Machine learning for building energy prediction has exploded in popularity in recent years, yet understanding its limitations and potential for improvement are lacking. The ASHRAE Great Energy Predictor III (GEPIII) Kaggle competition was the largest building energy meter machine learning competition ever held with 4,370 participants who submitted 39,403 predictions. The test data set included two years of hourly electricity, hot water, chilled water, and steam readings from 2,380 meters in 1,448 buildings at 16 locations. This paper analyzes the various sources and types of residual model error from an aggregation of the competition's top 50 solutions. This analysis reveals the limitations for machine learning using the standard model inputs of historical meter, weather, and basic building metadata. The types of error are classified according to the amount of time errors occur in each instance, abrupt versus gradual behavior, the magnitude of error, and whether the error existed on single buildings or several buildings at once from a single location. The results show machine learning models have errors within a range of acceptability (RMSLE scaled =< 0.1) on 79.1% of the test data. Lower magnitude model errors (0.1 < RMSLE scaled =< 0.3) occur in 16.1% of the test data. These discrepancies can likely be addressed through additional training data sources or innovations in machine learning. Higher magnitude errors (RMSLE scaled > 0.3) occur in 4.8% of the test data and are unlikely to be accurately predicted regardless of innovation. There is a diversity of error behavior depending on the energy meter type (electricity prediction models have unacceptable error in under 10% of test data, while hot water is over 60%) and building use type (public service less than 14%, while technology/science is just over 46%). Most of the instances of continuous error last longer than three days, and significant portions come from individual buildings and collections of buildings from the same site. This analysis forms the foundation for suggestions to reduce machine learning errors by collecting and using additional training data from onsite and web-based sources to improve the capability, accuracy, scalability, and usability of machine learning. Improvement of these metrics could enhance greater adoption in built environment applications.