header pic

Perhaps the BEST B1G Forum anywhere, here at College Football Fan Site, CFB51!!!

The 'Old' CFN/Scout Crowd- Enjoy Civil discussion, game analytics, in depth player and coaching 'takes' and discussing topics surrounding the game. You can even have your own free board, all you have to do is ask!!!

Anyone is welcomed and encouraged to join our FREE site and to take part in our community- a community with you- the user, the fan, -and the person- will be protected from intrusive actions and with a clean place to interact.


Author

Topic: HFA in general and this season

 (Read 2505 times)

medinabuckeye1

  • Legend
  • ****
  • Default Avatar
  • Posts: 8906
  • Liked:
HFA in general and this season
« on: August 25, 2021, 08:07:12 PM »
The thing about HFA that a lot of people don't think about is that it isn't static.  Rather, it varies by team.  

Years ago we did some data mining regarding home vs away records in an effort to determine which stadiums were the toughest road environments.  Our conclusions were surprising at first.  The teams with the biggest gaps between their home and road records weren't the teams with the stadiums that we typically think of as the loudest but they did tend to be middling teams.  


After thinking it through, this made sense to me.  If you think about it, HFA is only likely to be decisive in games between relative equals.  Ie, tOSU is ranked #1 and Illinois is ranked #14 in our current power rankings.  if those two teams played this year (they don't) it is safe to assume that nobody here would need to know where the game was to be played to come up with the prediction that #1 Ohio State should obviously beat #14 Illinois (in Columbus, Champaign, or on Mars, doesn't matter).  

HFA is more likely to be the determining factor in games among relative equals.  Ie, if #1 tOSU played #2 UW, more posters would be likely to pick UW if the game was in Madison.  Similarly, if #14 IL played #13 MSU a lot of us would want to know whether the game was in Champaign or East Lansing before making our prediction.  

Thus, it makes sense that HFA is biggest for teams closest to the middle of the pack.  Think of it this way:

  • Assume that we were going to play a 26-game double-round-robin
  • Further assume that the home team will win all games against teams up to three spots better than the visitor

Here is how that would play out:

Note that HFA is only decisive in three games each for the best and worst teams (tOSU and IL) while it is decisive in four games for the second best and second worst (UW and MSU) and five games for the third best and third worst (PSU and RU).  Meanwhile, HFA is decisive in six games each for the eight teams in the middle.  They each lose three games to inferior teams on the road and win three games over superior teams at home.  

In the real world, of course, it isn't nearly this simple.  For one thing, in the real world the gaps aren't uniform and also teams have good and bad days both at home and on the road but the above still provides a general explanation for why HFA is likely to be a bigger deal for a middling team than for a contender or a bottom feeder.  

Back to HFA being not static.  What I mean is that it isn't the same for #11 Purdue as it is for #4 Indiana.  What is best for each team is to play the teams most equal to them at home and the teams least equal to them on the road.  Assuming that our rankings are reasonably accurate then, fans of #4 Indiana should prefer:
  • #3 PSU at home
  • #5 Iowa at home
  • #2 UW at home
  • #6 M at home
  • everybody else on the road.  
Meanwhile, fans of #11 PU should prefer:
  • #10 UNL at home
  • #12 RU at home
  • #9 UMD at home
  • #13 MSU at home
  • everybody else on the road.  

Note from this that it is theoretically in BOTH teams' best interest to play the Bucket game on the road this year.  If our rankings are accurate than realistically #4 IU is going to beat #11 PU regardless of whether the game is in Bloomington or West Lafayette.  Both teams are theoretically better off playing that game on the road and reserving their HFA for games where it is more likely to be decisive (PSU, IA, UW, M for IU; UNL, RU, UMD, MSU for PU).  


This year's B1G schedule:


What jumps out at you?

medinabuckeye1

  • Legend
  • ****
  • Default Avatar
  • Posts: 8906
  • Liked:
Re: HFA in general and this season
« Reply #1 on: August 26, 2021, 02:58:57 PM »
I didn't think this post would generate much discussion but I thought that at least @betarhoalphadelta was enough of a statistics/cfb nerd to jump in . . .

