Score predictions for Central Florida | Page 3 | The Boneyard

Score predictions for Central Florida

Status
Not open for further replies.
rvwsleep said:
?????
How is that possible ?
While a good entry, not nearly as close as "SueBird". Pull out the calculator.
Expand ...
yeah I understand your margin but... I was looking at nailing the UConn score, it gets no better than that!
 
DaddyChoc said:
yeah I understand your margin but... I was looking at nailing the UConn score, it gets no better than that!
Expand ...

I understand this is not a huge deal but shouldn't there be some standard to base the winner on ?
And shouldn't that standard be the same week to week ?
I disagreed with the initial criteria of point differential to determine the winner but I acquiesced. Now, it seems as though you only have to pick UConn's score correctly to win.
How about this ? Next game we all get to choose a winner based on whatever criteria we choose !! That will really liven up the thread.
 
rvwsleep said:
I understand this is not a huge deal but shouldn't there be some standard to base the winner on ?
And shouldn't that standard be the same week to week ?
I disagreed with the initial criteria of point differential to determine the winner but I acquiesced. Now, it seems as though you only have to pick UConn's score correctly to win.
How about this ? Next game we all get to choose a winner based on whatever criteria we choose !! That will really liven up the thread.
Expand ...
That's what I do, I didn't creat this thread... the we have a winner pic is MY thang based on MY opinion. I actually led the predictions about 5 yrs ago but got lazy so it kind of went away but Winlots has rejuvenated it but I'm stuck with my own rules. The exact UConn score on a UConn board gotta account for something. I don't like the MOV style cause you could say 36-0 and win if the score is actually 84-48
 
I suggest (I thought someone else did earlier) that the three criteria be these:
First - margin -2 pt
second - UConn score 2 pts
Third - opponent's score 1 pts
This criteria allows for the breaking of ties and avoids the absurdity of someone missing the margin by twenty, but winning by nailing the UConn score. This also avoids the 36-0 problem cited above.
 
Last edited:
NewCorn said:
I suggest (I thought someone else did earlier) that the three criteria be these:
First - margin
second - UConn score
Third - opponent's score
This criteria allows for the breaking of ties and avoids the absurdity of someone missing the margin by twenty, but winning by nailing the UConn score. This also avoids the 36-0 problem cited above.
Expand ...

If you want an automated unbiased approach, simply use the rules from statistics

Define fractional error for each team as fe=[abs (score_guess - score)] / score
where abs stands for absolute value. i.e abs(-6) = 6.

Then define your score S as S=(1-fe) so a perfect guess score is S= 1

Then add the score of both teams guesses S_total=S_conn+S_opponent

A perfect score is 2 while errors in the guess (normalized by the true score) reduce the score in a controlled way.

In addition, this scoring mechanism is additive so one could imagine keeping a cumulative score over the season.
 
.-.
Is it really that important to "crown" a winner? We are all just guessing at the score. I think lots of people were close with UCONN in the mid to high 80's and UCF in the mid 40's. I think people are realizing that there is a rather substantial drop off when UCONN plays with 3 or 4 of the reserves and only 1 or 2 starters (see the 4th quarter). But I also think it's clear that as much as this team has "passed" every test, they are not going to dominate for 30+ minutes every game.

They are terrific and have probably exceeded literally everyone's expectations, but they are not winning by being dominating in every aspect (rebounding for example). Certainly a different experience for the fans, and fun for those who cherish closer games (by the way, I am NOT one of these people- I love seeing us crush EVERYONE)...
 
ericsandiego said:
Is it really that important to "crown" a winner? We are all just guessing at the score. I think lots of people were close with UCONN in the mid to high 80's and UCF in the mid 40's. I think people are realizing that there is a rather substantial drop off when UCONN plays with 3 or 4 of the reserves and only 1 or 2 starters (see the 4th quarter). But I also think it's clear that as much as this team has "passed" every test, they are not going to dominate for 30+ minutes every game.

They are terrific and have probably exceeded literally everyone's expectations, but they are not winning by being dominating in every aspect (rebounding for example). Certainly a different experience for the fans, and fun for those who cherish closer games (by the way, I am NOT one of these people- I love seeing us crush EVERYONE)...
Expand ...


I am not suggesting there is much skill in any guess. But if we have this competition, its fun to keep score.
 
eebmg said:
If you want an automated unbiased approach, simply use the rules from statistics

Define fractional error for each team as fe=[abs (score_guess - score)] / score
where abs stands for absolute value. i.e abs(-6) = 6.

Then define your score S as S=(1-fe) so a perfect guess score is S= 1

Then add the score of both teams guesses S_total=S_conn+S_opponent

A perfect score is 2 while errors in the guess (normalized by the true score) reduce the score in a controlled way.

In addition, this scoring mechanism is additive so one could imagine keeping a cumulative score over the season.
Expand ...
Good. This clearly works well and establishes with precision the relative accuracy of each poster's guess. Perhaps this is what would work best here.
I do think, however, that for those of us who were looking out the window at that girl walking by during the math lecture, the method I mentioned earlier is easier to follow. And perhaps it accomplishes much the same thing when my edit is included to give 2pts for margin, 2pts for UConn score, and 1pt for the opponent's score.
 
rvwsleep said:
While evidently not the winner, I wanted to recognize the MOST ACCURATE entry. Well done.
Expand ...
that's your winner... and no one will disagree it
 
NewCorn said:
Good. This clearly works well and establishes with precision the relative accuracy of each poster's guess. Perhaps this is what would work best here.
I do think, however, that for those of us who were looking out the window at that girl walking by during the math lecture, the method I mentioned earlier is easier to follow. And perhaps it accomplishes much the same thing when my edit is included to give 2pts for margin, 2pts for UConn score, and 1pt for the opponent's score.
Expand ...


I guess I never avoided math lectures and statistics is a good friend of mine. I imagine my approach is clear enough for anyone who wants to ponder it, it is easy to implement on an excel sheet for example etc. Of course, not my decision :rolleyes:

I would like to understand your approach. Are you saying, out of all entries, the person with best margin gets 2 pts,
the person closest to the UConn score gets 2 pts and the closest to the opponents score gets one?

If so, cant person A get 2 pts for best margin, person B get 2 points for closest UConn score and Person C get the 1 pt for
closest opponent score so we have a tie? :confused::confused:
 
.-.
eebmg said:
I would like to understand your approach. Are you saying, out of all entries, the person with best margin gets 2 pts,
the person closest to the UConn score gets 2 pts and the closest to the opponents score gets one?

If so, cant person A get 2 pts for best margin, person B get 2 points for closest UConn score and Person C get the 1 pt for
closest opponent score so we have a tie? :confused::confused:
Expand ...
I need to explain further. The closest to the margin is the winner. The UConn score is the first tie-breaker. The opponent score is the second tie-breaker. But as I have indicated, your method is excellent. I'm only explaining my idea, not pushing it.
 
Status
Not open for further replies.

Forum statistics

Threads
168,359
Messages
4,567,320
Members
10,469
Latest member
xxBlueChips


Top Bottom