Drew Pasteur is back Up!

bkmk1

Well-known member
He is on drugs with some of his picks and win by's: Brecksville over Tallmadge by 20? Wow. Where did that pick come from.
 

Just.here

Well-known member
He is on drugs with some of his picks and win by's: Brecksville over Tallmadge by 20? Wow. Where did that pick come from.
Probably because Brecksville wasn’t bad in 19 and 20 and tallmadge was. So when his model takes into account prior years, you get things like that.
 

bkmk1

Well-known member
Probably because Brecksville wasn’t bad in 19 and 20 and tallmadge was. So when his model takes into account prior years, you get things like that.
Unacceptable way to make his predictions. Might as well not make any. They are bogus. Thanks for the how to.
 

Just.here

Well-known member
Unacceptable way to make his predictions. Might as well not make any. They are bogus. Thanks for the how to.
He knows that. He’s said if it’s up to him he wouldn’t do anything this early in the year but the people wanted it anyway.
 

bkmk1

Well-known member
He knows that. He’s said if it’s up to him he wouldn’t do anything this early in the year but the people wanted it anyway.
Got it. I'm glad to know that. That makes sense regarding many of those picks. Thanks for that clarification.
 

Harrycrane

Well-known member
Got it. I'm glad to know that. That makes sense regarding many of those picks. Thanks for that clarification.
Some of his lines (. Predictions ) are pretty accurate , some not as much . He states that he expects 70 percent picking the winner in the first week or so and closer to 85 eventually . His initial lines seem to be a bit more accurate than the cal preps projections . He is usually more accurate overall than they are
 

Quest4Gold

Well-known member
Well hearing the one guy, my advice is if you don't like it, then don't go to it. Then you won't have to see those awful predictions
 

The Infidel

Well-known member
Early on Drew isn’t accurate and he knows that. Week 5 and on he’s pretty good! My favorite thing about his site is he will have all the playoff scenarios for everyone. He’s spot on.
 

bkmk1

Well-known member
Do you just not want to have his website? Stop complaining
We just disagree. His prediction website is useless. It is not based on any thoughts, watching of games, taking into account this years players. It's just based on last year stats and wins/loses. All of which mean nothing to me regarding a true prediction. Most Yappi posters can make better predictions because they are based on knowledge of their teams. Not computer plugged in garbage.
 

Jguth83

Active member
We just disagree. His prediction website is useless. It is not based on any thoughts, watching of games, taking into account this years players. It's just based on last year stats and wins/loses. All of which mean nothing to me regarding a true prediction. Most Yappi posters can make better predictions because they are based on knowledge of their teams. Not computer plugged in garbage.
Does one guy on yappi do every game? Do you get to see all of the matchups in 1 place on yappi? Nope
 
Unacceptable way to make his predictions. Might as well not make any. They are bogus. Thanks for the how to.
He's hasn't done it since before COVID year and decides to try something new, I don't blame him but usually when it is him predicting for himself he is one of the best. Give this some time and hopefully it'll be really good
 

Rangerfan

Well-known member
I did a statistical analysis about 10 years ago of the best methods to predict the number of computer points a team wil earn in the following year. After running through various methods and combinations, I was disappointed to find that the best way to predict how a team would do in the computer points as by looking at how they did the previous year.

So boring.
 

Bless Em All

Always Watching...
At the end of the day, it’s high school football. Too many variables come into play. Predicting is an art, not a science.
These guys like Drew are doing remarkably well considering the variables and the lack of real scientific data outside of scores.
I’m guessing that anyone that is not happy with his site is not looking at this thing for what it really is…fun. Not one game will be decided by any prediction or Yappi post.
 

bigkat

Well-known member
I did a statistical analysis about 10 years ago of the best methods to predict the number of computer points a team wil earn in the following year. After running through various methods and combinations, I was disappointed to find that the best way to predict how a team would do in the computer points as by looking at how they did the previous year.

