Thoughts About RPI (Check My Math)

[KidSilverhair]

While there's certainly excitement about the changes to playoff qualifying/seeding now taking every game into account, including non-district games, it seems to me the state's RPI system is actually going to make non-district games more important than district contests (in a way), at least when it comes to getting that wild card.

Obviously your own team's performance continues to be the top factor, as it should be. Heck, win all 5 of your district games and you're in the playoffs, no matter what happens (even if you're 5-4 overall). But for those teams fighting for the 7 (or whatever number) wild card spots, the results of your non-district games (and the non-district games of your other district members) are what's going to be the difference.

With 6-team districts, at the end of the season the district's overall winning percentage must be .500. It just is ... there will be 15 wins and 15 losses total in all district games. So, a team finishing 4-1 in the district (.800 winning percentage) will see their district opponents' win percentage at .440 (11-14). If you finish 3-2, .600, the rest of your district will be 12-13, .480. These numbers will ALWAYS be true, for EVERY district with 6 members across the state. Therefore, the district contribution to 75% of your RPI is unchanging, based on how you do in the district. A team in D-2 that finishes 4-1 in their district will have the exact same amount of RPI from their district games as a team that's 4-1 in D-7, or D-1, or any other district. It doesn't matter if you've got two teams at 4-1, or two or three teams at 1-4 ... every team across the state with an identical district record is tied in RPI as far as district games go.

So the only thing that's going to actually change RPI (for those teams with identical district records) is non-district games. Obviously doing better in the non-district to get a better overall record is the main thing (8-1, 4-1 in a district is getting a winning percentage boost over going 6-3, 4-1). But what you also want is for your other district members to do well in their non-district games, which increases their winning percentage, which helps your RPI.

Now in the big scheme of things this isn't rocket science. You want to make the playoffs and get a good seed? Win your games, that'll still do it. But the RPI also introduces this incentive to root for the other teams in your district to do well in non-district games, because it will help everybody's RPI. A district that has more teams going 3-1 or better in non-district games is going to have a big RPI advantage over a district with multiple teams going 0-4 or 1-3.

Anyway, outside of actually winning the district title, your district games turn out to be actually less important (sort of) than how your entire district does in non-district games. Just seems interesting, I think. But I'm weird that way.

 

[cjm3260]

In 8-Man the state chooses your out of district games. So they have the ability to "load" schedules or make them less friendly to the RPI System.

 

[Lonetreehawk]

In 8-Man the state chooses your out of district games. So they have the ability to "load" schedules or make them less friendly to the RPI System.

Normally I would say that state officials wouldn't let individual team's concerns influence their decision-making. Yet once again Regina & West Branch are in different districts, despite being a mere 7-8 miles down the same street from one another. Alledgedly, that has been due to the protests of one of their long-term coach's who wants it that way.

 

[screwloose]

In 8-Man the state chooses your out of district games. So they have the ability to "load" schedules or make them less friendly to the RPI System.

The state sets the schedule in all classes, with some input from the schools. But you're right historically. Opponents can be set that are traditionally stronger, therefore giving RPI benefit. However, each year is different and independent of the past. So a team that usually finishes no better than 2-7, could suddenly go 6-3. It's high school football. A player here or there can have a huge impact, including record. I understand what you're saying, but it's no 100% accurate.

 

[cjm3260]

A couple years ago it was said that the schools in 4A through A classes set their own non district games. Is that still true?

 

[Reasoned]

A couple years ago it was said that the schools in 4A through A classes set their own non district games. Is that still true?

My understanding is the schools get to "request" who they'd like to play in ND games (of course it helps if the AD's from both schools discuss ahead of time and request each other), the state tries to accommodate.

 

[cidhawkeye]

Normally I would say that state officials wouldn't let individual team's concerns influence their decision-making. Yet once again Regina & West Branch are in different districts, despite being a mere 7-8 miles down the same street from one another. Alledgedly, that has been due to the protests of one of their long-term coach's who wants it that way.

And one school gets 131 mile road trips and much longer on average travel while the other has a max distance that is shorter than the other schools average. And one gets a perpetual top 5 program in their district. Makes perfect sense.

 

[Iabeastmode]

And one school gets 131 mile road trips and much longer on average travel while the other has a max distance that is shorter than the other schools average. And one gets a perpetual top 5 program in their district. Makes perfect sense.

