KHSAA - A flaw found in the implementation of the RPI double play down exemption for football

[Toothpick]

I just saw this on KHSAA site.  Was this on here and I missed it? https://khsaa.org/10-06-23-football-playoff-schedule-request-and-football-rpi-clarification-and-clean-up/

[gchs_uk9]

It was within the link discussing Class 6A being wide open this season, although I think it's better to have it's own thread.

 

[theguru]

Here is what the KHSAA released:

FOOTBALL PLAYOFF SCHEDULING REQUEST

As teams start qualifying for the playoffs, we are asking for schools’ help in scheduling, especially in the first round. As all are aware, our supply of officials is not at the levels we were during the early 2000s when we had abundant numbers. Many of our associations, including some of our larger groups, have struggled to return to even their pre-COVID legals in terms of numbers. As such, staffing that first round which is always difficult, is likely very problematic, especially if all teams insist on trying to play on the same Friday night. Our assigners are very conscious of the need to only use those that feel are “playoff ready”, and really need your help on these dates.

There are distinct advantages to playing on Thursday or Saturday that negate the usual concerns about different preparation weeks, etc. Teams seem to be very adept at moving games during fall breaks, collegiate conflicts, etc. but don’t want to move for playoffs. Simply put, something needs to give. Even if all 96 games (up from 91 a year ago) only used five officials, that is 480 playoff-caliber officials, and that is without requests to play with seven. Last year, nearly 20 first-round games and many use that number exclusively. And that is a time when some of our officials who also work collegiate contests have to hold that weekend for playoff assignments at that level as well. 

The answer is fairly easy. Work with your opponent to play on Thursday or Saturday. If in fact, we get into a situation where we can’t staff the adequate number of officials, we will literally draw a class number from a hat and select a game or games that need to be moved. We don’t want it to get to that point. The students getting to play is far more important than any other concern.

As you move toward the playoffs this year, please have those discussions within your schools and explore options.

FOOTBALL RPI CLARIFICATION AND CLEAN-UP

At 1 p.m. ET today, the revised RPI calculations were posted on the KHSAA website and will continue to be updated hourly. There have been no changes in the calculations for the sports of soccer, volleyball, field hockey, basketball, baseball or softball.

However, a discovery was made this week of a flaw in the implementation of the double play-down exemption for football. Thankfully, with hundreds of games of live data, more tests on the data could be performed. Thanks to the watchful eye of a couple of our larger schools, this flaw was easy to trace and revise.

The beauty of a very public system is that it is both transparent and replicable. One of our assistant coaches who happens to teach match at Trinity High School, asked an innocent enough question about how a team in the largest class could have an OWP greater than 1.0. While mathematically possible in an extremely unusual situation, it certainly shouldn’t have applied to so many teams in the largest class. We are grateful that simple curiosity by Coach Michael Fox and his report to the office was able to be verified and the appropriate revisions made long before any impact on final standings would have been applied.

In review, it was determined that the class factor had been erroneously applied to the OWP and OOWP calculations when the change to the 2 play-down exemption was implemented. This error has been fixed and the revised calculations appropriately reflect both the 2 play-down exemption and the appropriate use of the class factor. Obviously, due to the potentially mathematically circular nature of applying that factor down the line, it has never been a part of the calculations and was inadvertently made for all schools regardless of class, illustrating the problem.

If you are curious about the calculations, remember there are two helpful links.

• First, is the main page for the RPI athttps://ly.khsaa.org/da5a.
• Secondly, you can go to the RPI page for the current season, athttps://ly.khsaa.org/yymgand click on any school, and basically “drill down” to any calculation.
• That’s the very basics of transparency.

We are very appreciative of Coach Fox and each of our member school coaches and assistants who have built “replicators” for the RPI to help in your planning and analysis and apologize for any inconvenience this may have caused.

[TimeWillTell]

Oh there are a ton of flaws.

I don’t understand how LCA is third? They beat CAL who beat Manual (6A RPI #1) who has beaten  SW (5-A #1) and St X.
 

I guess what I’m trying to say is this whole thing is messed up. 

[Runcible Owl]

Kudos to the KHSAA for the transparency in admitting this flaw. Not matching a new policy to the algorithm was an oversight. I'm certainly no KHSAA apologist, but I'll give them the Mulligan. 

