Trackwrestling State Qualifier Predictions

tw · February 15, 2019, 06:36:27 PM

Trackwrestling has used a data model to predict the outcomes of the sectional tournaments tomorrow and in turn the state qualifiers from each weight class in each division.

Division 1
http://www.trackwrestling.com/tw/PortalPost.jsp?postId=1298415132

Division 2
http://www.trackwrestling.com/tw/PortalPost.jsp?postId=1298434132

Division 3
http://www.trackwrestling.com/tw/PortalPost.jsp?postId=1298457132

This is just meant to add to the excitement for tomorrows eventS and give people something to talk about...

Micah · February 15, 2019, 06:54:30 PM

Can't wait to see state brackets and match-ups.

A Barnett vs Tijerina qtr final could be fun

Boggy · February 15, 2019, 11:33:33 PM

Quote from: Micah on February 15, 2019, 06:54:30 PM
Can't wait to see state brackets and match-ups.

A Barnett vs Tijerina qtr final could be fun

Majors can be exciting, I agree

Stripes · February 16, 2019, 05:37:59 AM

Quote from: Boggy on February 15, 2019, 11:33:33 PM
Quote from: Micah on February 15, 2019, 06:54:30 PM
Can't wait to see state brackets and match-ups.

A Barnett vs Tijerina qtr final could be fun

Majors can be exciting, I agree

😆

Micah · February 16, 2019, 04:26:02 PM

Quote from: Stripes on February 16, 2019, 05:37:59 AM
Quote from: Boggy on February 15, 2019, 11:33:33 PM
Quote from: Micah on February 15, 2019, 06:54:30 PM
Can't wait to see state brackets and match-ups.

A Barnett vs Tijerina qtr final could be fun

Majors can be exciting, I agree

😆

Or how about Henschel vs Whitehead qtrs

tw · February 16, 2019, 06:37:22 PM

Division I numbers are in, and the data model accurately predicted 173 of the 224 qualifiers. This translates to an accuracy percentage of 77% meaning on average less than 4 of the qualifiers in each weight class were not predicted to make it.

Stripes · February 16, 2019, 06:46:06 PM

Quote from: tw on February 16, 2019, 06:37:22 PM
Division I numbers are in, and the data model accurately predicted 173 of the 224 qualifiers. This translates to an accuracy percentage of 77% meaning on average less than 4 of the qualifiers in each weight class were not predicted to make it.

Stop posting here and release those state brackets.

tw · February 16, 2019, 07:14:31 PM

We have more than 1 person that works for TW now...

I am not the one taking care of building the brackets and releasing them, but with Division 3 the data model accurately predicted 128 of the 168 qualifiers which translates to an accuracy percentage of 76%.

tw · February 16, 2019, 07:32:42 PM

And finally Division 2 came in with the exact same accuracy count as Division 3. The model predicted 128 of the 168 qualifiers in Division 2 as well.

Jimmy · February 16, 2019, 10:50:54 PM

When grading students,76% is a D- . Not very impressive.

kpugh8680 · February 16, 2019, 10:57:51 PM

Quote from: Jimmy on February 16, 2019, 10:50:54 PM
When grading students,76% is a D- . Not very impressive.

Glad I didn't have your grading system in school. 76% is a C, so average.

crossface21 · February 16, 2019, 11:09:00 PM

What was the formula for actually coming up with these predictions?

babywhales · February 17, 2019, 06:51:37 AM

Models provide valid information , yet all models are still just that and not correct.

Important variables are selected yet some are deemed inessential.

They do not replace real life.

R2 in the high 70's is good.

With some refinement it may go up.

Nevertheless, heart, determination and will can never be a variable inputted into a model and will forever prove to be a challenge when predicting wrestling match outcomes

CLC FAN · February 17, 2019, 08:09:54 AM

Quote from: kpugh8680 on February 16, 2019, 10:57:51 PM
Quote from: Jimmy on February 16, 2019, 10:50:54 PM
When grading students,76% is a D- . Not very impressive.

Glad I didn't have your grading system in school. 76% is a C, so average.

Ugh... comments like this bother me so much. School grading is how many you can get correct when reciting known information that you've been taught. Projecting future events is completely different. Success is defined by the context of the situation you are in. You can't just transplant one definition of a successful percentage to another situation. If the Brewers win 120 games out of 160 some odd, will that be a ho-hum C of a season?

Jimmy · February 17, 2019, 05:03:24 PM

Kpugh, that wasn't my grading system . It was the grading system of the school I went to and many others back in the early 80's . 75-81=D,82-87=C,88-92=B,93-100=A. Below 75 was failing . If 75 is now average it explains why the USA has slid in the world rankings of academia. Expect less and you get less.