The difference between the Regina district and the West Branch district will be interesting to watch from an RPI standpoint. WB has 4 good teams and 2 bad teams in their district as opposed to IC which has 2 good teams and 4 bad teams. Will be more important for Regina and DNH to produce good non district games. They have to go in thinking that they will have a 0-9, 1-8, 2-7, and a 5-4 team in their district. I’m sure Regina will have Solon and DNH will have A-P. But the state will control the other 3 games.

 

[cidhawkeye]

The difference between the Regina district and the West Branch district will be interesting to watch from an RPI standpoint. WB has 4 good teams and 2 bad teams in their district as opposed to IC which has 2 good teams and 4 bad teams. Will be more important for Regina and DNH to produce good non district games. They have to go in thinking that they will have a 0-9, 1-8, 2-7, and a 5-4 team in their district. I’m sure Regina will have Solon and DNH will have A-P. But the state will control the other 3 games.

What happens if the state gives Regina or DNH 4 average to below average ND games, they lose to the other on a missed extra point, finish as a runner up in the district with the 15-15 opponents record in district, can they be left out?

 

[Iabeastmode]

What happens if the state gives Regina or DNH 4 average to below average ND games, they lose to the other on a missed extra point, finish as a runner up in the district with the 15-15 opponents record in district, can they be left out?

That’s what I’m wondering.

 

[KidSilverhair]

What happens if the state gives Regina or DNH 4 average to below average ND games, they lose to the other on a missed extra point, finish as a runner up in the district with the 15-15 opponents record in district, can they be left out?

If we look at the records from last year, Class 1A had four teams at 9-0, five teams at 8-1, and 7 teams at 7-2. Since your own winning percentage is by far the biggest component of RPI (yes, your opponents' winning percentage also counts for the same 37.5%, but with that spread among nine opponents your own record is still the biggest factor), I think if this year's rules were in effect last year, those would be your 16 playoff qualifiers. I haven't done the math at all, but it seems to me you'd have to have an absolutely ridiculous difference in strength of schedule for your RPI to lift a 7-2 team over an 8-1, or a 6-3 over a 7-2. An 8-1 record is a winning percentage of .889, while 7-2 is .778 and 6-3 is .667. That number is almost 40% of your RPI, so right away that's a significant difference to surmount. It appears to me RPI is only really going to sort out things between identical records.

If you finish 9-0, you're in, because that means you won your district. If you finish 8-1, odds are you'll be in - if you didn't win your district, there probably aren't enough other 8-1 finishers to lock you out of the 16 spots. It's the 7-2 finishers where we're probably going to get into RPI, and again, it's the non-district opponents (both yours and those of the rest of your district) that's going to be the difference. Just for grins, let's say Regina wins the district (even at 7-2, perhaps with losses to Solon and Xavier, but that won't matter because they were 5-0 in-district). And let's say DNH loses a close one to Regina for their only district loss, but drops another game along the way to also be 7-2. That puts DNH in RPI competition with all the other 1A non-district-winning schools at 7-2.

Let's use Iabeastmode's thoughts of the rest of the district finishing 5-4, 2-7, 1-8, and 0-9. If that is indeed the case, that means those four teams went 2-14 in non-district games, which really stinks for DNH's RPI. If, as you say, DNH's own non-district opponents aren't that great, their chances of beating out another 7-2 team for a playoff spot would be pretty low. Their RPI would also give them a worse seeding for the playoffs if they did make it, so tougher opponents, more road trips, etc.

Again, if either Regina or DNH finishes second in the district at 8-1 with the only loss to the other, well ... it's going to be pretty tough for an 8-1 team to miss the playoffs. It's possible, of course (multiple ties for district titles would cut down on the wild card possibilities), but unlikely. If you win your own games, the importance of your non-district schedule and your district's non-district schedule lessens.

That's what it looks like to me, anyway.

 

[PNation]

I think that the scary part would be in 2 districts have 3 way ties for champion that is 6 teams with a guaranteed spot plus 7 other district champs, leaving on 3 at large spots. You could have some very deserving teams left at home, like a DNH or Regina.

 

[Deezy_70]

Have you ever thought about you could have a school win their district championship but still finish the season with a losing record?

 

[roosterk]

Think about 8-player 8 team districts, (7/9) 78% of OWP & OOWP are the same teams so its going to be 50% . OOWP is going to be extremely close to 50% because 78% of those ND opponents OWP & OOWP are district team mates. OWP if you played 2 ND teams that finished 11-7 combined would be 0.193.

For the 9 team district It will be (7/8) 88% of OWP & OOWP are the same teams if they don't get a 9th game. So 62.5% of your RPI is already determined (half of 88% of 37.5% OWP possible and 25% OOWP possible). OWP would be 0.196 if you play a 8-1 ND team, it would take a 7-2 ND team to get a .193 OWP to match the 11-7 combined of a 8 team district. Nice advantage 8 team district.

