I recently stumbled across yet another Last.FM statistics based tool that I wanted to highlight for all of you. I’ve previously highlighted a bunch of really cool tools that utilize your Last.FM listening history, so make sure to check that out as well if music + data piques your interest!
Basics + Uploading Your Scrobbles
Shikaka (presumably named after the bat from Ace Ventura: When Nature Calls) is a free online tool that uses the Spotify API to give users a wide range of text and graph based stats based on their Last.FM listening history. It’s a bit more intensive to use than any of the other tools I’ve talked about in the past, mainly because you need to put your Last.FM username into the tool and then create a password that allows you to place your listening history into their database. From there, depending on the number of scrobbles you have in your library, the tool can take anywhere from minutes to hours processing all your data. When I first put my username into this tool, I had a little more than 64K scrobbles that needed to be imported and processed, and while it had made decent strides in about an hour, I was told to check back the next morning to give the tool time to process the data.
Shikaka ranks as one of the most thorough Last.FM based tools in terms of the raw amount of different metrics and charts. By my count, there are at least 45 different options to choose from: ranging from simple metrics like graphing your overall plays over time and artist and track rankings, to more complicated metrics like a recommendation system, milestones (i.e your 10,000th scrobble), time of day metrics, and seeing which tracks you usually play after another one. Even after a couple of hours of scrolling through all these different metrics, I’m still finding some very cool stuff hidden in this tool.
The metrics that I found the most interesting:
- The recommendation system: Shikaka will recommend artists to you based on your listening history, and there’s an option to take out artists that you’ve already listened to. The list looks pretty similar to the recommendations I’ve gotten from Last.FM and Spotify, so there’s definitely some good logic going on behind the scenes. I’m not sure how it’s calculated, and having that sort of information would be great (see below).
- Simple graphed metrics like plays per month or week are extremely valuable to someone not as familiar with Excel or Tableau (or any other visualization tool).
- Being able to look up metrics about what songs you played after a specific track is really interesting. In addition, seeing “old favorites” – i.e. songs that were played a few times years ago but not recently seems like a great way to rediscover old music that you haven’t given a spin in a while. As someone who’s played with the Last.FM data before, seeing these more complicated, logic based stats (would take some time to figure out in a tool like Tableau) in an easily accessible way is awesome.
Overall, there’s a ton to dig into here, and each person will get different things out of the tool, depending on how much analysis they’ve done in the past.
The Not So Good
From my research, this is actually one of the older and more well known Last.FM tools that has recently gotten a couple of upgrades and additions in recent months. I’m extremely happy to see that there’s still interest in a tool like this, and I hope that with Last.FM + Spotify being connected at the server level that more people will look towards tools like these to dig further into their tastes.
That being said, this looks like a tool that was made 15 years ago. The UI is extremely basic, which in some cases (such as the table views) works pretty well in spitting out information you want and nothing more. But, this is a clunky UI – loading times (especially with uploading the data, and switching from metric to metric) are an issue. Some of the graphical interfaces, which allow you to see things like line charts over time of your play history – rely on Flash, and aren’t able to be customized, making it difficult to get anything really visually pleasing out of this tool.
A lot of the metrics that revolve around running time also seem to be an issue. The tool told me that for my 16K+ tracks, it couldn’t find a running time for about half of them, making all the numbers regarding listening time pretty much worthless, which is unfortunate, since that’s the toughest metric for me to get a hold of in my analysis. Some of the metrics were unfortunately also broken. As an analyst, more descriptions about what different metrics like the “Song Repetition Index” or “AEP” would be helpful as well. But, with a little time and effort on the development side, all of these issues are things that can be fixed.
If you’re just starting to get into music analytics and looking into your listening history, Shikaka is a great place to start. It’ll give you high level metrics across a smattering of different topics. That being said, if you’re looking for something more tangible i.e. visual, I’d suggest different tools like Last FM Extra Stats, or better yet, downloading the CSV of your scrobbles yourself and creating your own visualizations! While I can’t say that I’ll use Shikaka on a regular basis because of my previous visualization experience, I’ll definitely be referring to it in the future for ideas about the sort of metrics that I can look further into.
Are you interested in music and analytics? Reach out to me for more information, and be sure check out my other blog posts revolving around analytics here. If you have any questions about anything I’ve talked about, or have ideas for other tools I can look into, please contact me as well!