I have performed backtest of simple strategy derived from (and very similar to) the Magic Formula and the Acquirer's Multiple for period between 2004.01.01 and 2018.01.01 (14 years) for US equities, which I called "gemalpha" (this is actually name of my product which allows you to invest according to this formula: https://www.gemalpha.com/). This is randomized version of the algorithm I have shared here. Selection of stocks in this algorithm is randomized (out of 100 stocks ranking the highest algorithm is randomly choosing 24 to buy). So every time you will try to run this algorithm you will get different results.
I have implemented this randomized version of my algorithm to test performance in case of randomized/noisy input and to rule out selection bias and curve overfitting problems. Results seem promising. Out of 10 runs I have performed, all of them were performing better than SP500. This is one of these runs.