" This data doesn't really say anything. The remarkable thing here is that medallion values aren't dropping. This really is terrible reporting.