812FB

I noticed an advocate that beliefs or once believed Sandy Hook didn't happen is letting us all know his opinion on this matter. 

Does anyone know if the IHSAA has plans to use this facility? I would think since it's labeled multisport we might not only see boys and girls soccer games but maybe even football and lacrosse. 

 

1 hour ago, Footballking16 said:

For me, a Cinderella team would be a bottom half Sagarin rated team that beats 2-3 top half rated Sagarin teams en route to a sectional championship, akin to a 13+ seed reaching the S16 of the NCAA basketball tournament.

If the field is cut in half does a Cinderella exist? Based on what you're saying here it doesn't appear to be the case?

Just now, Footballking16 said:

58-2 is the record between top half Sagarin teams and bottom half Sagrin rated teams. 

Cinderella is a myth yet it is a huge talking point as to why the all-in exists. 

Could you please separate the record? I'd like to see the record of the games like Columbia City over Sagarin 32 seed Ft. Wayne Wayne. Or Brebeuf over 21 seed Northview. I think it'll be interesting to see how the bottom half of the qualifiers hold up to the top half.

On 10/22/2022 at 6:06 AM, Footballking16 said:

Early indications are that top half Sagarin rated teams went 58-2 against bottom half Sagarin rated teams last night in class 1A-4A. Typical, but not at all surprising.

Dekalb #39 over #29 New Haven and #40 New Castle over #31 Beech Grove both out of 4A were the two “upset” winners. Will be back later with final score analysis to show just how lopsided last nights affairs were.

Cut the field in half people, please.

I'm wondering if you could also track win-loss records of those top half of the qualifiers against each other?  I'm interested to see if it's just as lopsided. Meaning the top half still dominates and the bottom half of the schools deemed worthy to qualify for a playoff system you are proposing. Obviously you'll run into issues like New Pal v Mt. Vernon. But for example, Brebeuf over Northview is an example I would like to see. Is that 58-2 number including Brebeuf's win already? Or do you not include it?

OK I've punched some numbers with the schools (14 total) that are currently 5A or recently 5A that I would imagine Castle and Evansville North will be jockeying with for a playoff qualifier and seeding. I compiled the last three years. The 2020 season has all kinds of issues (Cancellations etc. etc.) with it being that was the season of COVID. I was thinking of possibly throwing that year out but I was halfway through the process and decided to just include it. Some of these calculations might be off due to the rush and I by no means claim that the accuracy of these digits are legit. But I believe this will provide a nice picture of the system that OHSAA uses being applied to schools like Castle and Evansville North. 

Should a qualifier apply any school ranked 8 or below would be considered the bottom half. 

Bloomington South for some reason has an 8 game schedule. I didn't take exception to the formula with them. I applied the formula as if they played a 9 game schedule. This means they missed out on points. You'll see how this jeopardizes possible seeding for them below. 

Schools bolded won a sectional. Schools in italic won the regional.  

Current Season Breakdown 

1. [8-1] Whiteland 24.777
2. [7-1] Bloomington South 19.611
3. [7-2] Franklin    19.611
4. [6-3] Terre Haute South    15.444
5. [6-3] Seymour    14.444
6. [5-4] Bedford North Lawrence    11.444
7. [5-4] Bloomington North    11
8. [5-4] Evansville North    10.277
9. [5-4] Castle    9.833
10. [4-5] Columbus East    9.388
11. [4-5] Floyd Central    8.611
12. [2-4] Terre Haute North    3.555
13. [2-7] New Albany    3.388
14. [1-7] Jeffersonville    2.5

If all of these schools were in 5A and a qualifier that cuts half the field was implemented, the SIAC might not have representation in the tournament this year. Bloomington North and BNL might also not make the cut because I didn't include other 5A schools considered in the south bracket. This is assuming there would be a 'South' and 'North' bracket of course. 

