Official KHSAA RPI Rankings (10/10)

[Czar]

I’m not sure why the RPI has Johnson Central with a record of 4-0-1. We haven’t had a Covid cancellation this year. We couldn’t find a game for week 2 so it was a bye week for us. 

[NLTanker1]

Thanks a lot, Ballard. 🙄

[Voice of Reason]

I posted this information in another thread but I think it needs to be included here with more details. Hopefully this explains as simply as possible how the KHSAA RPI works.

All that matters in KHSAA RPI is wins and losses plus an added factor for game value, which is based on what class you and your opponent play in. Here is the quick summary:

- RPI is made up of winning percentage, opponents' winning percentage and opponents' opponents' winning percentage. Wins and losses are what matter. 

- In the winning percentage column, you get a zero if you lose the game. If you win the game, divide the class factor of your opponent by your class factor and that is your winning percentage for that game. Here is the factor used for each class:

• 1A = 1.323
• 2A = 1.521
• 3A = 1.749
• 4A = 2.011
• 5A = 2.313
• 6A = 2.660

- Game scores have no impact on RPI. 

- Quality of your opponent has no impact on RPI.

Here is a link with the details of the KHSAA's formula with an example of how they calculate the RPI:

https://khsaa.org/resources/RPI/rpiexample2018fbapb.pdf

 

[Voice of Reason]

Quick example of the winning percentage game factor:

If you play in 3A and you beat a 1A team, divide the 1A factor 1.323 by your 3A factor 1.749 and your winning percentage for that game is 0.756. 

If you beat another 3A team, 1.749/1.749=1. Your winning percentage for that game is 1. 

If you beat a 6A team, 2.660/1.749=1.52.

So you can see it is better to beat a team in a higher class (6A) than a team in a lower class (1A). Quality of opponent does not matter. A team gets the same number of winning percentage points for beating 5A Seneca as they do for beating 5A Douglass.

That is why it is best for RPI to schedule the highest class opponents you think you can beat. Remember, you get a zero for losing so you need to schedule teams you think you can beat.

 

[colonel-fan]

14 hours ago, Czar said:

I’m not sure why the RPI has Johnson Central with a record of 4-0-1. We haven’t had a Covid cancellation this year. We couldn’t find a game for week 2 so it was a bye week for us. 

I believe they are treating byes the same as a Covid cancellation game. CovCath also didn't play a game last week are are listed as 4-0-1.

[Yadi1957]

Maybe this year they will just draw for the playoffs?

[snakesnot_2000]

I know Mayfield is not #1 in 2A but if they can stay somehow at least to #2 they can host a semi game. One of the other teams, Somerset or LCA will be gone so it's a chance. I can see Mayfield finishing 9-1 in the regular season.

[gchs_uk9]

On 10/11/2020 at 9:12 AM, Voice of Reason said:

So you can see it is better to beat a team in a higher class (6A) than a team in a lower class (1A). Quality of opponent does not matter. A team gets the same number of winning percentage points for beating 5A Seneca as they do for beating 5A Douglass.

This, in my opinion, is the part of the RPI that needs to be fixed immediately. Teams should be incentivized for stepping up in competition. It is already hard enough for some schools in the state to find opponents willing to play. If you give schools an advantage if they play and/or beat better teams, maybe more teams are willing to answer the phone when Trinity, Covington Catholic, Johnson Central, etc. come calling.

[ATLCat]

26 minutes ago, gchs_uk9 said:

This, in my opinion, is the part of the RPI that needs to be fixed immediately. Teams should be incentivized for stepping up in competition. It is already hard enough for some schools in the state to find opponents willing to play. If you give schools an advantage if they play and/or beat better teams, maybe more teams are willing to answer the phone when Trinity, Covington Catholic, Johnson Central, etc. come calling.

Yep..

The out of state component of the RPI makes zero sense to me as well. With services such as Maxpreps/Calpreps.. there is no reason why an OOS opponent should be judged as a .500 team playing all .500 teams. This isn't 1984 and we are waiting on a newspaper to be mailed from another state and hoping a score is correct in it.

