Methdology for testing

[JiriHrdina]

Guys I'm starting to fart around with how to test various sim scenarios and want to share my thoughts on the approach.

- I will test a variety of scenarios with 10 full seasons simmed for each
- I will post the results for GMs to rank in terms of which scenarios they like dislike

For simming
- Injuries will be turned off
- All lines auto'd
- Suspensions turned off
- Only simming regular seasons to get us good sample size. Playoffs will not be including the testing.

The reason for this are twofold. If injuries are on the sim can't do a full season as it reaches a point where teams are incomplete due to injuries

Lines auto'd to put all teams on equal footing. We want to remove the aspect of line making as an element - and just evaluate the sim with every team using the same approach to lines.

Any thoughts or concerns?

[KevanGuy]

My first thought it to make sure the team rosters are "normal" before simming. What I mean is, someone hasn't loaded up all 60's-70s on left wing and has all 50 on right wing or something like that. Could a test roster file be made for download so we can set our rosters properly and then send back the lines file? Too much work?

[Knut]

Caps should be bottom 10 with our current roster... any simulation where we are higher than that is flawed.

[Caged Great]

Originally Posted by Hesla

Caps should be bottom 10 with our current roster... any simulation where we are higher than that is flawed.

All the SE teams should be in the bottom 15 in terms of points

[JiriHrdina]

Originally Posted by KevanGuy

My first thought it to make sure the team rosters are "normal" before simming. What I mean is, someone hasn't loaded up all 60's-70s on left wing and has all 50 on right wing or something like that. Could a test roster file be made for download so we can set our rosters properly and then send back the lines file? Too much work?

Or I can just set all the lines based on having guys in the right spots. I want to try to remove individual ability to have effective lines from the equation.

Any other thoughts?

[TurdFerguson]

I think you mentioned this in the other thread, but you are you still planning to remove coaches and moral? perhaps these are variables that will help to make up your scenarios?

[JiriHrdina]

Yeah those are essentially part of the different scenarios I'll run.
I'd like to post the results though as blind results - that is people won't know which results are from which scenario - to remove perceived bias over what people think the problem is.
And I'm also hoping people can look past what's best for their individual team v. what's the best outcome for the league overall to balance realism with appropriate randomness.

[KevanGuy]

Originally Posted by JiriHrdina

Or I can just set all the lines based on having guys in the right spots. I want to try to remove individual ability to have effective lines from the equation.

Any other thoughts?

Yeah for sure, was just trying to think of the fastest/easiest way to do that. Also, not sure what the current rosters look like, are there any released players that should still be/put back in the SIM?

[JiriHrdina]

Originally Posted by KevanGuy

Yeah for sure, was just trying to think of the fastest/easiest way to do that. Also, not sure what the current rosters look like, are there any released players that should still be in the SIM?

No there shouldn't be. When playoffs are on - the sim doesn't allow trades/releases or anything else. So they are frozen in their state following any changes made at the deadline.
Some teams are in pretty sh***y states because they dumped players at the deadline - but that's fine. In fact it helps because those teams should clearly be at the bottom of the standings consistently.