Apparently there are math wizards - how cool is it that a Trinity assistant coach, also a math teacher, was able to drill down into the minutiae when he sensed something was amiss. 

My take on RPI - It's doing (especially now) exactly what it's designed to do. The formula is clearly laid out for all to see. There is no "magic number" like CalPreps or Cantrell, or "data overload" like Massey, to keep everyone guessing, "How does that work?"

This year, like the previous five, and like all ratings/rankings, the RPI gets dialed in as the season progresses. The preseason CalPreps ratings were a head scratcher to me, just like every year, because there's a "magic number" carried over from the previous season.

Anyway, bragging rights and championships are won on the field. The angst over who is going to have home field advantage for two games, way, way, down the road is, imo, unnecessary. Plenty of teams have overcome the "hardship" of not being in the RPI top four, six, or even eight, and made it to the finals.

[gchs_uk9]

57 minutes ago, Runcible Owl said:

My take on RPI - It's doing (especially now) exactly what it's designed to do. The formula is clearly laid out for all to see. There is no "magic number" like CalPreps or Cantrell, or "data overload" like Massey, to keep everyone guessing, "How does that work?"

This year, like the previous five, and like all ratings/rankings, the RPI gets dialed in as the season progresses. The preseason CalPreps ratings were a head scratcher to me, just like every year, because there's a "magic number" carried over from the previous season.

In your opinion, which set of rankings are more accurate as of today in Class 6A:

Calpreps

1. Trinity
2. Male
3. Douglass
4. Manual
5. St. Xavier
6. Bryan Station
7. Ballard
8. Bullitt East
9. Ryle
10. PRP

KHSAA/RPI

1. Manual
2. Male
3. Bullitt East
4. Bryan Station
5. Trinity
6. St. Xavier
7. Simon Kenton
8. Ballard
9. PRP
10. Ryle

[16thBBall Fan]

7 minutes ago, gchs_uk9 said:

In your opinion, which set of rankings are more accurate as of today in Class 6A:

Calpreps

1. Trinity
2. Male
3. Douglass
4. Manual
5. St. Xavier
6. Bryan Station
7. Ballard
8. Bullitt East
9. Ryle
10. PRP

KHSAA/RPI

1. Manual
2. Male
3. Bullitt East
4. Bryan Station
5. Trinity
6. St. Xavier
7. Simon Kenton
8. Ballard
9. PRP
10. Ryle

Well this only goes to the Top 10 in each in CalPreps Douglass is 3rd and the RPI they are all the way down in 14th place behind #10 3-4 Ryle #13 2-5 Henderson Co. and #12 Tates Creek who is 4-3 but Douglass defeated earlier this season by a score of 43-0.

Two of the losses that Frederick Douglass has had one is against the #1 rated team in Ohio (all Classes) and currently ranked #32 in the nation and a loss to the #6 ranked team in Indiana.

[Runcible Owl]

11 minutes ago, gchs_uk9 said:

In your opinion, which set of rankings are more accurate as of today in Class 6A:

Calpreps

1. Trinity
2. Male
3. Douglass
4. Manual
5. St. Xavier
6. Bryan Station
7. Ballard
8. Bullitt East
9. Ryle
10. PRP

KHSAA/RPI

1. Manual
2. Male
3. Bullitt East
4. Bryan Station
5. Trinity
6. St. Xavier
7. Simon Kenton
8. Ballard
9. PRP
10. Ryle

Both serve their purpose. To that end, both are, equally "accurate."

[gchs_uk9]

1 minute ago, 16thBBall Fan said:

Well this only goes to the Top 10 in each in CalPreps Douglass is 3rd and the RPI they are all the way down in 14th place behind #10 3-4 Ryle #13 2-5 Henderson Co. and #12 Tates Creek who is 4-3 but Douglass defeated earlier this season by a score of 43-0.

Two of the losses that Frederick Douglass has had one is against the #1 rated team in Ohio (all Classes) and currently ranked #32 in the nation and a loss to the #6 ranked team in Indiana.

Which I believe proves the point that the RPI, while perhaps more transparent and without a starting number, is still extremely flawed when a team like Douglass is being punished for playing an extremely tough schedule.

