a good one

[Sjoerd Op 't Land]

Yates, Glen wrote/ schreef:

> What problem?
> 
> 82.845 rounded to 2 decimal places should round to 82.84
> <snip>
> Think about it, if you use the more simplistic method of always rounding
> perfect halves up, then you artificially skew your data upward. Perfect
> halves are exactly half way between the next lower number and the next
> higher one, so why would you always round up? The higher number is no closer
> than the lower one.
In MC you can choose:

round(2.5)     --> 3
statround(2.5) --> 2
 
> -Glen Yates
Regards, / Groeten,
Sjoerd