Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Odds on Favorite(According to CALPREPS) in the Six State Championship Games
#1
CAL  :    86%   chance  to cage  Bardstown's Tigers

Fred Douglass:   75%  chance to turn the Purples of Bowling Green  to a sad shade of Blue

Pikeville :    74%  chance to ram it up  Raceland's behind

Corbin :     72%   chance to show Haddix & the Rebels they ain't winning this Civil War either

Male :    65%   chance to let the Dogs feast on Bullitt East 

Beechwood :    62%   chance  to pull the wings off of Mayfield's Cardinals


NOTE:  These numbers represent the historical percent chances of a team winning when they have a given margin of victory.  For example, CAL is projected by the CP system to beat Bardstown by  16 points.  Historically, teams that have been projected to win by 16 points by the CP computer have won 86% of the time. Hence, CAL's  86% win probability.  


How many of these projections do you agree with???
[-] The following 1 user Likes Old School Hound's post:
  • Bull got out!
#2
IMO Calpreps is a joke with their predictions we probably shouldn’t even play the games just go ahead and award all of these heavily favored teams the trophy and call it a day, I think they’ll be lucky to get 50% of their predictions right.
[-] The following 3 users Like Jack Lambert's post:
  • 64SUR, Granny Bear, rookie57
#3
Pikeville wins
CAL wins
Beechwood wins
Boyle wins
Freddy D wins
Bullitt East wins
#4
I don't ever believe anything that comes from the state of California.   Fake News!  :)
[-] The following 2 users Like Rebel55's post:
  • Slic Ric, Westside
#5
Before folks go dismissing the CalPreps computer system, let me give the perspective from a pure mathematical and statistical point of view.

If you just blindly picked games throughout the season, picking the team with the better record, and in cases of equal records, picking the home team, you are going to have a percentage of correct picks of about  69.8% . 

In 2022, the CP computer system had a correct percentage in all games nationwide of 83.2% ( 42,089-8,497).  In Kentucky, it did even better with a correct percentage of  85% .  Without going into a lot of advanced statistics, confidence intervals, and statistical significance, let me just leave it at this---  anything in the 75-80% range would be considered quite good, so 85% is some damn good work by the CALPREP computer, which I nickname JTG("Jimmy the Geek"). 

Although six games is a small sample size, based on its overall work this season, "Jimmy the Geek"  should predict five of the six state final games correctly. The expected value of number of correct predictions would be 4.9  if we go by the  83.2% nationwide correct percentage for 2022  ;  5.1 if we go by the  Kentucky correct percentage for 2022.  So,  5 correct out of six looks favorable.

Based on its work predicting games in Kentucky this season , there is a 37.7% chance that it correctly predicts all six of these games and a 39.9% chance that it correctly predicts exactly five of the games correctly. Thus,  there is a 77.6% chance that the computer gets FIVE OR MORE correct !!!! NOTE: This is assuming the computer equals it's 85% correct projections in Kentucky games up to this point in the season.  However, that probability will be skewed high  because in the regular season you are going have more games that are easier to predict than these games are.  A better measure will be to use historical data , based on the percentage of teams that actually won games , given a predicted margin of victory by the computer.

When I get the time, I will compute the probability that five or more of these games will be predicted correctly based on historical data using the percentages given in the initial post. That will give us a more accurate probability than the overstated probability above.
[-] The following 1 user Likes Old School Hound's post:
  • Iam4thecats
#6
(11-28-2022, 11:41 AM)Old School Hound Wrote: Before folks go dismissing the CalPreps computer system, let me give the perspective from a pure mathematical and statistical point of view.

If you just blindly picked games throughout the season, picking the team with the better record, and in cases of equal records, picking the home team, you are going to have a percentage of correct picks of about  69.8% . 

In 2022, the CP computer system had a correct percentage in all games nationwide of 83.2% ( 42,089-8,497).  In Kentucky, it did even better with a correct percentage of  85% .  Without going into a lot of advanced statistics, confidence intervals, and statistical significance, let me just leave it at this---  anything in the 75-80% range would be considered quite good, so 85% is some damn good work by the CALPREP computer, which I nickname JTG("Jimmy the Geek"). 

Although six games is a small sample size, based on its overall work this season, "Jimmy the Geek"  should predict five of the six state final games correctly. The expected value of number of correct predictions would be 4.9  if we go by the  83.2% nationwide correct percentage for 2022  ;  5.1 if we go by the  Kentucky correct percentage for 2022.  So,  5 correct out of six looks favorable.

Based on its work predicting games in Kentucky this season , there is a 37.7% chance that it correctly predicts all six of these games and a 39.9% chance that it correctly predicts exactly five of the games correctly. Thus,  there is a 77.6% chance that the computer gets FIVE OR MORE correct !!!! NOTE: This is assuming the computer equals it's 85% correct projections in Kentucky games up to this point in the season.  However, that probability will be skewed high  because in the regular season you are going have more games that are easier to predict than these games are.  A better measure will be to use historical data , based on the percentage of teams that actually won games , given a predicted margin of victory by the computer.

When I get the time, I will compute the probability that five or more of these games will be predicted correctly based on historical data using the percentages given in the initial post. That will give us a more accurate probability than the overstated probability above.

