Hi folks. My name is Vivek B, and for about 2 years now i’ve been doing monthly graphs & analysis of the OPD crime stats. I load up the data from CrimeSpotting.Org into an access database, and look at it 16 ways from Sunday. I live in Rockridge, so I started off focusimg on Area 1 & our beats, mainly as I wanted to make sure the #s jived up with the feeling around the ‘hood. About 9 months ago I started doing this for all Areas, but kept a detailed analysis on the neighboring beats. My theory is that if crime moves due to higher enforcement, the bad guys are more apt to only move 1-2 beats away, rather than clear across the city to an unknown neighborhood.
I’m writing this guest post to not only let you know how crime has been doing Oakland wide, but also demonstrate how I’m using this to help our immediate beat. I find the #s for any given month wholly unsatisfying – gee, we had 102 calls for OPD service for troubling crimes in July. Is that good? Is that bad? How should I feel? How are other beats doing? Other areas?
But before we get started, let me give you a quick tutorial on how OPD is setup: There’s 3 geographic areas, as shown here:
Furthermore, each area has ~14 beats in it. To give you an aerial example, here’s Rockridge & the neighboring beats:
What I do is to take the full list of crimes, and break them into my own opinion of what the categories should be. I don’t worry about “all crime” as I think that’s a useless # – it treats armed robberies the same as jaywalking. So, I plot:
- Crimes against a person. (aka, violent plus some disturbing stuff)
- Crimes against property
- Crimes against person, property, & major drug offenses (not minor possession).
I plot 2 graphs, one for Area 1 vs 2 vs 3, and one for Rockridge & the 3 beats immediately surrounding it. What I’m looking for is lines that move in opposite directions (crime goes up in one area and down in another) as that’s a clue that crime is merely getting moved around, not actually reduced.
Enough chit-chat, let’s get started. Here’s the graphs from Jan ’08 to present.
Let’s start with crimes against a person:
See how all areas went up in July? Crime went up everywhere, and it was clearly noticeable.
And oy, Rockridge (12Y/13X) had a huge spike.
If you’re the type who likes #s…
You’ll note that I track medians, not averages. That’s because the mean is a useless #, a really bad or really good month can throw off the count. But a median looks for the middle ground, so a really extreme month won’t have an oversized impact.
Since I live in Rockridge, I was understandably concerned about that spike, so I dug further. I plotted the weekly stats for the past 90 days. I saw this:
If you look closely, two weeks were really bad, and killed the stats. But other than that, life was normal. So, i’m hoping that the situation was simply that 1+ roaming gangs chose 12Y/13X for their thuggery, rather than be a permanent increase in the level.
In the report I created for the NCPC, I have even more detailed graphs & tables on the types of crimes during those weeks, but this is already too detailed so i’ll move on to property:
Well, that doesn’t look so bad. Let’s see how the beats are doing:
Rockridge & the immediate beats edged up, and comparisons to the 2009 median is a mixed bag.
So, it seems that life is worse on the property crimes too, just not as bad as the spike in violent crimes.
And what do I do with this info? Well, on a monthly basis I publish it to my site, and based on what it says i take the appropriate action. This past month I decided the spike was bad enough that I sent out a special alert to a bunch of neighbors who typically don’t want to hear about crime, and asked them to take some pretty basic measures (ie, remember to lock the doors, leave nothing of value in plain sight in the car, avoid talking on the cellphone or iPod when walking home from BART, etc).
Hopefully that makes sense. If you have any questions or want to see some specialized graphs, let me know.
Vivek B. is an Oakland resident and creator of Rockridge Residents.org.