Alright, so check it out, today I’m gonna walk you through this little project I tackled – messing around with data after hearing about Mike Tyson’s car crash. No, I wasn’t there, just saw the news and my brain went, “Data time!”
First thing I did? Hit up Google. I needed to find some actual accident data, like statistics or reports, anything I could sink my teeth into. It wasn’t about Tyson specifically, but generally car crash info. Think I searched for stuff like “car accident statistics US,” “causes of car accidents,” “drunk driving accidents data.” You know, the usual drill.
Found a few decent sites. The National Highway Traffic Safety Administration (NHTSA) was a goldmine. They’ve got tons of raw data, like seriously detailed reports. Also checked out the Insurance Institute for Highway Safety (IIHS). They do a lot of testing and have some pretty clear summaries of accident causes and trends. State-level DOT websites are also good, since they tend to have facts about accidents in their state.
Once I had my data sources, I started downloading. CSV files, mostly. Big, messy CSV files. That meant it was cleaning time. I dragged everything into good old Excel. I know, fancy data scientists are gonna cringe, but it works! Started deleting columns I didn’t need – stuff like vehicle ID numbers, very specific location codes, you know, the filler. Then I had to fix the data types. Dates were sometimes strings, numbers were sometimes formatted weird… it’s always a pain.
Then the fun began. I started pivoting. I wanted to see things like:
- What are the most common causes of accidents? (Speeding, drunk driving, distractions, etc.)
- What time of day do most accidents happen?
- What’s the age range of drivers involved in accidents?
- Is there a correlation between weather conditions and accident frequency?
Excel’s pivot tables are your friend, trust me. It’s just drag-and-drop, but you can quickly get some pretty interesting insights. I used some charting in Excel too to visualize the data. Bar graphs, pie charts, the whole nine yards. It helps to see the numbers in a picture.
One thing I found particularly interesting was the correlation between time of day and accident type. Like, drunk driving accidents were way more common late at night and early in the morning (shocker, I know). Also, distracted driving accidents seemed to peak during rush hour. Makes sense, right? People are stressed, trying to get to work, and messing with their phones.
Now, it’s important to remember this isn’t some scientific study. I’m just a dude messing around with some data I found online. There could be biases in the data, errors in the reporting, all sorts of things. But it was still a cool way to learn a little more about car accidents and play around with data analysis.
Finally, I saved my Excel file and maybe sent it to a couple of friends. It’s not gonna change the world, but hey, it was a fun afternoon project, and maybe it made me a slightly safer driver.
