Toggle sidebar

Where My Nerds At?

W

Wrestleknownothing

All-Conference
Oct 30, 2021
1,685
4,471
113
TheREALSpooner said:
Not here.
Click to expand...
And yet you were the first to open and respond.

Thinking Think GIF
 
  • Haha
Reactions: InTheCircle, UNIHawk and Anon1746362370
mac119

mac119

Freshman
Sep 17, 2017
41
53
18
You went through the bonus point area pretty extensively, which was good for me to see this way. But the thing I've never liked about the binary model is that it assumes, as you said, "the wrestlers will finish in the exact order of the rank/seed" rather than an expected value. Since there is such a significant difference in the placing values between a 1st place finisher and a 12th place finisher, it leads to the other comment of "the binary model tends to overvalue top eight seeds".

I assumed I was just being a homer because the expected points makes Iowa look a little closer to the pack than the binary model, so its nice to see that stats back it up (and give me hope).
 
W

Wrestleknownothing

All-Conference
Oct 30, 2021
1,685
4,471
113
mac119 said:
You went through the bonus point area pretty extensively, which was good for me to see this way. But the thing I've never liked about the binary model is that it assumes, as you said, "the wrestlers will finish in the exact order of the rank/seed" rather than an expected value. Since there is such a significant difference in the placing values between a 1st place finisher and a 12th place finisher, it leads to the other comment of "the binary model tends to overvalue top eight seeds".

I assumed I was just being a homer because the expected points makes Iowa look a little closer to the pack than the binary model, so its nice to see that stats back it up (and give me hope).
Click to expand...
Now that you mention it I guess binary is not an accurate description either. It is really a unitary outcome model. Only the seed is assumed.
 
You must log in or register to reply here.
Top Bottom