3 THINGS THAT WILL MAKE THE CS.COM POLL BETTER !

[chris]

Anyone want to note how many people voted in the cs.com poll this week?

we had 13,763 votes this week for the poll

[Tailbone]

Not true.

I'm with the Hen on this one... more significant figures

Sorry, It IS true.
1.000000000000 may be 1 - to 12 significant digits...but it is still (exactly) one.

1.000000000001 is not exactly 1, but is a sufficient approximation in most practical applications.

What's more, an increase in significant digits doesn't always yield a more accurate result.....eg: rounding errors in calculators - some calculations that SHOULD result in a whole number may generate an incorrect fractional solution.

[JMU DJ]

I'm with the Hen on this one... more significant figures

Sorry, It IS true.
1.000000000000 may be 1 - to 12 significant digits...but it is still (exactly) one.

1.000000000001 is not exactly 1, but is a sufficient approximation in most practical applications.

What's more, an increase in significant digits doesn't always yield a more accurate result.....eg: rounding errors in calculators - some calculations that SHOULD result in a whole number may generate an incorrect fractional solution.

Which is why you use the correct significant figures. If a scale weighs out micro grams, you can't estimate to nano grams. What Hen was using was a simple example, if it's statistically relevant and accurate to go that far then that number is more accurate and correct than a whole number.



.... Why aren't we talking about football, my head hurts.