The limited record of reliable observations is definitely a challenge we often struggle to grapple with, especially, as Michelle's recent post discusses, since the climate system can have large, unpredictable variations (internal variability) over many decades. This is one reason why it is critical to have accurate computer models that capture the physical processes of ENSO, so we do not have to rely on the observed record alone. These models are pretty remarkable, but of course, we are always striving to get more observed data and to make better models.