I think you’re trying to extract meaningful insight from likely poor data. Without tags and high fines, I don’t believe any of the published data. I don’t even believe you can find “trend” from chite data.