Posted by cubcub113 on 10/12/2020 11:10:00 AM (view original):
Posted by Benis on 10/12/2020 10:32:00 AM (view original):
I think some people over exaggerate how much withholding some growth on a player impacts his chances of leaving. If the OP's player hadn't grown 6pts in LP, what does everyone think the chances of him staying increase by? Based upon the repeated conversations on the forums, it seems like people think it's like 10, 20, 30% increase in chances of staying. But in reality it's like a difference of single digit %. Probably like 1-2% difference.
It's like if you lost a roll when you were at 60% and the response would be - Dude, you should have sent 200 more AP so your chance would increase to 62%.
Unless you're on the cusp of changing from On the Fence to Likely Going (or similar type change), the position change on big board has a pretty minimal affect of chances of leaving early.
So the key Benis, is when you're dropping a guy so that there's a chance 60 guys might declare before him. I dropped a JR from 84 to 97 through the season, and he went from 30% or so to 3%, and it ended up that the 93rd guy was the 60 pick, so he didn't even make a decision on drat night. That what my spreadsheet simulates.
8.5.1
Yeah I think this where I'm losing you. Maybe it's the way we're assuming the big board and EE selection process actually works.
Your assumption is that it goes down the line and gives each player a predetermined chance of going early based upon class and location on big board right? And then it does it's yes/no scenario on down the line until it fills all 60 draft spots (graduating seniors are 100% obvs).
I think this makes sense but maybe i'm just super duper slow (most likely) but the actual results don't seem to really mirror this. You would think it's much more of a linear degradation of chance to leave early as you climb the big board but the final results don't really indicate this.
For example, from my spreadsheet where I've tracked 842 Soph EEs and 1232 Jr EEs, the % of Jr EEs from a cluster ranked in the 50-60 range is about the same as a 85-95 range. And then the same thing is true for the Soph EEs from 50-60 is about the same as 85-95.
What I mean is - out of the 135 Jrs who were ranked from 50-60, 36% of them left early. Then out of the 94 Jrs ranked from 85-95, 38% of them left early. Out of the 101 Sophs ranked from 50-60, 8% left early while out of the 97 Sophs ranked from 85-95, 9% left early.
If it worked like your example, you would see a sharp drop in chance to leave early as you get closer and closer to the #100 ranked player but you don't so I'm trying to make sense of it so let me know your thoughts!