2021 Season

1. [6-2] Bloomington South 16.055
2. [7-2] Castle     15.555
3. [6-3] New Albany    14.333
4. [5-3] Whiteland    13.611
5. [6-3] Bloomington North    13.5
6. [4-5] Columbus East    8.722
7. [5-4] Evansville North    7.944
8. [4-5] Bedford North Lawrence    7.722
9. [3-6] Franklin    7.722
10. [3-5] Jeffersonville    6.611
11. [3-6] Floyd Central    6.388
12. [3-5] Seymour    6
13. [3-6] Terre Haute North    5.388
14. [1-9] Terre Haute South    0.666

Whiteland beat Columbus East twice, once in the regular season and again in the first round of the sectional.
An 8 total game schedule Bloomington South beat Castle in the first of the sectional. 

2020 Season

1. [8-1] Whiteland    22.777
2. [7-2] Columbus East    21.666
3. [7-2] Floyd Central    15.833
4. [5-3] Castle    12.833
5. [6-2] New Albany    10.777
6. [5-4] Evansville North    10.666
7. [4-3] Bloomington South    9.833
8. [4-5] Bloomington North    7.944
9. [4-5] Terre Haute North    6.388
10. [3-5] Jeffersonville    5.166
11. [3-7] Seymour    5.055
12. [2-7] Terre Haute South    4.722
13. [3-5] Franklin    3.611
14. [2-7] Bedford North Lawrence    3.333

This year is a tough one to evaluate in a substantial way due to COVID. Jeffersonville won its Sectional this year. They most likely would have been cut from the field according to the points accumulated. But again, 2020 was an anomaly.

Overall, I think it's safe to say the 5A SIAC schools probably wouldn't face any major issues with getting any sort of shaft with this type of system. Getting a 1 seed might be a challenge for them. And I would consider a 5-4 school be on the bubble anyway. But I think they'd be just fine with the current schedule they play. Which is actually a good thing for fans of the SIAC. Tango already mentioned Rivalry Week down here. He's 100% correct about the enthusiasm. The folks on the west side would probably blow a gasket if the West Side Nut Club Championship was no longer a thing because of something like this. They would absolutely lose their poop. It wouldn't be pretty. No seriously. For real.

16 hours ago, DumfriesYMCA said:

….I think I actually like this system. 
 

would be cool to actually see the numbers for Indiana.  It just seems like a pretty cut and dry way to do it….you get rewarded for winning….and if the teams you beat, can also beat good teams, you are rewarded for that too….all without actually needing to see what the team names were.  
 

it could be done listing everyone as “Team 1” all the way to “Team 1000” with no bias 

I haven't punched the numbers yet but Evansville North and Castle probably get screwed with OHSAA's system. When I get the time I plan on breaking down the area to see just how much these two 5A schools miss out on points as opposed to the 5A schools north of I-64. 

1 hour ago, Footballking16 said:

The purpose of this post wasn't necessarily to stir discussion for potential bubble teams like Harrison. Any format that cuts the field in half and/or uses a qualifier is going to leave some teams out, that's the nature of the business. That's also why it makes regular season games for meaningful. Teams can play their way in or out the last few weeks of the season.

Rather this post was designed to illustrate that the all-in isn't necessary because Cinderella runs in  Indiana High School football don't exist. Detractors of a qualifier believe that everyone should be included because the postseason signals a sign of new life and teams that were non-existent in the regular season are somehow going to turn it on come "week 10". It doesn't work that way. There will be a few upsets (5 or less) in the first round where a bottom half Sagarin rated team beats a top half Sagarin rated team and you will see some bottom half teams advance a round, sometimes two, given the nature of the blind draw that allows for bottom half Sagarin rated team to play other each other, but when the dust settles, sectionals are almost exclusively chalk. If the sectionals were seeded all the way through accordingly, the myth that Cinderella exists would instantly die. You'd have running clocks in almost every game that feature the top 2 teams playing the bottom 2 teams. 

I might be mistaken but isn't a bubble team also considered a Cinderella? I'd be interested to see not only the bottom half stats but also the stats bubble teams that would or wouldn't qualify. I'm willing to throw down some cash that you'll see running clocks with a field cut in half as well. I could be mistaken but I like my chances. So if we do indeed see the same type of stats of what you deem Cinderella teams with bubble teams should the field be cut again?

Personally, I just enjoy watching football. And I'm interested to see how 812 football does during the tournament. Just like any tournament on the planet the good teams weed out the bad teams week after week. Some teams are better than others. That's sports. 

