They didn’t accurately forecast the last election, they backtested their method on previous elections and declared themselves accurate within .03%. They may claim they have been doing this for 30 years, yet I find no evidence of that. Their website wasn’t even active in 2008.
"The Monmouth University Poll was sponsored and conducted by the Monmouth University Polling Institute from August 4 to 7, 2016 with a national random sample of 803 registered voters. Interviews were conducted by a live caller in English, including 402 drawn from a list of registered voters (200 landline / 202 cell phone) and 401 using random digit dial (200 landline / 201 cell phone). Monmouth is responsible for all aspects of the survey design, data weighting and analysis. The final sample is weighted for age, gender, race and partisanship based on voter list and U.S. Census information. Data collection support provided by Braun Research (field), Aristotle (voter list sample), and SSI (RDD sample). For results based on this sample, one can say with 95% confidence that the error attributable to sampling has a maximum margin of plus or minus 3.5 percentage points (unadjusted for sample design). Sampling error can be larger for sub-groups (see table below). In addition to sampling error, one should bear in mind that question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of opinion polls."
And see post #64 to see Monmouth's skewed poll weight.