• Hi Guest, want to participate in the discussions, keep track of read/unread posts access private forums and more? Create your free account and increase the benefits of your eGriz.com experience today!

Just how good are eGrizzers at predicting scores?

CrunchGriz

CrunchGriz

Well-known member
I get a kick out of the wild range of scores predicted here each week for the upcoming game, so I thought I'd check out how accurate all of you/us are at predicting the final scores. Here are the results from this season's games (UC Davis predictions to date, all others include all non-joke predictions, with the means rounded to the nearest point, and Griz scores shown first in pairs):

South Dakota:
Mean: 35-24 (score difference from actual: -3 for victor)
Median: 35-24
Mode: 31(!)-17(!)
Standard Deviation: 6.82 Griz, 5.95 Opponent

North Alabama:
Mean: 38-19 (score difference from actual: -25 for victor)
Median: 28-20
Mode: 38-17(!)
Standard Deviation: 7.46 Griz, 5.24 Opponent

Oregon:
Mean: 16-54 (score difference from actual: +6 for victor)
Median: 16-56
Mode: 10-63
Standard Deviation: 8.53 Griz, 11.71 Opponent

Monmouth:
Mean: 46-15 (score difference from actual: +11 for victor)
Median: 45-16
Mode: 45-17
Standard Deviation: 8.64 Griz, 4.93 Opponent

UC Davis:
Mean: 34-32
Median: 34-34
Mode: 31-35
Standard Deviation: 8.32 Griz, 9.35 Opponent

Conclusions: I'll let someone who understands this better than I give you an explanation of what the statistics say, but in general it looks like Griz fans are fairly swayed by what happened the week before (guessing similar scores for North Alabama and South Dakota, with Monmouth's total very similar to both of the preceding FCS teams, and Oregon getting a huge bump from the Nevada game the week before UM-UO). Fans here tend to make wildly more divergent score guesses for the projected victor as opposed to the projected loser, which may not be surprising considering how lopsided the projected score differences have been for every game except the UC Davis game.

Interesting tidbits: Not a single prognosticator had the Griz scoring 61 or better in the UNA game, and similarly no one gave Monmouth a score higher than 25. Only the whimsical maroon-glasses wearing handful held the Ducks under 36.

In 228 prognostications for the first four games, only one game was guessed exactly right, by a single person.

And it looks like we think UCD and the Griz are going to overtime. ;)
 
Nice work. Maybe if you have some free time you could do the same on bobcatnation - see which board is more crazy. We could then be part of the Town Pump brawl of the wild competition.
 
So it's safe to say that predicting a game's outcome is really hard? :D

On a semi-related note, STATS has JMU at Elon as the FCS game of the week. I thought for sure that honor would go to our game at UC Davis, but I guess the coaching staff connection between JMU and Elon is a better story. STATS predicts a 27-21 JMU win. Any guesses/thoughts as to the outcome of that game versus UM/Davis? I predict that JMU will take Elon to the woodshed with a score more like 35-17 while our game will be a hard fought battle and a score around 38-35 (and hopefully Griz win). We'll see how wrong I am on Saturday! ;)
 
CrunchGriz said:
I get a kick out of the wild range of scores predicted here each week for the upcoming game, so I thought I'd check out how accurate all of you/us are at predicting the final scores. Here are the results from this season's games (UC Davis predictions to date, all others include all non-joke predictions, with the means rounded to the nearest point, and Griz scores shown first in pairs):

South Dakota:
Mean: 35-24 (score difference from actual: -3 for victor)
Median: 35-24
Mode: 31(!)-17(!)
Standard Deviation: 6.82 Griz, 5.95 Opponent

North Alabama:
Mean: 38-19 (score difference from actual: -25 for victor)
Median: 28-20
Mode: 38-17(!)
Standard Deviation: 7.46 Griz, 5.24 Opponent

Oregon:
Mean: 16-54 (score difference from actual: +6 for victor)
Median: 16-56
Mode: 10-63
Standard Deviation: 8.53 Griz, 11.71 Opponent