847badgerfan

  • Administrator
  • Hall of Fame
  • *****
  • Posts: 25215
  • Liked:
Re: HFA in general and this season
« Reply #2 on: August 26, 2021, 03:05:39 PM »
What jumps out at me is it seems UW has a favorable mix of H/A here.
U RAH RAH! WIS CON SIN!

Honestbuckeye

  • Team Captain
  • *******
  • Posts: 5796
  • Liked:
Re: HFA in general and this season
« Reply #3 on: August 26, 2021, 03:27:49 PM »
First of all great analysis Medina.

On the Michigan schedule I think you have it wrong for the Ohio State game because it’s at Michigan this year.

Following up that it looks to me like Michigan does catch a break with most of their tough games against better teams being at home.

The badgers have it pretty tough like 847 just said.
Get your facts first, then you can distort them as you please.
-Mark Twain

medinabuckeye1

  • Legend
  • ****
  • Default Avatar
  • Posts: 8906
  • Liked:
Re: HFA in general and this season
« Reply #4 on: August 26, 2021, 03:40:45 PM »
First of all great analysis Medina.

On the Michigan schedule I think you have it wrong for the Ohio State game because it’s at Michigan this year.

Following up that it looks to me like Michigan does catch a break with most of their tough games against better teams being at home.

The badgers have it pretty tough like 847 just said.
Thank you.  


My chart reads the other way.  Look at the column not the row, ie:
  • tOSU is HOME for PSU on 10/30
  • tOSU is AT IU on 10/23
  • tOSU is AT M on 11/27
  • etc
And the Badgers . . .
What jumps out at me is it seems UW has a favorable mix of H/A here.
Agreed, their schedule is remarkably favorable at least according to our preseason Power Rankings.  They are #2 and:
  • miss #1 tOSU
  • get #3 PSU at home
  • miss #4 IU
  • get #5 IA at home
  • get #6 M at home
  • visit #7 MN
  • get #8 NU at home
  • miss #9 UMD
  • get #10 UNL at home
Only one road game against a team in the top-10 in the league?  Wow!


medinabuckeye1

  • Legend
  • ****
  • Default Avatar
  • Posts: 8906
  • Liked:
Re: HFA in general and this season
« Reply #5 on: August 26, 2021, 03:46:44 PM »
Following up that it looks to me like Michigan does catch a break with most of their tough games against better teams being at home.
I don't know what to think of Michigan's schedule.  We have them ranked #6 but they have the highest standard deviation of rankings indicating that our votes are all over the place.  Here is their schedule:
  • tOSU at home, 11/27
  • UW away, 10/2
  • PSU away, 11/13
  • IU at home, 11/6
  • IA miss
  • n/a
  • MN miss
  • NU at home, 10/23
  • UMD away, 11/20
  • UNL away, 10/9
  • PU miss
  • RU at home, 9/25
  • MSU away, 10/30
  • IL, miss

If they perform up to their recruiting then getting tOSU at home is a big break because maybe they can win that and they should be able to beat everybody else regardless of location.  

If they perform more as they have been, then I see this as a pretty rough schedule because they probably can't beat tOSU even at home then they get their next two toughest opponents (UW and PSU) on the road in two VERY loud stadiums so that almost makes 6-3 a ceiling.  


betarhoalphadelta

  • Global Moderator
  • Hall of Fame
  • *****
  • Posts: 12188
  • Liked:
Re: HFA in general and this season
« Reply #6 on: August 26, 2021, 03:56:33 PM »
I didn't think this post would generate much discussion but I thought that at least @betarhoalphadelta was enough of a statistics/cfb nerd to jump in . . .
I'm just waiting to see if Purdue is good enough for HFA to matter. If we're really the 11th team in the conference this year, it's gonna be a LONG season.

Brohm eventually needs to start earning those damn paychecks.

BuckeyeAvenger

  • Recruit
  • **
  • Posts: 97
  • Liked:
Re: HFA in general and this season
« Reply #7 on: August 26, 2021, 04:01:54 PM »

Nice analysis Medina!

