 Stochastic time tracking is a relatively new type of time tracking. Let's take a look at how it builds upon traditional time tracking systems. These systems usually involve recording what you're doing over constant intervals of time, usually each 6, 30, or 60 minutes. However, this system has some drawbacks. Firstly, most people don't actually spend their time in discrete chunks. Focuses shifted back and forth between things. Quick things are done for just a few minutes and things are done that can't really be categorized into predetermined categories. However, tracking what you're doing on a per second or per millisecond basis would be almost impossible and would certainly result in more time spent tracking time than actually doing things. Generally, the more complex your time tracking system, the more time you spend tracking your time. One alternative approach to manual time tracking is to automatically detect what you're doing and annotate your time based on that. As such as rescue time do this by keeping track of what websites and apps you use and automatically categorizing your usage into categories like productive and distracted automatically. This is a considerable improvement. However, there are some limitations here too. There's no generally applicable way to determine if a given website or app is productive or distracting since different people use them for different purposes. As such, there's a degree of manual classification required. Classifications can also change over time as websites and apps are used for different purposes. Furthermore, what people are doing on a certain website is also relevant. Professional networking on Twitter is productive, but using it to read the latest news might be distracting unless you're a journalist in which case it might be productive. It's hard to capture this subtly automatically and it still suffers from overly broad categorization. Still, automatic time tracking can be quite effective and it can certainly be an improvement over manual tracking. I recommend checking it out and seeing if it works for you. Stochastic time tracking systems are an alternative method of time tracking which is both efficient as you don't need to spend a lot of time tracking and accurate with detailed tags. I'll explain that in a bit. It sends a ping at random intervals to you asking what you're doing right at that moment. Then instead of trying to fit what you're doing into a single category you can enter multiple tags to describe exactly what you are doing. Tagging is a much more extensible system than these rigid categories most other time tracking systems enforce. Most time tracking systems use an average ping gap of 45 minutes. This means that on average each ping will be 45 minutes apart. This means that every 24 hours you get around 32 pings. Note that with stochastic time tracking pings aren't at constant intervals otherwise it would be trivial to game the system by only working right before a ping occurs. It would also result in productivity going down right after a ping since there wouldn't be another one for a while so there would be no chance of that time being sampled. Also the size of the gaps between pings can vary quite considerably. While the data for a single day might be worse than with traditional interval tracking there are many benefits to stochastic time tracking. It's a lot harder to make mistakes in stochastic tracking systems since you don't need to be constantly updating a system whenever you switch tasks. The time tracking system actively asks you what you're doing but more importantly you've got really high precision data. The time tracking system can accurately know when you gaze off or check your email for just a few minutes. It gives you insight into if your email checking is a huge distraction or not. There's also a lot of cool graphs that you can generate from a tag time log. I have a link to one such graph generator in the description. The first implementation of such a system was called tag time. It's an open source implementation written in pearl. You might want to read the blog post about it from the author. Link in the description. I found it quite interesting. I recently written an open source web version called tag time web. I'm currently working on making tag time web better. I have links to both in the description. I hope you found this interesting. Thanks for watching.