Monmouth:
Mean: 46-15 (score difference from actual: +11 for victor)
Median: 45-16
Mode: 45-17
Standard Deviation: 8.64 Griz, 4.93 Opponent

UC Davis:
Mean: 34-32
Median: 34-34
Mode: 31-35
Standard Deviation: 8.32 Griz, 9.35 Opponent

Conclusions: I'll let someone who understands this better than I give you an explanation of what the statistics say, but in general it looks like Griz fans are fairly swayed by what happened the week before (guessing similar scores for North Alabama and South Dakota, with Monmouth's total very similar to both of the preceding FCS teams, and Oregon getting a huge bump from the Nevada game the week before UM-UO). Fans here tend to make wildly more divergent score guesses for the projected victor as opposed to the projected loser, which may not be surprising considering how lopsided the projected score differences have been for every game except the UC Davis game.

Interesting tidbits: Not a single prognosticator had the Griz scoring 61 or better in the UNA game, and similarly no one gave Monmouth a score higher than 25. Only the whimsical maroon-glasses wearing handful held the Ducks under 36.

In 228 prognostications for the first four games, only one game was guessed exactly right, by a single person.

And it looks like we think UCD and the Griz are going to overtime. ;)
Click to expand...

Can you proved standard error please. It is more relevant here than SD.
 
grizpsych said:
CrunchGriz said:
I get a kick out of the wild range of scores predicted here each week for the upcoming game, so I thought I'd check out how accurate all of you/us are at predicting the final scores. Here are the results from this season's games (UC Davis predictions to date, all others include all non-joke predictions, with the means rounded to the nearest point, and Griz scores shown first in pairs):

South Dakota:
Mean: 35-24 (score difference from actual: -3 for victor)
Median: 35-24
Mode: 31(!)-17(!)
Standard Deviation: 6.82 Griz, 5.95 Opponent

North Alabama:
Mean: 38-19 (score difference from actual: -25 for victor)
Median: 28-20
Mode: 38-17(!)
Standard Deviation: 7.46 Griz, 5.24 Opponent

Oregon:
Mean: 16-54 (score difference from actual: +6 for victor)
Median: 16-56
Mode: 10-63
Standard Deviation: 8.53 Griz, 11.71 Opponent

Monmouth:
Mean: 46-15 (score difference from actual: +11 for victor)
Median: 45-16
Mode: 45-17
Standard Deviation: 8.64 Griz, 4.93 Opponent

UC Davis:
Mean: 34-32
Median: 34-34
Mode: 31-35
Standard Deviation: 8.32 Griz, 9.35 Opponent

Conclusions: I'll let someone who understands this better than I give you an explanation of what the statistics say, but in general it looks like Griz fans are fairly swayed by what happened the week before (guessing similar scores for North Alabama and South Dakota, with Monmouth's total very similar to both of the preceding FCS teams, and Oregon getting a huge bump from the Nevada game the week before UM-UO). Fans here tend to make wildly more divergent score guesses for the projected victor as opposed to the projected loser, which may not be surprising considering how lopsided the projected score differences have been for every game except the UC Davis game.

Interesting tidbits: Not a single prognosticator had the Griz scoring 61 or better in the UNA game, and similarly no one gave Monmouth a score higher than 25. Only the whimsical maroon-glasses wearing handful held the Ducks under 36.

In 228 prognostications for the first four games, only one game was guessed exactly right, by a single person.

And it looks like we think UCD and the Griz are going to overtime. ;)
Click to expand...

Can you proved standard error please. It is more relevant here than SD.
Click to expand...

Standard Errors (first numbers in each pair are for Griz, second for opponent):

South Dakota:

0.896 0.781

North Alabama:

0.898 0.630

Oregon:

1.172 1.609

Monmouth:

1.247 0.712

UC Davis:

1.189 1.335
 
You must log in or register to reply here.

Latest posts

Back
Top