So boring.
too much time on your hands..LOL!!!
 

birdog

Well-known member
It is interesting that Pasteur has Big Moe #11 in the state regardless of Division, # 9 in D-1 and favors Massillon by 2 (giving the Tigers a 55% chance to win the game.) I think this forecast flies in the face of what most but not all posters have projected. There are many other interesting projections here, so I guess we just have to let things play out. GBM
 

TriangleMan

Well-known member
Article from today's The Repository.

Wooster professor predicts outcome of every Ohio prep football game​

Drew Pasteur poses for a photo in Taylor Hall at The College of Wooster where he is a professor and co-chair of the Mathematical & Computational Sciences Department.


Tim Botos
The Repository

WOOSTER ‒ Ah, the return of high school football.

In Ohio, that means more than crunching pads, marching bands, student cheering sections and concession stand munchies. Again, it will be a gleam from the computer screen of The College of Wooster math professor Drew Pasteur, thousands of lines of code, probability, statistics and matrix algebra.

After being sidelined for the past two seasons, Pasteur's mathematical high school football crystal ball is back in the game.

For the uninitiated, Pasteur's Fantastic50.net website, which is available for free, uses math formulas to project the outcome of every game in the state throughout the season, rank teams and analyze every school's ongoing playoff scenarios.

For a cult of coaches, players and fans, who routinely visited the website for more than a decade, its return has triggered a collective round of applause — largely because the site is so darned accurate.

The news has even prompted the creation of a thread aptly titled "Drew Pasteur is coming back!" on the football forum at Yappi.com, a popular message board for high school sports in Ohio.

On it, posters bow to Pasteur's prognosticating prowess: "Drew Pasteur doesn't pay taxes, taxes pay Drew Pasteur;" "Drew tells the weather when it’s allowed to rain" and "The sun doesn't wake up Drew every morning, he tells the sun when it's time to get up."

The 46-year-old Pasteur laughed it off.

He's no guru; he's a math geek.

"But some of the 'Moneyball' stuff has made that cooler than it was 20 or 30 years ago," Pasteur noted, referring to the popular movie that exemplified how analytics have invaded pro sports.

College of Wooster professor Drew Pasteur looks at this week's high school football matchups on his Fantastic50.net website. He uses probability and statistics to forecast the scores of every high school football game in Ohio throughout the season and playoffs.


Why did Fantastic50.net take an unexpected hiatus?​

"It was just a couple of chaotic years," Pasteur explained.

He'd operated the website for a dozen years from computers in his office on the third floor of Taylor Hall on The College of Wooster campus, and from his house across the street.

But he had to pull the plug in 2020 and 2021.

First, he came down with COVID-19. The lingering effects were so bad that Pasteur — an avid runner — could walk no more than a mile for about a year. Then, his wife, Heather, was diagnosed with cancer. Throw in taking care of their two boys and chairing the math department, and he simply didn't have the time or motivation for the website.

But he's stronger now.

His wife is in remission.

Pasteur was ready to re-launch Fantastic50.net. So this summer, he tweaked the formula. He incorporated more historical data. The results, he said, should prove even more accurate.

This week, he posted initial rankings and predictions for week one games of the more than 700 teams in Ohio. In a couple of high-profile local games, Pasteur has McKinley as a 17-point underdog vs. Mentor; Massillon is a 2-point favorite against Cincinnati Moeller.

Soon, the site will include a lot more data: A continuing look at each team's playoff scenarios, should it finish with a record of 6-4 vs. 7-3 or 8-2 and so on, as well as the mathematical probability of which opponent a team will face in the first round of the playoffs.

Joe Eitel, who runs joeeitel.com and tracks results of every game in the state to determine playoff computer rankings for the Ohio High School Athletic Association, is glad Pasteur's back.

"I had enough people emailing me and calling me, asking 'Where's Drew?' and 'What happened to Drew?'" he said.
College of Wooster professor Drew Pasteur talks about his Fantastic50.net website. He uses probability and statistics to forecast the scores of every high school football game in Ohio throughout the season and playoffs.


