First, this is very noisy data to make any predictions from. (I guess it's a good thing to have so few murders that it's hard to make generalizations.) March 2014 alone had 5 murders compared to three last year.
Second, this isn't a 52.9% drop, even if you exclude May, which makes the drop from 17 to 9. It's a 47.1% drop.
This is a very low quality article.
But linking to data for 5 months and then only using the 4 you want to use to skew numbers seems pretty ridiculous to me.
And then in the same page there's data for drug/narcotics violations which isn't even mentioned because it is going up (21% more) even though there are supposedly plummeting marijuana convictions. Also they stated that the crime rate in Denver is going down, yet the data shows the exact opposite.
Whole article seems like cherrypicking at its finest.
One of the arguments against legalization is that once marijuana is legal, we would see a significant _increase_ in crime.[1] So a __statistically__ insignificant result is not a __politically__ insignificant result.
1: From TFA: ""Expect more crime, more kids using marijuana and pot for sale everywhere," said Douglas County Sheriff David Weaver in 2012."