CalPreps Predictions points to CCS Playoff parity - Maybe?

[FBAddict]

This morning as I surveyed the major collapse in the WCAL playoff team success, I also noticed a number of upsets in other divisions in the CCS playoff field. This led me to explore all the sections and their upset factor, if you will, according to Calpreps predictions. What i can see suggests to a strong CCS parity and also a very weak NS section parity. The alternative option of thinking is that certain sections had strange anomalies from other sections.

The supposition I am using is that the more parity a section uses in its seeding teams and divisions for playoffs, the more likely one will see upset games. I am calling an upset where Calpreps predicted a win and the opposing team won instead. Here what I see so far, starting with the SJS section first since they started their playoff rounds last week.

SJS had 28 games in their first round with 4 games going opposite to CP prediction - 14% upset factor.
SJS had 25 games in their second round with 3 games going opposite to CP prediction - 12% upset factor.

NS had 12 games in their first round with 0 games going opposite to CP prediction - 0% upset factor.

NCS had 22 games in their first round with 3 games going opposite to CP prediction - 14% upset factor. One more game is scheduled for today.

CCS had 18 games in their first round with 9 games going opposite to CP prediction - 50% upset factor.

This likely suggests a very strong parity for the CCS seeding process and an extremely non-parity seeding for the NS section.

The NS section moved away from a Calpreps seeding model to a strict enrollment criteria for divisions and a win'loss quotient for seeding. This process created the least parity among teams.

The CCS section is the far end of the spectrum from the seeding process used in NS section.

The SJS and the NCS sections are somewhere in the middle leaning toward the very predictable NS section model. We know both of these sections use enrollment as a major criteria for division placement along with varying criteria for strength factors. So I wonder if we want very predictable playoffs or playoffs where the winner is in question.

Or maybe CCS is simply an anomaly this year?

 

[colhenrylives]

The CCS football placement formula is so complex math experts at the California Institute of Technology have yet to fully fathom its convoluted permutations. Just ask the PAL Booster, a CCS seeding guru who remains baffled by some of the 2023 bracketing.

 

[Cal 14]

The CCS football placement formula is so complex math experts at the California Institute of Technology have yet to fully fathom its convoluted permutations. Just ask the PAL Booster, a CCS seeding guru who remains baffled by some of the 2023 bracketing.

There was only one case where things didn't make sense, although many of us think we know how that was generated.

 

[Cal 14]

This morning as I surveyed the major collapse in the WCAL playoff team success, I also noticed a number of upsets in other divisions in the CCS playoff field. This led me to explore all the sections and their upset factor, if you will, according to Calpreps predictions. What i can see suggests to a strong CCS parity and also a very weak NS section parity. The alternative option of thinking is that certain sections had strange anomalies from other sections.

The supposition I am using is that the more parity a section uses in its seeding teams and divisions for playoffs, the more likely one will see upset games. I am calling an upset where Calpreps predicted a win and the opposing team won instead. Here what I see so far, starting with the SJS section first since they started their playoff rounds last week.

SJS had 28 games in their first round with 4 games going opposite to CP prediction - 14% upset factor.
SJS had 25 games in their second round with 3 games going opposite to CP prediction - 12% upset factor.

NS had 12 games in their first round with 0 games going opposite to CP prediction - 0% upset factor.

NCS had 22 games in their first round with 3 games going opposite to CP prediction - 14% upset factor. One more game is scheduled for today.

CCS had 18 games in their first round with 9 games going opposite to CP prediction - 50% upset factor.

This likely suggests a very strong parity for the CCS seeding process and an extremely non-parity seeding for the NS section.

The NS section moved away from a Calpreps seeding model to a strict enrollment criteria for divisions and a win'loss quotient for seeding. This process created the least parity among teams.

The CCS section is the far end of the spectrum from the seeding process used in NS section.

The SJS and the NCS sections are somewhere in the middle leaning toward the very predictable NS section model. We know both of these sections use enrollment as a major criteria for division placement along with varying criteria for strength factors. So I wonder if we want very predictable playoffs or playoffs where the winner is in question.

Or maybe CCS is simply an anomaly this year?

It's the design and intent.

SJS is not really interested in competitive equity, nor has the NCS been historically. I guess there have been discussions now that the NCS will move to a fluid divisional situation to move teams up and down depending on the amount of success that year, but we'll see if it really moves forward.

No, the CCS situation is not an anomaly. I've stated many times that the CCS and SS have similar issues, mainly the number of high-powered private schools. It is not coincidental that both sections have adopted strong competitive equity solutions.

