quick little winning % by strength of schedule going way back

[Millhouse]

I'm throwing out Gerg and Covid. I'm going back to Maloney and I'm only looking at buckets with multiple seasons

this SOS is measure is from college football reference, it's scaled to be a measure per game. an sos of 5 is 5 points harder per game than an sos of 0

When SOS is less than 4, winning percentage is 58%

When it's >=4, winning percentage is 42%

Schedule doesn't explain everything, coaching explains a lot of course

Mac had all tough schedules EDIT WRONG Mac's first 4 were hard, last 5 were easy
Maloney had all but one
Pasqualoni only had one
Marrone had none
Shafer had none
Babers had two (first two)

Babers has been soft out of conference so overall the schedules don't look bad but the tough parts of the schedule decimate his teams. Kind of a new problem. But overall schedule strength doesn't explain babers's struggles.

 

[Cusefan95]

Mac had all tough schedules
Pasqualoni only had one

I would not have predicted this. I thought the narrative was Mac did as well as he did because we were scheduling more cupcakes, then some of P's struggles were due to overscheduling (like, uh...'98). Doesn't look like data supports that.

 

[FrancoPizza]

How did Mac have tough schedules? 87 for instance they only played 1 ranked team the entire season. PSU.

 

[Millhouse]

How did Mac have tough schedules? 87 for instance they only played 1 ranked team the entire season. PSU.

I screwed up! I'm sorry. Numbers are right but i eyeballed the coaches wrong.

Mac was tough at the beginning, easier at the end

 

[MadNY3]

I screwed up! I'm sorry. Numbers are right but i eyeballed the coaches wrong.

Mac was tough at the beginning, easier at the end

Tough/hard is kind of relative, no?

First four years, he was winning 2-4 games a year and we were a bottom 30 team.

Final couple years, he built a Top 15 program.

Schedule strength gets easier when the team goes from suck to pretty good.

 

[Eric15]

How did Mac have tough schedules? 87 for instance they only played 1 ranked team the entire season. PSU.

1984 was the hardest schedule in program history. 1981 4th hardest. 1982 6th hardest. But then 1987 was the 72nd hardest.

 

[Millhouse]

Tough/hard is kind of relative, no?

First four years, he was winning 2-4 games a year and we were a bottom 30 team.

Final couple years, he built a Top 15 program.

Schedule strength gets easier when the team goes from suck to pretty good.

Strength of schedule is determined by how well the teams you play do against the teams they play (and so on)

georgia won the last two national championships and had the 8th and 11th highest SOS so being great doesn't make your schedule strength easier. year before bama had the 3rd hardest schedule and won it all

here's a nerdy explanation. the college football reference site limits the margin of victory and treats close wins the same as 7 point wins, I believe. but once they do that adjustment, the SOS matrix works the same way from there

A very simple ranking system » Pro-football-reference.com blog » Blog Archive

 

[Millhouse]

I think of SOS like a point spread. We can't afford to give points away and if we're playing a schedule 5 points harder than average - we can't really overcome that generally.

now if you do over come it, like in 2001, you end up with a really great season where we were the 9th best team in the country with the 13th best record. Dwight Freeney had the best season of any player in Syracuse history, I think. What would we have been without him...

 

[Cusefan95]

From the linked article:

"...when you tweak the method to strengthen its weaknesses, you also weaken its strengths. In particular, if you use a modified margin of victory, the numbers don't have as nice an interpretation."

That really gets to one of the issues with statistics generally, you can often manipulate them to produce the results you want. This is a chronic issue in scientific research right now, where scientists are showing an amazing ability to find the results they want (or the people funding the research want them to find...) - but its also leading (IMHO) to the replication crisis in the sciences.

Syracuse has successfully shaken my belief that aggregated analytics can get you close to a projected point spread this year. Starting with Clemson, we've deviated so far from what an expected point spread should be based on any publicly available modelling I could find...it seems to hold better for NFL (maybe because there's more betting so wisdom of the masses applies better in the NFL than in college?). My best guess is the four wins led modelling to massively overvalue Syracuse, people figured out we were garbage and the modelling was wrong - and it took until THIS WEEK for the computer models to catch up to Syracuse's "true value". (True value" meaning "CDRW doesn't see the field").

 

[FrancoPizza]

Syracuse has successfully shaken my belief that aggregated analytics can get you close to a projected point spread this year.