8-1 is better than 7-1, 7-2 is better 6-2 so another advantage for WP.

(I think my #'s are correct, correct me if I messed up)

 

[KidSilverhair]

My brain hurts, and I've done way too much math here for no real benefit, but I think I have found an odd consequence of this RPI formula. It does have to do with the 5-common-opponents in the district conundrum.

Anyway, with everything else being equal (your non-district record is .500, your district mates' non-district record is .500, your OOWP is .500), a team that goes 7-2 but 4-1 in their district actually would have a lower RPI that a team that was 7-2 but 3-2 in their district.

This has to do with the fact that your extra district loss raises the overall winning percentage for your district opponents. In other words, like I said in the OP, it appears your non-district games hold a little more weight than your in-district games.

For the 7-2/4-1 team I figure an RPI of .5972, while the 7-2/3-2 team is at .6021. Again, that's figuring a .500 record for your non-district opponents, a .500 non-district record for other teams in your district, and a .500 record for your opponents' opponents. It's possible - likely, even - that I didn't go deep enough into OOWP and how that would change between a 4-1 district record and 3-2 ... but I'm not interested enough to dig that deep. So take all of this with a grain of salt, and realize there's a huge number of variables going into RPI, like your opponents' opponents' percentage being determined by the results of 81 games, including games you were involved in because after all, you are one of your opponents' opponents ... It's kind of like "Inception."

 

[cidhawkeye]

My brain hurts, and I've done way too much math here for no real benefit, but I think I have found an odd consequence of this RPI formula. It does have to do with the 5-common-opponents in the district conundrum.

Anyway, with everything else being equal (your non-district record is .500, your district mates' non-district record is .500, your OOWP is .500), a team that goes 7-2 but 4-1 in their district actually would have a lower RPI that a team that was 7-2 but 3-2 in their district.

This has to do with the fact that your extra district loss raises the overall winning percentage for your district opponents. In other words, like I said in the OP, it appears your non-district games hold a little more weight than your in-district games.

For the 7-2/4-1 team I figure an RPI of .5972, while the 7-2/3-2 team is at .6021. Again, that's figuring a .500 record for your non-district opponents, a .500 non-district record for other teams in your district, and a .500 record for your opponents' opponents. It's possible - likely, even - that I didn't go deep enough into OOWP and how that would change between a 4-1 district record and 3-2 ... but I'm not interested enough to dig that deep. So take all of this with a grain of salt, and realize there's a huge number of variables going into RPI, like your opponents' opponents' percentage being determined by the results of 81 games, including games you were involved in because after all, you are one of your opponents' opponents ... It's kind of like "Inception."

I volunteer you to help the Boone nozzles with this process because they just don't seem to get it.

 

[KidSilverhair]

So take all of this with a grain of salt, and realize there's a huge number of variables going into RPI, like your opponents' opponents' percentage being determined by the results of 81 games, including games you were involved in because after all, you are one of your opponents' opponents ... It's kind of like "Inception."

I think I was wrong here. OOWP comes from the winning percentage of your 9 opponents' 9 opponents ... so while it's based on the winning percentage of 81 teams, that means it's actually ... 729 games?

(Obviously not 81 individual teams and 729 individual games, because there's a lot of overlap what with common opponents and all, not to mention the 9 games you played are included in there too, and now my eyes are glazing over and it's all turning black, so black ........)

 

[screwloose]

Your math is off here, Kid.

All the district record come out to 15-15, that's correct. But for the 4-1 district team(Team A) to end 7-2, they would be 3-1 in non-disrict. The 3-2 district team(Team B) to end 7-2, they would be 4-0 in non-district.

Each finishes 7-2, .7778.

Assuming everyone besides these two teams finishes 2-2(.500) in non-district, the extra win by Team B to get them even overall with Team A, would end equal with their district opponents at 23-22. If each of their non-district opponents also were exactly even, 18-18 overall, each team's OWP would be 41-40.

Each OWP 41-40, .5062

Assuming each OOWP opponent is .500, which isn't possible because of the odd number of games. But still, they would be essentially equal in your scenario.

Each OOWP 364-365, .4993

These two teams would be exactly equal, except for the head-to-head game, which Team A would have the advantage.

