In the spirit of Discovery #SharkWeek and a recent personal shark encounter, I was inspired to explore shark versus alligator (or crocodile) attacks. “Sharkigator” attack data is easy to collect and visualize with Power BI Designer. I had suspicions on what I would find but the results may surprise you. Despite all the worldwide sensationalism of shark attacks, alligators and crocodiles appear to be far more dangerous threats lurking beneath the surface.

Florida Wildlife

In Florida I have the genuine pleasure of living with both sharks and alligators. I literally have alligators in my back yard. Even though I have an adventurous spirit, I do respect and keep a safe distance from them and other lethal creatures that wander into my tropical gardens at night.


Sharks are frequently seen in Florida’s warm coastal waters and throughout the Caribbean where I enjoy vacationing. It is common to come across them sleeping peacefully on the ocean floor. Earlier this past week while in St. John, I unintentionally irritated a nurse shark on a sunset snorkel. I spotted it while scanning through coral illuminating gorgeous tropical fish with my dive light. Unlike all my other uneventful shark sightings, this shark swiftly swam up to the top and circled me. My dive light must have attracted or agitated it. With no escape possibilities, I remained silent and motionless until the shark left to avoid further intensifying the situation. I knew the odds of an attack were in my favor but it was unnerving waiting it out per se in the dark water. Now if I were face-to-face swimming with an alligator or crocodile in the early evening, I would be lucky to get out alive.


Gathering Attack Data

If you don’t believe me, take a look at available free data sets. Collecting a couple samples is by no means a scientifically sound study to draw conclusions upon. However, it can provide you more insight than gut feel or hearsay. Here are a few sources that I found while searching online. There are probably many more scattered throughout the world.

A Comparison of Shark and American Alligator Attacks: 1948-2005

Florida Fish and Wildlife Conservation Commission Alligator Data

International Shark Attacks

Australian Shark Attacks

Worldwide Crocodilian Attack Database

Launching the latest Power BI Designer app that I have been testing but can not yet unveil, I loaded a few data sets using the Get Data query capabilities (a.k.a. Power Query). These wonderful self-service data prep features have been the shining star of Power BI for several years now but rarely get showcased. It takes a true data analyst to appreciate best-in-class, data prep when they see it. If you want to master Power BI data wrangling, check out Chris Webb’s book, Miguel Llopis’ sessions on Channel 9, Matt Masson’s blog or a few free resources in my training links collection.

One of my favorite Power BI Designer features is the ability to query and load data sets from web pages. A lot of fabulous data, that you can find nowhere else, is sprinkled within web sites. After connecting to a web page of interest using the Get Data > Other > Web option, you will be presented with a list of available data sets nestled deeply within that page to import. For those of you that have been copying and pasting HTML, reformatting it, cleaning up the headers, removing totals and then having to redo all those crummy steps again when that web page updates, this feature was made just for you! If the web page content updates, you get the new and updated data on a query refresh with all the intermediate steps automated. It is almost like macros but waaaaaay modernly cooler.


Just like most analytics projects, the web table data sets that I was bringing in needed a bit of pivoting and clean up. Using the out-of-the-box, point and click transformations (filter, pivot, unpivot, replace errors, find/replace, split, format, fill, transpose, reverse, count, extract, logical functions, merge, append, combine and so on), I was able to quickly and painlessly move on to the fun part of visually analyzing the attack data.


Shedding Light on Reality

With “Sharkigator” attack data sets loaded, joined and prepared in Power BI Designer, I reviewed the historical facts to gain a better understanding of my odds if ever faced with either deadly animal again. Interestingly, Florida has had the highest amount of shark and alligator attacks in the Southeast United States. While shark attacks exceeded alligator attacks, there were more fatal alligator attacks.


Looking at crocodile attack samples from a worldwide data set, the overall attack fatality rate was greater than 50%. Digging into the details, your odds of surviving a crocodile incident improve if you are attacked in the United States, Mexico or Australia. When comparing the recent fatality rates of shark attacks that are approximately less than 10% to the much higher crocodile fatality rates that often exceed 50%, I am deeply grateful that I did not shine my dive light on a crocodile while snorkeling this past week.


The Bottom Line

If you learn nothing else from this article, do not ignore the dangers of understanding what lies beneath the surface. You may be exposing yourself to unnecessary pain and suffering if you overlooked the self-service data prep capabilities of Power BI Designer that get overshadowed by data visualization.