Quality of win to me is the biggest miss of the system and has been since Day 1.

If I could waive a magic wand, here are the changes I would pursue:

1. Use Calpreps for a more accurate out of State ranking. If a school schedules an out of state foe, they can submit evidence of classification. Meaning if Ashland wants bonus points for playing a 6A equivalent school when they defeat George Washington, WV... Ashland can submit proof before the game documenting that George Washington's enrollment would make them a 6A school in Kentucky.

2. Three District results are used per team. If you lose a District game it MUST be one of those three results. If you go undefeated only the top three highest RPI value games are used and the lowest value(s) of wins are dropped. For the fourth value, either the team has an option of using the omitted District victory, or they can choose to use for a second time the lowest RPI result from their Non-District schedule. This helps teams who have no choice but to play in five team Districts against weaker competition as opposed to teams who play in stronger Districts. For teams in 3-team Districts (see Hazard/Pikeville).

3. Each result gets a bonus for a neutral site victory and a larger bonus for a road victory. This would need to be a minimal multiplier in conjunction with whatever "Quality of Opponent" value is assigned. Simply put, if Mercer County goes to CAL and beats them it should be weighted more heavily than if DeSales hosts CAL and beats them. 

4a. Quality of victory would give a multiplier bonus to each result for wins greater than 3 TD. You cant reward on shear size of victory or that has negative consequences, but beating a team 3 TD's should weigh more heavily than someone else beating them by 7 points.

4b. Same thing for quality of loss, you get penalized more for losing by 3 TD's + than losing by 6-17 points, or losing by less than 6 points. 

5a. Quality of opponent needs to be addressed as mentioned above. The first variable I would reward is you get a positive bonus multiplier if the team you defeat reached the seeding round of the playoffs the year before (won their District). That multiplier increases if they won their region or reached the Finals.

5b. On the final calculation, a bonus is awarded for NON-DISTRICT wins versus teams in the Top 5 and Top 10 respectively in their class in RPI.

6. Finally, for teams who are unable to find someone to fill a 10th game, facing teams who forfeit, or in the case of this year face a COVID cancellation. There is a one game exemption where they can choose to either duplicate the value of their "best loss" or their "worst win". COVID exemptions withstanding, if you are unable to play a 10-game schedule it unfairly effects a team since this is a culminative rating. 

 

 

[ColonelMike]

At this point in a messed-up COVID season, there are NO numerical ratings worth the paper they are printed on at this time.

For that matter, at this point in ANY season, there are no numerical ratings worth posting.    Not enough data.

[Voice of Reason]

48 minutes ago, ColonelMike said:

At this point in a messed-up COVID season, there are NO numerical ratings worth the paper they are printed on at this time.

For that matter, at this point in ANY season, there are no numerical ratings worth posting.    Not enough data.

 I disagree. Five games into the season gives you enough to have a pretty good read on most teams.

[theguru]

51 minutes ago, ColonelMike said:

At this point in a messed-up COVID season, there are NO numerical ratings worth the paper they are printed on at this time.

For that matter, at this point in ANY season, there are no numerical ratings worth posting.    Not enough data.

 

[gchs_uk9]

50 minutes ago, ColonelMike said:

For that matter, at this point in ANY season, there are no numerical ratings worth posting.    Not enough data.

No numerical ratings are worth posting at this point? Which of these from Calpreps should be deleted immediately because they are egregiously wrong?

[Cheese]

I like where it shows strength of schedule. 

[Hearsay]

Some members placing way, way too much emphasis on "who you could schedule" when criticizing RPI.  There are so many factors in scheduling, including distance, cost, school schedule, where you're spending money for your other school teams to go, team compatibility, not to mention it is a zero-sum gain only for lower class teams.  I would love for Bardstown to take a crack at Belfry, but we went to Pikeville last year and played pretty well, but the cost was very high for us.  I would love for Bardstown to try their hand at Male or Boyle (we used to play them years ago), but why would they play us? Not to mention should we risk getting trounced/injured just for a shot at a fraction higher RPI? Its really not at simple as saying "just pick up the phone and schedule better competition."