I came across this great article about a group of Reed students in a statistics course that reverse engineered the US News College Ranking System in oder to determine if Reed was being penalized by not filling out info for the the proprietary ranking system. How did they validated it? Statistics of course.

They noted that without being able to derive the statistical model it is uncertain to determine if schools are being properly ranked and there is significant errors with a weighted-sum model (Clarke 2004). As such, they decided to reverse engineer a model to predict school rankings. In doing so, with nearly 94% accuracy, they illustrate that Reed College is the only outlier in predictions, indicating that they might be reaping the penalty by not playing by the games of US News. Regardless, it’s amazing real world statistics and very thoroughly done. Well done Huaying Qiu and Wenxin Du.