I keep a fairly detailed log book of my personal and professional life in slim black notebooks , each one of which lasts me about 10-12 days. Each book has 60 blank lined pages, and over the years I’ve divided the book into different sections. Page 1 is a list of what I eat everyday (three main meals and two snacks), page 2 is for new ideas (often only a few lines there, but sometimes full) , and page 3 is a weekly planner (of the next four weeks). After that, starting page 4, is a log of each day. On the top margin of each page I summarize the main events that happened on that page, a maximum of five. At the end of the book (pages 58 and 59) I keep an index of the whole log book, which is a list, page by page of the main events on each page. On the last page a list of books read, movies/series watched and important events/moments that week.

**Indexes and events**

In 2014 I had a total of 3578 events. I know this because I put the index of every book in an excel file and can count up all the events. So the only thing that ends up getting copied to the computer is the index of each book (the main events on each page). Here I include an excel template to explain this better. It has 9 columns by 60 rows. The first column is the book number (quite repetitive but useful later), followed by the page number, date, event numbers 1 to 5 (five columns) and the total number of events on that page (I use the excel formula CountA to just count the number of filled cells on each row). You’ll notice a line in the Excel template after page 53, which is where I include summaries of things I’ve read.

I fill in this excel as I finish each log book. The whole process of writing the index in the first place may take 20-30 min and copying it to excel 10 min. So at the end of last year (2014) I had an excel that is 297 columns wide (33 books times 9 columns). Now the question is what to do with the information.

First of all, if I want to analyse it statistically I need it all vertical, which I do by telling excel that the rows underneath the first book are equal to the cells in the subsequent columns after the first book (e.g., cell A62=cell J2, B62=K2, etc).

Once I have it all vertical, I can sum the total events per that year and put it into a stats program to calculate events per day, per week or per month. I use a program called Statistix, quite powerful, fast and takes up minimal space. To analyse any variation per day or per month in the stats program, I first ask Excel to tell me the day of the week and the month based on the date (which is helpful for graphs where you want averages per day of the week or month). The equation is “TEXT((cell with date); “dddd”)” for days of the week and “TEXT((cell with date); “mmmm”)” for month of the year. Since some stats program don’t like text, I then convert the day of the week and month of the year into numbers, simply copying the column with the text day of week into another column and replacing Monday by 1, Tuesday by 2, etc. The same with months (January by 1 etc.). This is a nice moment to see if there are any errors, since sometimes there are two days on the same row, or some dates have been punched in wrong or are incomplete. It’s a drag if there are two days in the same row, so I try to make every day start on a new page in the notebook.

I’ve included two graphs about the events in 2014, per day of the week (Figure 1), or month of the year (Figure 2). On average I have 1-2 events per page. The variation during the week is not large, and not statistically significant (Figure 1). There are more important variations throughout the year, with significant differences during vacations (Figure 2).

Unsurprisingly, during vacations (and weekends) I found that I use up fewer pages, or write less things down in my notebook (I need some time to relax!). Somewhat counter-intuitively however, on vacation there are more events per page, or higher “event density” per page when I’m more relaxed. So technically, for me, a sure sign that I’m working is that I have fewer events per page (range 1-2). Work is more monotonous but at the same time it drags out time, and to some it may seem like that time is more significant. Maybe that is part of the reason why many people would rather be working…

Ok, last thing with the days and the months. Using the stats program I can get the total events per day (summary statistics) or per month. That allows me to see, on one hand, the most eventful day of the year (the 10th of May, the first day of the quantified self conference in Amsterdam, Figure 3) and the most eventful months (May and September, mostly related to the conference and to periods when I give the most classes, Figure 4).

So, I’m sure this system can be improved, and it may appear to be laborious, but really its just a habit now and does not really take me that much time. I just find it interesting that I found this way of doing things, and I think it helps me to plan things and in general to live a happier life. But who knows, I may just be going bonkers.