Free Republic
Browse · Search
News/Activism
Topics · Post Article

To: inquest

Yahoo's popular vote total still shows 51% to 48%. They are truncating, not rounding. It's actually 51.52 to 48.48. I'd round that to 52-48..


51 posted on 11/05/2004 8:06:51 AM PST by scottinoc
[ Post Reply | Private Reply | To 1 | View Replies ]


To: scottinoc
Yahoo's popular vote total still shows 51% to 48%. They are truncating, not rounding. It's actually 51.52 to 48.48. I'd round that to 52-48.

Given that the difference is 3%, 51-48 seems quite reasonable. Assuming a perfect two-way split, I'd think the following would be the best roundings:

Winner's share Report
below 50.25 50-50
50.25-50.75 50-49
50.75-51.25 51-49
51.25-51.75 51-48
51.75-52.25 52-48
I don't know if anyplace formalizes this procedure, but it's based on much the same principle as the 'rounding' algorithm used for PostScript® rendering: symmetrical rounding would halve the effective resolution for stroke widths. Assymetric rounding results in a slight loss of positional accuracy, but enhances stroke-width accuracy which is often more 'important'. Likewise when showing winning/losing percentages, what really matters is the difference. Using straight rounding for both would result in the differences always being even numbers. Using assymetric rounding means the difference will be accurate to the nearest unit.
187 posted on 11/07/2004 12:10:32 AM PST by supercat (If Kerry becomes President, nothing bad will happen for which he won't have an excuse.)
[ Post Reply | Private Reply | To 51 | View Replies ]

Free Republic
Browse · Search
News/Activism
Topics · Post Article


FreeRepublic, LLC, PO BOX 9771, FRESNO, CA 93794
FreeRepublic.com is powered by software copyright 2000-2008 John Robinson