A simple example we use is the most likely high temperature is some location in the world. I send you the Trinidad with a clip board, a hat, a chair and a pencil. Your job for 365 days to to write down the high temperature of the day. When you get done you make a histogram of the values that occur and the frequency of those values.
What you see is there is a most likely value. In Trinidad that value is something around 87 degrees.
Now I'm sending you to Cody Wyoming to do the same thing. What you'll discover is the value is about the same. The Most Li key value, means the value that occurs most often - somewhere close to 87 degrees.
This Most Likely value is the number that lives in the DURATION field of any scheduling tool. It's the number of days the work will take. But of course the variance of that Most Likely - the Mode of a probability distribution - is not the same in Cody as it is in Trinidad.
So why all this statistics and probability stuff in project management, Why not just write down some number, it's going to be wrong any way, we're always late and over budget. For that very reason, we need to understand what the variances are on our estimates. Without this knowledge, we are going to be late and over budget with 100% confidence.
So why the R book. R is a free statistical analysis tool. You can do a lot of things in Excel, but R has capabilities you only do in VBA, and that's a pain. R provides you with all that is needed to figure out of the past performance data is going to provide any confidence that you can complete on time and on budget with some level of confidence.
One thing R does is compute correlations between random variables. The drivers of most project grief come from correlated variances. One late all late (or most late). Risk is the same way.
Next time you're sitting in the conference room with your colleagues and planning the next round of work for the project, think in terms of probability and statistics and that most likely (mode) temperature in Cody, rather than Trinidad.