I wouldn’t go that far. Starting with Clemson the internal dynamics inside the team collapsed and those types of things cannot be modeled by computers.

nothing is perfect.

 

[mscanlo]

Strength of schedule is determined by how well the teams you play do against the teams they play (and so on)

georgia won the last two national championships and had the 8th and 11th highest SOS so being great doesn't make your schedule strength easier. year before bama had the 3rd hardest schedule and won it all

here's a nerdy explanation. the college football reference site limits the margin of victory and treats close wins the same as 7 point wins, I believe. but once they do that adjustment, the SOS matrix works the same way from there

A very simple ranking system » Pro-football-reference.com blog » Blog Archive

Georgia had the 11th hardest schedule last year? Does that include the SEC Championship?

Don't get me wrong, they played Oregon and Tennessee (i.e. two top ten teams) during the regular season, but outside of that, I am struggling with who was good?

2021 season I can buy a top 10 SOS

 

[Cusefan95]

Georgia had the 11th hardest schedule last year? Does that include the SEC Championship?

Don't get me wrong, they played Oregon and Tennessee (i.e. two top ten teams) during the regular season, but outside of that, I am struggling with who was good?

2021 season I can buy a top 10 SOS

South Carolina and Mississippi State both finished the year ranked. Another big factor is they didn’t play any terrible teams. Everyone in the SEC had at least 5 win, GT had 5 wins…even their FCS opponent Sanford went 11-2. Decent number of top teams with the rest being middling is going to be a strong SOS.

 

[OrangeTarheel]

I wouldn’t go that far. Starting with Clemson the internal dynamics inside the team collapsed and those types of things cannot be modeled by computers.

nothing is perfect.

I’m curious what you mean by internal dynamics? Like when starters get injured and the backups don’t have the same established rapport with the rest of the team?

 

[Millhouse]

From the linked article:

"...when you tweak the method to strengthen its weaknesses, you also weaken its strengths. In particular, if you use a modified margin of victory, the numbers don't have as nice an interpretation."

That really gets to one of the issues with statistics generally, you can often manipulate them to produce the results you want. This is a chronic issue in scientific research right now, where scientists are showing an amazing ability to find the results they want (or the people funding the research want them to find...) - but its also leading (IMHO) to the replication crisis in the sciences.

Syracuse has successfully shaken my belief that aggregated analytics can get you close to a projected point spread this year. Starting with Clemson, we've deviated so far from what an expected point spread should be based on any publicly available modelling I could find...it seems to hold better for NFL (maybe because there's more betting so wisdom of the masses applies better in the NFL than in college?). My best guess is the four wins led modelling to massively overvalue Syracuse, people figured out we were garbage and the modelling was wrong - and it took until THIS WEEK for the computer models to catch up to Syracuse's "true value". (True value" meaning "CDRW doesn't see the field").

If you think the problems with statistics are bad, wait until you see the problems with not using them

I didn't know what model you were using.

 

[Millhouse]

Georgia had the 11th hardest schedule last year? Does that include the SEC Championship?

Don't get me wrong, they played Oregon and Tennessee (i.e. two top ten teams) during the regular season, but outside of that, I am struggling with who was good?

2021 season I can buy a top 10 SOS

It includes all their games

 

[TheCusian]

I'm throwing out Gerg and Covid. I'm going back to Maloney and I'm only looking at buckets with multiple seasons

this SOS is measure is from college football reference, it's scaled to be a measure per game. an sos of 5 is 5 points harder per game than an sos of 0

When SOS is less than 4, winning percentage is 58%

When it's >=4, winning percentage is 42%

Schedule doesn't explain everything, coaching explains a lot of course

Mac had all tough schedules EDIT WRONG Mac's first 4 were hard, last 5 were easy
Maloney had all but one
Pasqualoni only had one
Marrone had none
Shafer had none
Babers had two (first two)

Babers has been soft out of conference so overall the schedules don't look bad but the tough parts of the schedule decimate his teams. Kind of a new problem. But overall schedule strength doesn't explain babers's struggles.

It’s a part of the equation, just like any measure of any coach in CFB. It’s got too many teams with too small a data point year to year.

There are a lot of other issues that absolutely play a part. NE declining in FB talent, recruiting becoming more national, etc. Bad coaching or mediocre coaching with mediocre talent with some injuries will kill any one.

This job needs an exceptional coach that is a good fit for the region.