1 hour ago, Muda69 said:

The answer is 3.66.  

Or 42.  Take your pick.

 

Well, it goes without question 42 is the answer, to everything.

I was under the assumption based on the conversation we've been having only one programed CRS (a computer) vs an application written to play chess is the discussion. I'm now at the understanding you have a desire to take averages of three separate programs (computers)? If that's the case, that somewhat balances out the most obvious defect that all CRS have in the state of Indiana. It still isn't effective and solved by "a computer" or computers. It's taking the subjective human element into solving the math objective by intervening with separate algorithms, that are not as effective due to unit-containment. This also points out the inefficiencies in taking multiple programs written to get the outcome, unlike an application being written for a game of chess and the stark differences between the two math objectives.

15 hours ago, Footballking16 said:

Computer polls are only imputing and calculating data. There’s no bias in the algorithms like you see with the human element. A computer model is much more likely to spit out an accurate reading for a team like Luers despite being 4-5 than a coach in Southern Indiana who has nothing to go by other than W-L record. You’re taking the bias out of the equation by using computers. Computers aren’t a 100% science but they’re generally better than the human element.

 

I’ve always proposed in my format that a tenth regular season game be added (for those detractors who think that high school athletes are somehow entitled to a tenth game) and the field then be effectively cut in half. That leaves landlocked conferences like the SIAC and SAC one OOC game a year.

 

Sagarin unfortunately does not factor in GS win over the team from Kentucky or any of the data from the teams they played. Hurts GS in this scenario. It significantly helped a team like Chatard in the Massey Poll who played Elder and Christian Brothers, two teams who are both highly rated in national polls.

Yeah, these are all good ideas to address the unit-containment issue. 3-6 Evansville Harrison might end up above the cut off threshold to get punched in if they'd play a game outside of the SIAC. But proposing a game to fix the unit-containment won't address Sagarin's second out of state opponent defect. Which you have pointed out regarding Gibson Southern. And all SIAC schools would absolutely have to avoid scheduling any out of state schools, which might be an issue down here in 812 SIAC land.

13 minutes ago, Footballking16 said:

Using Sagarin, yes Norwell is the #1 rated team in 3A.

I use Sagarin because its readily available and it's unique only to Indiana High School football.

If a playoff format was ever implemented using a predictive rating measure, I highly doubt Sagarin would be used as it doesn't factor in out of state competition. Massey and Calpreps do however. 

OK, that makes sense. I was wondering why you were using Sagarin as an example. I appreciate your response here. Frankly, I think all systems have their defects here in the state of Indiana. I'll cover that below. I personally could be sold on a 'take an average' of the three systems approach. However, if I'm being completely honest, I could really be sold to a 'committee of coaches' (CoC) approach. I say this because I tend to think coaches have a better pulse on the IHSAA football landscape then some dudes in the Bay Area, CA or Jefferson City, TN. But more importantly the CoC will cover the computer ranking system (CRS) defect. 

The major issue with using a CRS in the state of Indiana is there are two metro areas/conferences (SIAC and SAC) that have self-contained units. This is basically the statistical math kiss of death for any statistician on the planet. They can't solve this problem effectively. However this can be effectively solved by subjective pollsters or a CoC.

Sidenote: 
Man, using Sagarin, Gibson Southern would get the shaft at an overall 7 seed when they're ranked 3 in both Calpreps and Massey. Instead of hosting a regional game they're most likely on the road.

34 minutes ago, hhpatriot04 said:

I assume Norwell is the No. 1 or 2 team in its sectional.

What's the problem?

Is Norwell the overall #1 seed for the entire tournament?

18 minutes ago, foxbat said:

And therein lies the problem.  It's not whether a computer can do the math, it's the question of the algorithm employed ...  and, as this thread has shown, we haven't even gotten anywhere close to algorithms yet.

I cosine this hypotenuse.

14 minutes ago, Muda69 said:

Of course the algorithms would be different, Chief.  I never claimed otherwise.  But both are problems that can effectively be solved by a computer.  