[gchs_uk9]

3 minutes ago, Runcible Owl said:

Both serve their purpose. To that end, both are, equally "accurate."

Douglass is the 14th best team in Class 6A because the KHSAA can't find out what Archbishop Hoban and Indianapolis Cathedral's records are. That is far from accurate and that's the one with playoff implications.

[theguru]

25 minutes ago, gchs_uk9 said:

In your opinion, which set of rankings are more accurate as of today in Class 6A:

Calpreps

1. Trinity
2. Male
3. Douglass
4. Manual
5. St. Xavier
6. Bryan Station
7. Ballard
8. Bullitt East
9. Ryle
10. PRP

KHSAA/RPI

1. Manual
2. Male
3. Bullitt East
4. Bryan Station
5. Trinity
6. St. Xavier
7. Simon Kenton
8. Ballard
9. PRP
10. Ryle

When I look at it Calpreps is clearly superior and it is not even close. 

[Runcible Owl]

12 minutes ago, 16thBBall Fan said:

Well this only goes to the Top 10 in each in CalPreps Douglass is 3rd and the RPI they are all the way down in 14th place behind #10 3-4 Ryle #13 2-5 Henderson Co. and #12 Tates Creek who is 4-3 but Douglass defeated earlier this season by a score of 43-0.

Two of the losses that Frederick Douglass has had one is against the #1 rated team in Ohio (all Classes) and currently ranked #32 in the nation and a loss to the #6 ranked team in Indiana.

1. You left out the loss to Corbin? 

2. Win margin makes zero difference in RPI calculations. If it did, and out of state teams counted, a 6-45 loss to Hoban wouldn't help.

3. Risk/reward. What better prepares the Broncos for post season play - the best in out of state competition, or "lesser" in state teams? 

4. How many of the teams current ahead of FD are going to make it to the third round when RPI comes into play?

5. Teams can win or lose out of state regional/national games and win championships, e.g. Trinity.

6. Decisions were made, knowing the consequences of playing out of state teams. Cause and effect for RPI purposes.

7. I think Douglass is a great team. I wish them nothing but success. I like that they're representing Kentucky against powerhouse teams, and I believe those experiences will serve them well come playoff time.

[Runcible Owl]

39 minutes ago, gchs_uk9 said:

Douglass is the 14th best team in Class 6A because the KHSAA can't find out what Archbishop Hoban and Indianapolis Cathedral's records are. That is far from accurate and that's the one with playoff implications.

It's not that they can't, they don't have to.

[16thBBall Fan]

I agree with a lot of what you are saying, but the better teams can go on the road and win. I just think they are better ways than the RPI especially for 6A teams. I think that good teams can win but it is not fair to the school or a school's student when everyone knows by watching teams that they a school is a lot better than another school, but they have to go on the road when they and their school deserves home games. Also, unless they are some big changes in the next few weeks there will be some 6A teams in northern KY advancing and teams that could beat any of those schools by double digits having to play each other based on RPI.

I know as an old man I probably did not explain that very well. It can make a difference a few years ago based on RPI in 5A Douglass had to go to Southwestern to play instead of at home it was a mud bowl. Douglass won the game by 1 point but I'm fairly confident that if that game had been a home game and played on the turf it would have not been near that close.

I am sure they are other examples in other classes. I just provide the ones I am familiar with. I just think if you have a program that can be used to measure all schools no matter the state it would be a better system than what KHSAA is currently using. If I understand this statement today, they were mistakes found by coaches I don't think you would have to worry about something like this if they were using CalPreps.

[Maroon Beech]

Concerning the first point of rescheduling Playoff games due to a shortage of game officials, this is quickly becoming an issue. Here in Tennessee, 1/2 of the Regions (Districts in Kentucky) must play their last Regular season game on Thursday and these same Regions must also play their first round game on the following Thursday. The 3rd and 4th place teams in our Region #7 (suburbs of Nashville) will have to travel 3 1/2 hours away to the Memphis area (Region #8) on a Thursday night. They should be home by 2am or so Friday morning, ready to go to class at 7:30. I fear this shortage of referees situation is only going to get worse.