The SS has adopted a model nearly exclusive to the use of Calpreps.com. I say nearly exclusive because some higher-rated teams still got bumped due to lower-level league automatic qualifiers. The results in that section are similar to that of the CCS where there have been a lot of higher seeds getting knocked out in the early rounds.

You and I both have statistics backgrounds, so we know that in a Gaussian distribution (bell curve for everyone else), the middle always has sporadic overlaps, which is what I would call a good number of the "upsets" on Calpreps. To me, any game with teams rated within 0-4 points are basically toss-ups. Before the SS playoffs started, there were several divisions that had ratings span not exceeding 5 points across all 16 teams.

In my opinion, the CCS should go ahead and move on to a similar model. The old playoff point system has always wanted to reward those who won leagues and played a tougher schedule. However, Calpreps does the latter part more honestly. Should playing Alvarez be considered the same as Valley Christian? Leland the same as Los Gatos? No. We're talking about 30-point ratings gaps.

If the section still wants to give A, B, and C leagues different number of automatic bids, that would be fine. But the playoff point system was needed when Calpreps didn't exist. Now that it does, it does a better job and clearly there will still be room for a number of the fun upsets that we enjoy. Additionally, there would be no more shenanigans possible.

 

[PAL Booster]

It's the design and intent.

SJS is not really interested in competitive equity, nor has the NCS been historically. I guess there have been discussions now that the NCS will move to a fluid divisional situation to move teams up and down depending on the amount of success that year, but we'll see if it really moves forward.

No, the CCS situation is not an anomaly. I've stated many times that the CCS and SS have similar issues, mainly the number of high-powered private schools. It is not coincidental that both sections have adopted strong competitive equity solutions.

The SS has adopted a model nearly exclusive to the use of Calpreps.com. I say nearly exclusive because some higher-rated teams still got bumped due to lower-level league automatic qualifiers. The results in that section are similar to that of the CCS where there have been a lot of higher seeds getting knocked out in the early rounds.

You and I both have statistics backgrounds, so we know that in a Gaussian distribution (bell curve for everyone else), the middle always has sporadic overlaps, which is what I would call a good number of the "upsets" on Calpreps. To me, any game with teams rated within 0-4 points are basically toss-ups. Before the SS playoffs started, there were several divisions that had ratings span not exceeding 5 points across all 16 teams.

In my opinion, the CCS should go ahead and move on to a similar model. The old playoff point system has always wanted to reward those who won leagues and played a tougher schedule. However, Calpreps does the latter part more honestly. Should playing Alvarez be considered the same as Valley Christian? Leland the same as Los Gatos? No. We're talking about 30-point ratings gaps.

If the section still wants to give A, B, and C leagues different number of automatic bids, that would be fine. But the playoff point system was needed when Calpreps didn't exist. Now that it does, it does a better job and clearly there will still be room for a number of the fun upsets that we enjoy. Additionally, there would be no more shenanigans possible.

Great post. I think when you re designing a play-off system or placement in leagues you have to ground it around a core principal. I think competitive equity is fairer, better for the sport and much better for the kids and coaches involved. Times have changed. School size has become largely irrelevant to determining the strength of high school football programs. Schools vary greatly in terms of the focus, resources and commitment they make to sports in general and football in particular. I think most leagues are doing a pretty good job in getting this right in trying to make competitive equity leagues before the season starts. It is easier for play-offs where you have the seasons whole body of work to order teams. Any other system would likely lead to more blow-outs and less competitive leagues and play-offs.

I agree with CAL14 that in moving exclusively to Calpreps it will simplify everything a great deal and make a fairer system. It is important to point out that in the CCS we are pretty close to this already. If that was the method used this year 39 of the 40 teams would have been the same as 35 teams are determined (appropriately in my opinion) as automatic qualifiers form league play. The only change would have been Aptos would have made it as an at-large team instead of Mountain View.

In terms of brackets, while there would have been some seeding changes, the open and DI fields would be the same. DII would have had Palma and Menlo-Atherton in DII instead of DIII with Carmel and Christopher in DIII instead of DII. The only other change I see is that Seaside would have been DV instead of DIV and Woodside would have been DIV instead of DV.

I think the CALpreps projected spreads are indicative of competitive equity working. All Dii -DV games are -projected to be within 7 points, which in my opinion means they have a good chance to be highly competitive games. The only two game with a higher projected spread are the DI semi-final where Los Gatos is projected to win by 10 points over SI and the Open final where Serra is projected to win by 30 over Wilcox. Serra would have a similar projection over anyone in the section this year.