It is better to be roughly right than exactly wrong.

The following was first posted by Dean Morel on the Motley Fool Stock Advisor board “Becoming an Investing Master” on 5th Feb 2007.

Buffett said something along the lines of it is better to be roughly right than exactly wrong. The idea behind this is you should be giving yourself large enough margin of safety that it doesn’t matter if your figures or estimates are a bit off.
Bill Mann from the Motley Fool wrote some similar thoughts on How to Be Exactly Wrong.

One way to hone your roughly right skills is to come up with ranges in a matrix style. Whatever metric you like to use simply lay out a matrix with low, med and high figures to get a good range. I’ll show you an example later.

Here is another example of the same concept. In this post by Hewitt Heiserman the author of It’s Earnings That Count shows how he uses the mean earnings estimate as a high estimate and then builds in low and medium estimates to provide a margin of safety.

I cobbled together a spreadsheet based on Hewitt’s thoughts. Paul Szydlowski helped me automate the task of calculating PIV and ER. The spreadsheet is based on Jim Gillies’s DCF Calculator. If anyone wants a copy please feel free to grab a copy of my spreadsheet with Microsoft as an example.
Another example is using p/e ranges and estimates ranges

In this matrix p/e is down the side and earnings estimates along the top. This provides a quick look at possible price ranges. The range may seem crazy, but that was for a high growth company which I thought could stumble. Realistically I was looking at the central figures, i.e. $20.65 to $30.97. If you do this for other companies you may want to use p/e increments of 2. The two decimal places in the prices should not be confused for accuracy, it is simply Excel and I recommend you round numbers when doing exercises like this.

Yahoo provides mean, high and low estimates and the number of analysts.
You also have past history, how has the company performed in relation to estimates in the past. Earnings.com is a good source for past earnings history.

With practice you can start coming up with your own estimates for companies.
This is especially true for companies that you own and get to know well.