Community Spring Cleaning week is here! Join your fellow Maveryx in digging through your old posts and marking comments on them as solved. Learn more here!

Weekly Challenges

Solve the challenge, share your solution and summit the ranks of our Community!

Also available in | Français | Português | Español | 日本語
IDEAS WANTED

Want to get involved? We're always looking for ideas and content for Weekly Challenges.

SUBMIT YOUR IDEA

Challenge #103: Just another game?

PhilipMannering
16 - Nebula
16 - Nebula

There is evidence supporting the fact the fact that the Superbowl is less predictable. I used a linear regression model, but I wouldn't mind knowing which models I should use under which circumstances. Solution attached,

 

Spoiler
challenge_103 - Snip.PNG
Ilías
7 - Meteor

Fly Eagles Fly!

Ilías
7 - Meteor

Since the Patriots have no defense to speak of I just went with the offensive variables. Even the predicted score would not have been enough for the Pats.

 

Fly Eagles Fly!

JosephSerpis
17 - Castor
17 - Castor

Challenge Completed

Inactive User
Not applicable
 
estherb47
15 - Aurora
15 - Aurora

Standard predictive analysis here. While I'd love to be snarky and claim that I only used offensive stats because the Pats have no defense, the real reason is in the p-values. They were statistically significant.

I explored whether removing some "outlier" games had any impact - and they resulted in worse p-values, so I went with all of the data excepting the superbowl games for modelling.

Oh, and FLY EAGLES FLY. Though I am actually a Giants fan ;)  

MMehdi
8 - Asteroid

Challenge #103 resolved

 

Spoiler
I used the Association analysis tool to select 3 predictor variables.
Based in the difference between :
-  The predicted scored points VS actual points  in regular season matches
-  The predicted scored points VS actual points  in Super Bowl matches

I found that the predictions were pretty accurate for regular matches but not for Super Bowl ones, I guess it is because of the whole motivation and adrenaline levels behind those special games !




Workflow (#103)Workflow (#103)

 

SeanAdams
17 - Castor
17 - Castor

Used a few different kinds of predictive models - linear regression seemed to do best.

 

 

jasperlch
12 - Quasar

Solution attached. Thanks!

LordNeilLord
15 - Aurora

I don't use predictive tools at all so this was a bit of a struggle for me but I got there in the end

 

Spoiler
103.PNG