We must keep in mind that we are always attempting to match our explanations to experimental
data. The explanation that fits the data best should be the current favorite,
but new theories are continually being suggested and tested. Because we are working
with such a wide variety of atoms and molecular structures, it is unlikely that a single,
simple approach will work for all of them. Although the fundamental ideas of atomic
and molecular structures are relatively simple, their application to complex molecules is
not. It is also helpful to keep in mind that for many purposes, prediction of exact bond
angles is not usually required. To a first approximation, lone pairs, single bonds, double
bonds, and triple bonds can all be treated similarly when predicting molecular shapes.
However, better predictions of overall shapes can be made by considering some important
differences between lone pairs and bonding pairs. These methods are sufficient to
show the trends and explain the bonding, as in explaining why the H - N - H angle in
ammonia is smaller than the tetrahedral angle in methane and larger than the
H- 0 - H angle in water.