 

[PNation]

Jeff Linder tweeted out that he is working on who would have made the playoffs last year with the new system, and he mentioned that so far he has found out that a 5-4 4A team that played down twice (I assume Western Dubuque who played Wahlert and Beckman) would get in over a 7-2 with a lesser sos (Pleasant Valley) even though PV beat WD.

 

[atomic87]

Jeff Linder tweeted out that he is working on who would have made the playoffs last year with the new system, and he mentioned that so far he has found out that a 5-4 4A team that played down twice (I assume Western Dubuque who played Wahlert and Beckman) would get in over a 7-2 with a lesser sos (Pleasant Valley) even though PV beat WD.

This ought to be interesting ... I was wondering if BCMoore or some other computer whiz could run these scenarios.

 

[dtk913]

This ought to be interesting ... I was wondering if BCMoore or some other computer whiz could run these scenarios.

I would really like to see 2A. I am guessing it would have looked a lot different.

 

[KidSilverhair]

Your math is off here, Kid.

All the district record come out to 15-15, that's correct. But for the 4-1 district team(Team A) to end 7-2, they would be 3-1 in non-disrict. The 3-2 district team(Team B) to end 7-2, they would be 4-0 in non-district.

Each finishes 7-2, .7778.

Assuming everyone besides these two teams finishes 2-2(.500) in non-district, the extra win by Team B to get them even overall with Team A, would end equal with their district opponents at 23-22. If each of their non-district opponents also were exactly even, 18-18 overall, each team's OWP would be 41-40.

Each OWP 41-40, .5062

Assuming each OOWP opponent is .500, which isn't possible because of the odd number of games. But still, they would be essentially equal in your scenario.

Each OOWP 364-365, .4993

These two teams would be exactly equal, except for the head-to-head game, which Team A would have the advantage.

Dear God, I wasn't even thinking about these two 7-2 situations being in the same district. I was just pondering that a 7-2/4-1 team from D-X would have a worse RPI than a 7-2/3-2 team from D-Z, with all else being equal.

And to think I only started this to get some idea of what an RPI should be for the various 8-1/7-2 teams with all opponents being .500 ... I really fell down into quicksand there, didn't I?

 

[screwloose]

Here's some 4A for last year to see how the RPI would he looked.

I'm better than Jeff Linder.

 

[atomic87]

So it works just as good (or better) than points for 4A. Interested to see how 3A-A will work out. Hope somebody's on it.

 

[KidSilverhair]

I swear this is the last I'm gonna post on this for a while, but I thought maybe people would want to know what a baseline RPI would be.

This is for 6-team districts. For opponents' WP, I figured the 4 non-district opponents finished at .500 (18-18). The other teams in the district all went 2-2 in non-district games, and I figured the OOWP at .501 (365-364). So these RPIs are as close to baseline as you can get: if your non-district opponents have a better record, or your district members do better than 2-2 in non-district, your RPI would be higher. And the reverse is also true.

Therefore, with everything being as close to .500 as I can figure it, here's where RPI will be:

• For a team finishing 8-1, 4-1 in district: .6391.
• For a team finishing 7-2, 4-1 in district: .5975.
• For a team finishing 7-2, 3-2 in district: .6024.

This ought to provide a little perspective about what are good and bad RPI numbers, I hope.

 

[KidSilverhair]

Neat work, screwloose. That really shows how having a strong district like D-6 gives you a real RPI advantage. Also, look at Newton - they were 7-2/3-2, which should make for a decent RPI with all things being equal, but that OWP of only .333 killed their chances.

You kind of see what you need - win 6 or more games with the OWP about .500 or better to get a shot. Or win 5 games as long as that OWP is about .600, like Western Dubuque. If your OWP is down around .460-.450, 6 wins won't do it.

 

[KidSilverhair]

Looking at those again ... Newton was the first team out with a .5508 RPI. Everybody .5600 or better got in. Also, don't know how Linder might have been figuring things to say Pleasant Valley wouldn't have made it, because an RPI over .6000 is pretty much a lock.

Okay, now I really am done for a while. Damn you, screwloose, and your interesting spreadsheets!

 

[PNation]

I was wrong about Linder and Pleasant Valley, I assumed when he said a 7-2 team he meant one that made it last year, it was Newton.

 

[Reasoned]

So CR Wash 5-4 and Davenport Central at 6-3 would stayed home and Dubuque SR 6-3 & Western Dubuque 5-4 got in.

Wish I could have seen the spreadsheet screwloose posted because from prior posts and the above quote, the proposed RPI system seems to do as advertised. Having played two of the above four teams, it's entirely possible the other two would have fared better in the PO's...maybe. I still contend with 16 teams, the adopted PO format and the proposed RPI, the two best teams in the state will make the PO's...but not necessarily in the championship game.

