Tue Oct 24, 2023 12:48 pm
I went to the managers records page and sorted on winning % and then chose the top 5 managers with 700+ leagues. They were BC15NY, Goreds54, nomadbrad, The Last Druid, tmfw30.
I then went to my file of league standings that goes through late August 2022 and chose the 100 or more win seasons (12-team) for the above managers since late 2019 (so time range is start dates from late 2019 to August 2022). I manually looked at the playoffs of each of those seasons and recorded the following:
name/lost-semis/lost finals/champs/champs %
BC15NY 8 4 6 33%
Goreds54 3 7 5 33%
nomadbrad 8 7 16 52%
The Last Druid 2 2 5 56%
tmfw30 4 2 5 45%
Total 25 22 37 44%
Total winning % in semis: 70%
Total winning % in finals: 63%
Prior to pulling this data, my expectation for the championship percentage of a 100+ team was about 32% without any additional knowledge about who the manager was. For managers with an established winning record, like those above, my expectation was around 36%. My expectations were based on simulations using a methodology modified from that described in "Analyzing Baseball Data with R" by Marchi, Albert, and Baumer.