Linear regression showing increase in Lego UCS value

Fcbarcelona101 offers this fantastic study on the value of Lego Ultimate Collector Series kits.


What I did in this particular case is select a popular theme that had a manageable amount of retired sets so that I would not need to spend several days collecting data, and the UCS seemed to be the most useful choice. For my independent variable I selected Change Over Retail, and the dependent variables that would help me determine it ended up being the following:

Price Per Piece

Price Per Gram

Piece Count

MSRP

Total Number of Sets Sold in the Past 12 Months

Approximate Numbers of Years Retired

There are several other variables that more than likely have an impact in how much a set increases in value over its retail price, like amount of exclusive pieces and actual sales numbers from its entire production run, but they are very hard or even impossible to quantify. The ones I selected are probably the most important of the ones we easily have access to.

Thanks, Joel!