Boss, which one of these programs effectively solved the problem?
3A class rank of Norwell - Sagarin: 1 Calpreps: 6 Massey: 4

 

15 minutes ago, Muda69 said:

Not really.  It all really just boils down to math.  And that's is really all a computer is, a very fast machine that is good at manipulating zeroes and ones.

 

 

Yes really. Creating algorithms for a game of chess has a completely different math objective than the subjective logic of creating a ranking system for sport teams. Excuse the pun but these are two separate objectives to solve mathematically. 

28 minutes ago, slice60 said:

Wrestling sectionals are seeded based upon a set of previously agreed-upon criteria. There is no seeding after the sectional level-- the Regional brackets are based on Sectional placement, the Semistate brackets are based on Regional placement, the State brackets are based on Semistate placement.

How is that criteria agreed-upon? A computer ranking system?

Just now, JQWL said:

I've never heard anything positive when wrestling coaches talk about how their tournament is seeded. That was actually what popped in my mind when it was mentioned here.

Sure you're gonna get complaints. Which makes sense. But are they somewhat fair seedings?

4 minutes ago, Muda69 said:

If a computer can now "solve" the game of Chess it can rank a few high school football games.

 

Yeah, apples and oranges as someone has already pointed out. 

OK, I'm hearing a complete lack of faith in a committee of coaches seeding IHSAA's football tournament but are cool with 15 coaches seeding the entire IHSAA's wrestling tournament? That just doesn't make sense to me but whatever.

Again, I'm Switzerland when it comes to this stuff. But I've got to admit, relying on a single computer ratings system (CRS) isn't something that I would want to happen. Especially when there are at least 3, that I know of, out there to use right now. And all three of them have flaws. The one CRS, in my opinion, that has the most glaring flaw (Sagarin) is the one that is being pushed or used for an example makes zero sense to me. I can't get on board having that system determining whether or not a school like Evansville Harrison is either playing this Friday or watching. I might get on board if an average rank using the combination of the three are used but that's just me. 

A lot of talk about Sagarin to solve a complex problem.

Could someone please explain to me why a computer ranking system would be used over a committee of coaches when determining seeding and/or qualifiers?

I'm Switzerland regarding this topic however I'm not sure using Sagarin or only one system in general is as an example helping ones argument.  For example does a school like 3-6 Evansville Harrison qualify for the tournament using Sagarin? In Calpres and Massey they are class ranked 33 and 30 respectively whilst ranked 34 in Sagarin. If using Massey do they qualify?  Averaging out all three systems they are class ranked 32.333. Does that make them qualify?

Most importantly, this Friday they will face a 6-3 Boonville team. This game is intriguing for this type of topic because it's not a stretch to say Harrison has a legitimate chance to win that game. However, in a qualifier, using Sagarin or Calpreps they don't even have a chance to play this Friday? 

I'm curious here. I'm under the assumption ranking systems really wouldn't play a role in determining the tournament seeding or qualifier but more so coaches accomplishing the task?

I'm posting this to keep a record on how these pan out throughout the tournament. Also, there has been discussion regarding seeding and/or qualifiers for the tournament. I've noticed Sagarin is being the most common quoted so I thought it would be interesting to see which system had a better accuracy rate at the end of the 2022 season. 

1A

2A

3A
 

4A
 

5A
 

6A
 

The 812FB Top 20 rankings is my opinion and based off not only the eye test but the averages of Sagarin, Calpreps and Massey ranking systems.

1. East Central [7-2]
2. Bloomington South [8-1]
3. Evansville Reitz [9-0]
4. Gibson Southern [9-0]
5. Linton-Stockton [9-0]
6. Lawrenceburg [8-1]
7. Owen Valley [9-0]
8. Evansville Memorial [6-3]
9. Jasper [6-3]
10. Evansville Mater Dei [6-3]
11. Southridge [8-1]
12. Castle [5-4]
13. Columbus North [4-5]
14. Evansville North [5-4]
15. Terre Haute South [6-3]
16. North Decatur [9-0]
17. Bloomington North [5-4]
18. Tecumseh [8-0]
19. Northview [6-3]
20. Boonville [6-3]

HM: South Dearborn [7-2], Vincennes [4-5], Heritage Hills [5-4], Charlestown [8-1], Sullivan [5-4], Perry Central [8-1]

Sectional Top 20 & HM Matchups
#6 Lawrenceburg [8-1] at South Dearborn [7-2]
Vincennes [4-5] at #4 Gibson Southern [9-0]
Perry Central [8-1] at #10 Evansville Mater Dei [6-3]

Game of the Week: #16 North Decatur [9-0] at Tri [8-1]
 

The 812FB Top 20 rankings is my opinion and based off not only the eye test but the averages of Sagarin, Calpreps and Massey ranking systems.

