Have you done any testing on these proposed numbers?pjfoster13 wrote:Goalies are WAY better than you think, at every level. I've noticed that relative to EHM07, the NHL goalie pool is calibrated more towards the higher end which is good. However, people don't realize how close it needs to be as you scale downwards. Kids ENTERING junior (15 year olds) should be all 11s and 12s, guys leaving junior are 13s and 14s, guys in AHL and Europe headed towards the NHL as backups are 15s and 16s, guys establishing themselves as starters are the 17s and 18s and your handful of Lundqvists and Rasks and Quicks etc are 19s and 20s. At the elite level, you're bean-counting between stuff like 19 glove and 20 glove or 18 balance and 19 balance, and that accounts for the difference between .925 and .930 kind of stuff. Also, in general there are TONS of replacement level goalies in real NHL (and at every level, really). For every Henrik Lundqvist or Tuukka Rask with .920+ career save % there's 100 Anton Khudobins and Michael Hutchinsons and Niklas Svedbergs and Andrew Hamburglers- random guys who seemingly come out of nowhere, including being undrafted, to put up ~.910 career save percentage. 1% is the margin of error between elite and replacement level, it's an incredibly small target to aim for. But in general, every goalie is good, and imo EHM has always missed the mark on that. Anders Lindback's ratings (even though he sucks) should be much, much closer to Henrik Lundqvist's than the average person might intuitively think they should be. He should have enough 15s and 16s to land between .900 and .915% (As a side note worth mentioning, re-calibrating the goalies at every level will presumably go a long, long way towards lowering the elevated goals-for totals which most users are noticing)
This is for EHM07, but it should get you started http://www.ehmtheblueline.com/forums/vi ... =13&t=4920pjfoster13 wrote:I would love to learn more about how the "hidden" attributes function, such as PA and consistency and how those items could be re-calibrated as well. Does anyone have any resources on the hidden attributes?
Also, there's some great knowledge/resources in the Data Editing Forum http://www.ehmtheblueline.com/forums/vi ... &sk=t&sd=d