Can I master mental arithmetic in one month?
It’s a new month so time to start my next monthly challenge. I did mention in my last post that I would be learning to use Tensorflow during May, but then I realized I’m about to spend a week in the wilds of Alaska, where a power supply will be hard to come by, let-alone an internet connection. So for May I’ve decided to switch out Tensorflow and instead attempt to ‘master’ mental arithmetic. I’ve had a busy week to start to the month and we’re 4 days deep in May already before I even started on the challenge, so I’d better get cracking.
Why mental arithmetic? Well, I’ve got a scientific background and have always been fairly decent at maths, but schools stop putting focus on mental arithmetic beyond a certain age and it’s not required at all for any advanced maths courses. This means that now whenever I’m put in a situation where some quick mental maths is required, I just feel very ‘rusty’. As though I should be able to get to the answer quicker, but the cogs in my mind are just moving too slowly. Now I did spend some time working in strategy consulting, which really sharpened my rough mental maths — I could quickly tell you that 61 was around 20% of 325, but I wouldn’t be able to tell you that it was actually 19%. One of my consulting peers on the other hand, could tell you within a couple of seconds that it was in fact 18.77%. Whilst that level of accuracy often wasn’t necessary, there was always a part of me that was a bit impressed, as though I’d just witnessed some magic.
Baseline: To make this challenge measurable I want to quantify my improvement in some way. I had a look around for some online tests, but none of them seemed comprehensive so I decided to make my own. I devised a test of 24 questions across 6 question types: addition, subtraction, multiplication, division, “% of” (e.g. 4% of 80) and “as %” (e.g. 14 as a % of 90). Each category has 4 questions of increasing difficulty, using randomly generated numbers of increasing orders of magnitude. I gave myself 10 minutes to answer as many as I could. Of the 24 questions I managed to answer 15 within the time limit, and 12 of those correctly — so an overall score of 50%.
Goal: Setting a goal is going to be a little arbitrary as I don’t know what someone ‘good’ at mental arithmetic would have gotten in my custom-made test, but I’ll go with 80%. The questions I didn’t answer were of course the harder ones in the test, so I think 80% will actually be quite a stretch. Over the course of the month I’ll see if I can find some way to grade myself against the wider population, but I’ll use my own test to grade my progress.
Tactics: At this stage I don’t really know the best approach to take, but I think there will be a couple of key components. Firstly, there may be some fundamentally different approaches to thinking about mental arithmetic which I could adopt — for example some maths-genius kids imagine an abacus in their mind allowing them to visualize the calculation. Secondly, from a few minutes researching it seems that there are a number of ‘tricks’ you can use to help simplify calculations and increase speed. Finally I’ll need to practice, practice, practice — to create that muscle-memory for identifying the best method and then solve the component calculations rapidly.
I think this will be an interesting one and don’t really know what to expect so we’ll see how it goes.
