I'm the parent of a kid who took part in the 2015 school lottery in Amsterdam, and since I had skin in the game I went as far as studying the original nobel pri

I'm the parent of a kid who took part in the 2015 school lottery in Amsterdam, and since I had skin in the game I went as far as studying the original nobel prize winning papers on the DA algorithms to understand exactly what was going. This article is actually a really good summary, the amount of misinformation out there being spread by journalists who did not fully grok the (admittedly complex) system is astonishing. One thing that the article only touches on is the "preferred placement" that some schools managed to bring in from the old system, e.g. Montessori kids get preference for Montessori schools. This ruins the elegance and inherent fairness of the algorithm. In particular it makes it possible to do some form of strategic selection again if you have a preferred placement option that you can trade. This tends to benefit kids with clever parents disproportionately, as the article notes. A "lesson learned" for policy makers should be that even though an algorithm performs better on being "fair" (Multi-DA), it can still be better to choose an algorithm that is easier to explain. (They did end up switching to a different algorithm the next year)