When Data Goes Bad
In the process of determining what data I would include in the Sonic Compendium, I ran quite a few test pivot tables. That made testing the data easier to find what would and wouldn’t work. Here is an example of a test that I definitely couldn’t include.
This looks like it would have been a good chart to include. It shows an interesting angle people might not have thought of. Liz Shaw, of all people, having the the best “per story” usage numbers was something I would not have thought was possible. But that’s where this data gets it wrong.
The number is 13 uses in one story, Inferno. But Liz was in four stories total. In none of the other stories did she use a sonic device. So in the spreadsheet I work with, it gave me an average use of 13. In reality, Liz’s average should be calculated as 13/4 which equals 3.25. That knowledge instantly invalidated all of the numbers on the chart.
It behooves people to be careful when running spreadsheet tests and getting some great looking numbers. They may not be what you think.
I will continue to check and recheck the numbers for the book to make sure I did not include any other erroneous tables as above. The target date is holding true and the book should be completed by May of this year.