'I'm fascinated by the statistics behind it.'​

Eitel said Pasteur's site complements his own.

"I'm kind of a numbers geek, too, .... so I'm fascinated by the statistics behind it," he added.

Pre-pandemic, Pasteur's site would log about a million page views over the last two weeks of a typical high school football season — he said traffic spiked heavily on Friday nights, as visitors sought insight on how the just-concluded games would impact the playoff picture.

The site was so spot-on that in the past, head coaches would contact him, asking how adding a certain team to the schedule for the following season might ultimately affect playoff chances.

An untold number of coaches gobble up the site's data.

Sandy Valley head coach Brian Gamble said he and his staff are frequent Fantastic50 visitors. Especially because the Cardinals have been in Division V playoff chases in recent years.

"That (math) is way above my pay grade," Gamble said. "I'm a P.E. teacher, so I'm thankful for any help."

Matt Gulling, Tuslaw's head coach, knows the site well.

"We've looked at it for sure ... especially when you get to about weeks four or five, it's really accurate," he said.
Pasteur agreed his site's accuracy increases as the season progresses. In the past, his predictions have been better than 80% on the mark statewide during a season's latter weeks.

In simple terms, Pasteur's system can simulate outcomes of tens of thousands of scheduled matchups within a matter of minutes. Beyond that, Pasteur said the math is "too messy" to explain.
Drew Pasteur poses for a photo at The College of Wooster where he is professor and co-chair of the Mathematical & Computational Sciences Department.


How did Drew Pasteur get started?​

Pasteur never played high school football.

"I weighed about 80 pounds as a freshman," he said of his days at Enloe High, near Raleigh, North Carolina.

So, he became team statistician. Later, as a high school teacher, graduate student, athletic trainer and doctoral candidate in North Carolina, Pasteur came up with a system to rank high school football teams, no matter their size — that's where the Fantastic50 moniker came from.
"I wanted to find out how some smaller schools would do against larger schools," he said.

In Ohio, teams are divided into seven divisions, based on the number of boys in the school. Division I is the largest; Division VII the smallest. At season's end, seven state champions are crowned.

Pasteur still is not certain how the best small schools would fare against larger elite teams, because high school football remains largely regionalized, with teams mostly playing schools of similar size during the season.

That being said, small school behemoths, such as Kirtland and Marion Local, both in Division VI this season, regularly appear in Pasteur's ranking of the top-50 teams, no matter the division.

Pasteur said he believes they'd hold their own against top teams two or three divisions above. Generally, he said the gap between the best teams in Divisions II through VII is small.

The largest gap, he said, exists between Division I and II. The enrollment differences are almost staggering. The largest Division I school, including OHSAA competitive balance adjustments, has Cincinnati St. Xavier with a male enrollment of 1,691, compared to the 605 enrollment figure for Westerville South, the largest Division II school.

Pasteur doesn't make money off Fantastic50.

He doesn't condone gambling, tough he admits the best oddsmakers are those in Las Vegas.

High school football predictions, he added, can be more difficult than projecting NBA or Major League Baseball, because there are so few games and the gap between best and worst teams is far greater than in pro sports.

"How do you project the margin for a mismatch?" he asked.

From experience and dabbling with his formula the largest spread you'll see this season is 49 points, even if the math says a team is 80 points better. His tinkering has taken into account a team will typically insert reserves and resort to sportsmanship when a game gets out of hand.
Pasteur is a fan, too.

"I like the way it brings communities together; there's nothing like Friday night football," he said.

In the past, he's taken in multiple games in the area. But with his oldest son now a sophomore band member at Wooster High, his Friday nights are limited to watching the Generals.

Reach Tim at 330-580-8333 ortim.botos@cantonrep.com.On Twitter: @tbotosREP
 
.
Top