1. East Central
2. Bloomington South
3. Evansville Reitz
4. Gibson Southern
5. Linton-Stockton
6. Owen Valley
7. Southridge
8. Evansville Memorial
9. Evansville Mater Dei
10. Jasper
11. Lawrenceburg
12. Castle
13. Columbus North
14. North Decatur
15. Evansville North
16. Bloomington North
17. Vincennes
18. Boonville
19. Terre Haute South
20. South Dearborn & Tecumseh

HM: Northview, Columbus East, Heritage Hills, Charlestown, North Posey

Week 9 Top 20 & HM Matchups
#3 Evansville Reitz at #9 Evansville Mater Dei
#18 Boonville at #4 Gibson Southern
#11 Castle at #8 Evansville Memorial
#17 Vincennes at #10 Jasper
#12 Lawrenceburg at #20 South Dearborn
#19 Terre South at #13 Columbus North

Game of the Week: 4A #2 Indianapolis Roncalli at #3 East Central

Hats off to this site. Looks like a lot of work is put into it and I'm grateful to have another outlet to look up what others in 812FB think!

The 812FB Top 20 rankings is my opinion and based off not only the eye test but the averages of Sagarin, Calpreps and Massey ranking systems. 

1. East Central
2. Bloomington South
3. Evansville Reitz
4. Gibson Southern
5. Linton-Stockton
6. Owen Valley
7. Evansville Memorial
8. Evansville Mater Dei
9. Evansville North
10. Southridge
11. Lawrenceburg
12. Castle
13. Vincennes 
14. Columbus North
15. Jasper
16. Bloomington North
17. North Decatur
18. Boonville
19. Terre Haute South
20. South Dearborn

HM: Tecumseh, Northview, Heritage Hills, Sullivan, Columbus East

Week 7 Top 20 & HM Matchups
#7 Evansville Memorial at #13 Vincennes
#15 Jasper at #9 Evansville North
Heritage Hills at #10 Southridge
#16 Bloomington North at #14 Columbus North

11 hours ago, SEAL_63 said:

I'll preface by saying I too find it a little shocking to see Lincoln as high as they are when compared to the best in the state (but not necessarily complaining haha). The Alices are a legit threat in 3A no doubt, but there's a lot left to prove just in the SIAC alone, not to mention Sectional 30.

But to question VL's ranking you absolutely have to look at Reitz at #9, since the two teams are so closely tied together.  VL's near defeat of the Panthers in Week 2 seems to have carried A LOT of weight. Reitz faces Memorial this Friday followed by VL/Memorial in Week 8...so we're about to get a more realistic picture of this. 

Overall though I'm not about to question Sagarin's feel for the SIAC, as they've been low balled the past two years (as explained by @Alduflux). Me personally, I would knock off 10-15 points for SIAC teams and that would put you in the ballpark.

Massey has got it all wrong. No way that Reitz and VL are 50 spots apart. 

Yeah no knock on your Alices was intended. You got a solid ball club up there in Knox County. That QB is something special. And I agree, I personally think Sagarin has weighed the SIAC 10-15 points too high. 

19 hours ago, DumfriesYMCA said:

If we move to seeding it would entirely be up to the coaches of the given class.  Maybe they break it down by regional or semi state…who knows

 

but when coaches are asked about the topic they often can heard saying something along the lines of “if 15 wrestling coaches can sit down and seed an entire tournament across multiple weight classes with a hundred kids then I think it’s possible for football coaches to sort out seeding a sectional” 

Thanks for sharing. That would make sense.