I've recently been running a game created by Henrik Kniberg with different teams. The idea behind the game is that it proves how bad multitasking can be even on the most simple of tasks. In this case the task is writing a persons name. I won't go into the details too much here as they can all be found on the crisp website. The game has worked fairly well in all cases and hopefully has helped to make the participants realise the serious downside of multitasking. This was obviously what we were hoping to achieve, however, what got drawn to my attention was the different attitudes that shone through during the game.
The game starts by posing a question; how long does it take to write a name and what factors influence this? People will generally guess around the 5 second mark, although the exact figure doesn't really matter, factors influencing the time to write it will vary from group to group but they generally won't include multitasking. We then ask therefore how long does it take to write 5 names which normally comes in around 25 seconds. Next step is to carry out an iteration where the team writes all the names at once (multitasking), obviously this iteration results in a time much larger than the estimate. We then ask which of these factors caused this inaccuracy of the estimate, clearly multitasking is the key here. We will then run a second iteration where the work in progress is limited to 1. This second iteration is the key to the whole game as it shows how much more quickly the names are delivered when you are not context switching.
The interesting point I've found comes after the first iteration when you ask which of the factors influenced the inaccuracy of the estimate. I've found that so far people fall into two groups. The first group will realise that something has gone wrong in the first iteration and will concentrate on getting to the bottom of what happened. The second group are more interested in trying to justify why their estimate was out and will often cry wolf. Obviously this is a contrived example and is designed to lead you down a certain path but these reactions I've found quite interesting. They seem to separate the people that recognise a problem and try to do something about it and those whose intial reaction is to explain why it wasn't their fault and who's it was. The million dollar question is how do we encourage people to think less like the latter group and more like the first. In the agile environment that we're trying to nurture we want to accept that sometimes things do go wrong and that it's not so important who is to blame for this but that we recognise what caused the problem and then adapt to try to ensure that it doesn't happen again. Going one step further we want people to pro-actively recognise when the way they are working could be improved and making those changes themselves.