Back in about 1997, while lecturing at the University of natal, I ran some computer simulations to establish how often you could win the solitaire card game “Perpetual Motion”. I wrote up the results but never published them — which journal would have thought it important enough if I tried? But in 2009 I uploaded the paper to Arxiv so anyone could read it.
I also added some notes to the Wikipedia page. Years later, Mark Masten emailed me to say that his simulation had arrived at quite different results, primarily because we play by different rules.