Frankly, most discussions of HFA treat it as a static constant number, that applies to them at home. I see the wisdom in how it effects “in the middle” teams more than teams at the top, or bottom. A lot of podcasters and talking heads are expecting a pronounced HFA this season, as a reaction to finally having fans in the seats, again. I assume you disagree, looking at it from a strictly statistical point of view?

longhorn320

  • Legend
  • ****
  • Posts: 9329
  • Liked:
Re: HFA in general and this season
« Reply #8 on: August 26, 2021, 04:04:03 PM »
I didn't think this post would generate much discussion but I thought that at least @betarhoalphadelta was enough of a statistics/cfb nerd to jump in . . .
Im missing the point of this discussion

is it to say HFA exists or doesnt exist

is it for a betting guide
They won't let me give blood anymore. The burnt orange color scares the hell out of the doctors.

Brutus Buckeye

  • Hall of Fame
  • *****
  • Posts: 11238
  • Liked:
Re: HFA in general and this season
« Reply #9 on: August 26, 2021, 04:17:14 PM »
The CFL will provide a lot of data points, even more than usual, as they are mostly playing against their own division this year because of Covid. Many teams will be playing one another thrice during the regular season, with maybe a fourth meeting during the playoffs. 
1919, 20, 21, 28, 29, 31, 34, 35, 36, 37, 42, 44
WWH: 1952, 54, 55, 57, 58, 60, 61, 62, 63, 65, 67, 68, 70, 72, 74, 75
1979, 81, 82, 84, 87, 94, 98
2001, 02, 04, 05, 06, 07, 08, 09, 10, 12, 13, 14, 15, 16, 17, 18, 19

medinabuckeye1

  • Legend
  • ****
  • Default Avatar
  • Posts: 8906
  • Liked:
Re: HFA in general and this season
« Reply #10 on: August 26, 2021, 04:35:34 PM »
Im missing the point of this discussion

is it to say HFA exists or doesnt exist

is it for a betting guide
HFA definitely exists.  If you look at records, nearly every team has a better winning percentage at home than on the road even if you exclude OOC games where part of that is due to the bigger/better teams generally hosting.  Here, for example, are Ohio State's H/A records against league opponents over 50 years from 1970-2019:
  • Michigan:  .660 at home, .500 away
  • Illinois:  .682 at home, .826 away EXCEPTION
  • Wisconsin:  .762 at home, .690 away
  • Indiana:  .958 at home, .929 away
  • Minnesota:  .952 at home, .947 away
  • Michigan State:  .722 at home, .789 away EXCEPTION
  • Northwestern:  .941 at home, .938 away
  • Iowa:  .889 at home, .781 away
  • Purdue:  .933 at home, .611 away
  • Penn State (includes some non-league games pre-1993):  .812 at home, .571 away
  • Rutgers:  1.000 at home, 1.000 away
  • Nebraska:  1.000 at home, .667 away
  • Maryland:  1.000 at home, 1.000 away

Looking at 13 league opponents:
  • Against nine of them the Buckeyes did better at home
  • Against two of them the Buckeyes did the same H/A
  • Against two of them the Buckeyes did better on the road

No, it isn't for a betting guide, it is for discussion of the projected impact of HFA on THIS season.  So, for example, we discussed that Wisconsin's schedule appears to be quite favorable because they get nearly all of the teams where HFA might be decisive at home.  


medinabuckeye1

  • Legend
  • ****
  • Default Avatar
  • Posts: 8906
  • Liked:
Re: HFA in general and this season
« Reply #11 on: August 26, 2021, 10:02:15 PM »
My comparison of Ohio State's winning percentages at home and on the road vs league opponents was something I thought I should explore more so here I have ignored the three newest members (UNL, RU, UMD) because there are not enough games for the comparison to be statistically valid and ranked each of the 10 league teams that have been in the league for at least half of the last 50 years by the size of the gap between tOSU's home record against them and tOSU's road record against them:

Three oddities jump out at me:

