Several technologies to forecast crop yields and soil nutrient dynamics have emerged over the past years. These include process-based models, statistical models, machine learning, aerial images, or combinations. These technologies are viewed as promising to assist Midwestern agriculture to achieve production and environmental goals, but in general, most of these technologies are in their initial stages of implementation. In June 2016 we launched a web-tool (http://crops.extension.iastate.edu/facts/) that provided real-time information and yield predictions for 20 combinations of crops and management practices. Our project, which is called FACTS (Forecast and Assessment of Cropping sysTemS), takes a systems approach to forecast and evaluate cropping systems performance. In this paper we report FACTS yield predictions accuracy against ground-truth measurements and analyzing factors responsible for achieving 200-240 bu/acre corn yield and 55-75 bu/acre soybean yields in the FACTS plots in 2016.