So you are telling me there is chance!
[-] The following 1 user Likes Grannyshot's post:
  • rookie57
#7
I expected Corbin to be the favorite over Boyle, but that projection is a little higher than I was expecting.
#8
I hope Boyle can keep it close just for the sake of all the people that will be in attendance paying damn near $20 for a ticket.
#9
Again, here's what the percentages mean. The computer gives Pikeville a 10 -point margin of victory over Raceland(which seems reasonable, if not, even a little conservative). Historically, in all the hundreds of thousands of games that the CP system has projected, 74% of those teams that were rated 10- point favorites actually won those games. Hence, Pikeville is given a 74% chance of winning based on the historical data points. It's science. It's math. It's real.
#10
I agree with you old school hound but if you go in the Raceland and Pikeville thread they don’t have a chance because they are not as talented as Raceland
#11
Calpreps picked Male(first time they played in regular season), Manual, Central Hardin and Bryan Station to beat Bullitt East. Bullitt East didn't win half of them, they won ALL of them. So I doubt I'll pay much attention to the NUMBERS.
[-] The following 3 users Like BEChargers's post:
  • Bull got out!, Rebel55, toolaidback
#12
(11-28-2022, 10:58 PM)BEChargers Wrote: Calpreps picked Male(first time they played in regular season), Manual, Central Hardin and Bryan Station to beat Bullitt East.  Bullitt East didn't win half of them, they won ALL of them.  So I doubt I'll pay much attention to the NUMBERS.


Sample size of four .    

42,089 - 8,497 (83.2%)
[-] The following 2 users Like Old School Hound's post:
  • BEChargers, Iam4thecats
#13
(11-28-2022, 11:38 AM)Rebel55 Wrote: I don't ever believe anything that comes from the state of California.   Fake News!  Smile

I have the same opinion of the fruit and nut state.
#14
Calpreps is 2-1 so far and Massey is 3-0
#15
Calpreps finishes 4-2
Massey finishes 5-1
[-] The following 1 user Likes Iam4thecats's post:
  • Bull got out!
#16
(12-04-2022, 11:08 AM)Iam4thecats Wrote: Calpreps finishes 4-2
Massey finishes 5-1
Granny Bear went 6-0.  Going to keep an eye out on who she likes for 2023.  Big Grin
#17
(12-04-2022, 11:08 AM)Iam4thecats Wrote: Calpreps finishes 4-2
Massey finishes 5-1

Which games did they miss......
#18
(11-28-2022, 11:31 AM)DukeBoy Wrote: Pikeville wins
CAL wins
Beechwood wins
Boyle wins
Freddy D wins
Bullitt East wins

Nicely done
#19
(12-04-2022, 06:06 PM)BEChargers Wrote:
(12-04-2022, 11:08 AM)Iam4thecats Wrote: Calpreps finishes 4-2
Massey finishes 5-1

Which games did they miss......

Computer cant see that a team like Bullitt East is for Real!
[Video: http://https://www.youtube.com/watch?v=MYxAiK6VnXw]
[-] The following 1 user Likes Bull got out!'s post:
  • BEChargers
#20
CALPREPS losses were Corbin, Male
Massey loss was Male (https://masseyratings.com/hsf/ky/games)

Based on the original post with the historical percentages, CALPREPS picking 4 winners was not unexpected based on numbers I computed. The numbers I computed showed CALPREPS was just as likely to pick 5 or 6 wins vs 4 wins or fewer. I would have made the over/under based on CALPREPS predictions at 4.5 wins.
#21
(12-04-2022, 06:59 PM)kybred Wrote: CALPREPS losses were Corbin, Male
Massey loss was Male (https://masseyratings.com/hsf/ky/games)

Based on the original post with the historical percentages, CALPREPS picking 4 winners was not unexpected based on numbers I computed. The numbers I computed showed CALPREPS was just as likely to pick 5 or 6 wins vs 4 wins or fewer. I would have made the over/under based on CALPREPS predictions at 4.5 wins.

Thank you @kybred, I actually knew which ones they missed.  I was being a little bit of a smart a$$.  Appreciate it.
[-] The following 2 users Like BEChargers's post:
  • Bull got out!, Rebel55
#22
My dog picked Boyle over Corbin and didn’t need a computer from California to tell her. She preferred eating out of the black bowl than the red bowl.
[-] The following 1 user Likes Rebel55's post:
  • Bull got out!
#23
(12-04-2022, 10:58 PM)Rebel55 Wrote: My dog picked Boyle over Corbin and didn’t need a computer from California to tell her.  She preferred eating out of the black bowl than the red bowl.

Yep, sure.....[Image: https://i.pinimg.com/originals/c0/20/29/...ffb104.gif]
#24
(12-04-2022, 11:04 PM)Bull got out! Wrote:
(12-04-2022, 10:58 PM)Rebel55 Wrote: My dog picked Boyle over Corbin and didn’t need a computer from California to tell her.  She preferred eating out of the black bowl than the red bowl.

Yep, sure.....[Image: https://i.pinimg.com/originals/c0/20/29/...ffb104.gif]
LOL that Hilarious! That got a chuckle out of me.
[-] The following 1 user Likes Rebel55's post:
  • Bull got out!

Forum Jump:

Users browsing this thread: 1 Guest(s)