  • Purdue:  It just makes no sense for Ohio State to have THAT big of a HFA against the Boilermakers over the 50 years from 1970-2019.  In the vast majority of those seasons the Buckeyes were vastly better than the Boilermakers so HFA should have been irrelevant.  Nevertheless, there they are with, statistically, the biggest HFA against tOSU.  Note also that for no apparent reason the Boilermakers have the third best home record against the Buckeyes behind only Michigan and Penn State.  
  • Michigan State:  The negative variance is odd but the Buckeyes went 13-5 against them in Columbus and 15-4 against them in East Lansing so we are really only talking about a 1.5 game difference.  Over the course of 50 years that is just some randomness (like tOSU's phenomenal 1998 and 2015 teams both losing at home to MSU).  
  • Illinois:  The negative variance is odd and this one is more than a couple games.  The Buckeyes went 15-7 against them in Columbus and 19-4 in Champaign.  Most of this stems from an odd three-year stretch from 1992-1994 during which the home team went 0-3.  If you flip the results of those three games then tOSU would be 17-5 against them in Columbus and 18-5 against them in Champaign.  That is still a negative variance but only barely.  

Other than that I think the chart is pretty much what you would expect.  Penn State, Michigan, Iowa, and Wisconsin are near the top because they are programs that, over the 50 years examined here most frequently had teams good enough for HFA vs the Buckeyes to possibly be decisive while NU, MN, and IU are at the bottom because for the vast majority of the 50 years examined here the Buckeyes would have beaten those teams no matter where the games were played.  


As a double-check, here is the same chart for Purdue:

Four things jump out at me:

  • Ohio State:  See above.  
  • Michigan:  Same reasoning as tOSU.  
  • Illinois:  Maybe Illinois just sucks at home in general because they are better against both tOSU and PU on the road?  OTOH, the difference is not significant being 11-12 at home and 11-10 on the road.  
  • Wisconsin:  Like Illinois, the gap is not large, 6-13 at home vs 7-13 on the road.  


Seeing Illinois with a negative variance against both tOSU and PU got me curious so I did the chart for them as well:


It is odd that they are better against four of the ten on the road but note that the negative variances are generally smaller than the positive variances against the other six teams so they are still substantially better overall at home than on the road.  I think that the negative variances for Illinois against tOSU and Michigan are mostly a product of small sample sizes in terms of Illinois wins over the Buckeyes and Wolverines.  In the 50 years looked at here the Illini played the Buckeyes and Wolverines a combined 86 times and went 16-68-2.  That is not a lot of wins so if they just happened to have good seasons and/or catch tOSU/M on bad seasons in Columbus/Ann Arbor then that creates this oddity.  

betarhoalphadelta

  • Global Moderator
  • Hall of Fame
  • *****
  • Posts: 12188
  • Liked:
Re: HFA in general and this season
« Reply #12 on: August 27, 2021, 10:19:10 AM »
Possible explanation...

Teams with apathetic fan bases have more significant HFA in "big" games that the fans show out for, compared to generic conference games where attendance and excitement is tepid?

medinabuckeye1

  • Legend
  • ****
  • Default Avatar
  • Posts: 8906
  • Liked:
Re: HFA in general and this season
« Reply #13 on: August 30, 2021, 09:59:23 AM »
Possible explanation...

Teams with apathetic fan bases have more significant HFA in "big" games that the fans show out for, compared to generic conference games where attendance and excitement is tepid?
That could explain why Purdue has such a dramatic HFA bump against Ohio State and Michigan but they have almost zero HFA bump against Penn State and it really falls apart when we look at Illinois.  The Illini have "negative HFA" against both Ohio State and Michigan.  

Maybe Purdue and Illinois fans are both somewhat apathetic but in a different way, perhaps:
  • Purdue fans are generally apathetic but show up for big games (ie, Ohio State/Michigan)
  • Illinois fans are completely apathetic so when Ohio State or Michigan come to town they end up taking over the stadium?
Just a thought?


 

Support the Site!
Purchase of every item listed here DIRECTLY supports the site.