 

[screwloose]

Sorry if anybody wasn unable to see the speadsheet temporarily. I accidentally trashed it. It should be restored now.

 

[PNation]

Also, did Assumption opt to play up in 3A? Their enrollment (314) puts them in 2A.

That was Assumption's number for this year, next year's number has not been released yet, and I know that assumption had a very small senior class (under 100) so they may be a 3A school by enrollment.

 

[Reasoned]

Sorry if anybody wasn unable to see the speadsheet temporarily. I accidentally trashed it. It should be restored now.

Great, thanks for posting that. So assuming your calculations are correct..., it appears to me utilizing the RPI is a superior way to get to the "best" 16 teams, as opposed to taking the top 2 finishers in each district (as was the case in 2017). I like what the state is trying to do, going to district play and now opening up the borders to really make things interesting for scheduling.

 

[Reasoned]

That was Assumption's number for this year, next year's number has not been released yet, and I know that assumption had a very small senior class (under 100) so they may be a 3A school by enrollment.

I would not be surprised to learn that Assumption requested to remain in 3A (even though their enrollment classification is 2A), due to competitive pride and scheduling.

 

[PNation]

I doubt that, they chose to play 2A baseball and basketball a few years ago when they were still in 4A football due to it still being conferences.

 

[PNation]

It's 9-11 from the previous year, so essentially 10-12 of the year that they are actually in affect.

 

[screwloose]

I don't know why there wasn't some way to incorporate class size into the formula. Even if it was one win/one loss swap per class up or down. The only problem with it is the lower class teams would be lobbying for more of those games and the higher class teams would want to avoid them more.

Bottom line is, winning is by far the greatest factor into RPI formula and it's not even close. It's the one factor each team has the most control over. Plus is also has the widest variance as far as the percentage(0.0000 to 1.0000) goes. And lastly, it accounts for three of the eight units(37.5%) factored in at the end. OWP also accounts for three units, but the variance is likely going to stay between 0.3333 to 0.6667, which keeps most everybody in the same area. The OOWP is even tighter and accounts for two units out of eight(25%), spanning 0.4500 to 0.5500. Of course, there can and, likely, will be outliers to these numbers. But you'll be hard pressed to see more than a couple per class.

 

[Reasoned]

I keep reading the state "sets" the schedule, I think it's more accurate to say the state "confirms" the schedule. If not mistaken, most schools AD's are currently hard at work lining up their 4 (in 4A) ND opponents. In the case of out of state opponents (new this year), obviously the request will only come from the IA school but there will have been a discussion with the out of state AD to make sure a particular date is open (which could prove a real challenge for the 2018 season since many out of state schools have already set their schedules).

 

[La_Fleur]

So if you were a team scheduling your non-district slate, would it be better to schedule teams you're almost certain you would beat and get the wins, or teams that should be good that you may or may not beat? I'm talking strictly in terms of RPI and making the playoffs and not something like playing challenging games because you think it'd help you in the long run

 

[smallcenter]

So if you were a team scheduling your non-district slate, would it be better to schedule teams you're almost certain you would beat and get the wins, or teams that should be good that you may or may not beat? I'm talking strictly in terms of RPI and making the playoffs and not something like playing challenging games because you think it'd help you in the long run

Both. Coaches are looking for teams that they can beat, that will go one to winner their district (ideally), or finish second. You want the wins for yourself, and you want them to win every other game they play. This is a BIG problem with not having a differentiation between classes. Example: Xavier will schedule Regina, who they will beat. Regina is going to go on to win the rest of their games (except for maybe a Solon match up). So Xavier goes and takes on a school two classes below them, but gets the same credit. Vice versa, if Regina knocks off a school two classes bigger, they get no credit beyond a win, even though they are at a huge disadvantage. (This is not an attack on Xavier or Regina, just the example that came to mind first).

 

[GOPANTHERS23]

I keep reading the state "sets" the schedule, I think it's more accurate to say the state "confirms" the schedule. If not mistaken, most schools AD's are currently hard at work lining up their 4 (in 4A) ND opponents. In the case of out of state opponents (new this year), obviously the request will only come from the IA school but there will have been a discussion with the out of state AD to make sure a particular date is open (which could prove a real challenge for the 2018 season since many out of state schools have already set their schedules).

The state does the schedule. The school just sends in a list of opponents they prefer to play. The schools have no say technically who they will get. The state does the final say on who will be your non district games