They also produce some of the best football teams in California so winning that bracket gives you a great argument to be national champs if you get past DLS.
New CCS play-off Proposal for 2019
- Thread starter PALbooster
- Start date
They also produce some of the best football teams in California so winning that bracket gives you a great argument to be national champs if you get past DLS.
There is one problem I have with this proposal.
Calpreps has a policy that an undefeated team must remain rated higher than any team it defeated, regardless of what the rest of the season looked like. As such, the ratings for both Carmel and Wilcox were inflated by Salinas and Valley Christian, respectively. Carmel was never a 28 (rated) team and Wilcox was never a 46 team.
Once that team loses, all bets are off and they fall to wherever they would normally be. In these cases, Carmel fell to about 18 and Wilcox to 36... both of which are about right. The Chargers have an opportunity to work their way back up, it seemed pretty clear to me that by week 4 or so, Valley Christian has surpassed them, but the Calpreps ratings weren't allow to reflect that until Wilcox lost.
So, teams' ratings could get inflated if they get one big win early in the year and are able to generally coast for the remainder. That screws up the true ratings and therefore, rankings.
Calpreps has a policy that an undefeated team must remain rated higher than any team it defeated, regardless of what the rest of the season looked like. As such, the ratings for both Carmel and Wilcox were inflated by Salinas and Valley Christian, respectively. Carmel was never a 28 (rated) team and Wilcox was never a 46 team.
Once that team loses, all bets are off and they fall to wherever they would normally be. In these cases, Carmel fell to about 18 and Wilcox to 36... both of which are about right. The Chargers have an opportunity to work their way back up, it seemed pretty clear to me that by week 4 or so, Valley Christian has surpassed them, but the Calpreps ratings weren't allow to reflect that until Wilcox lost.
So, teams' ratings could get inflated if they get one big win early in the year and are able to generally coast for the remainder. That screws up the true ratings and therefore, rankings.
Heck with the computers, it can’t be that hard to seed best 8 and so on throughout the section with the new system. Palboosters Wife May divorce him with the time he has to put in.
The bigger issue is clumping an A team with a C team in D5 or a Carmel with WCAL teams. The equity seems to disappear.
The bigger issue is clumping an A team with a C team in D5 or a Carmel with WCAL teams. The equity seems to disappear.
I don’t think Carmel will get grouped in with the WCAL playing in the Gabilan. With games against the big 4 there will be much more data for the computers to properly rank them; so if they were still ranked high enough to get in to D1 at that point they probably are good enough that they need to be there.
The bigger problem is on the back end of this thing.. I don't really know what it even means to be "worthy of" playing for state title but take a look at the D4/5 brackets.. A few good ones but the majority I am not sure about.. Bellarmine??? There needs to be some type of safeguards put in place..
8) Half Moon Bay 30 vs 1) Bellarmine 40
5) Leland 33 vs 4) Christopher 35
6) Gonzales 33 vs 3) Carlmont 35.33
7) Aragon 30 vs 2) Milpitas 37.5
Division 5
8) Mt. Pleasant 16.5 vs 1) Willow Glen 27.5
5) Saratoga 22.83 vs 4) Del Mar 24.5
6) Santa Clara 20.83 vs 3) Piedmont Hills 25
7) Soledad 19 vs 2) Monte Vista Christian 26.5
8) Half Moon Bay 30 vs 1) Bellarmine 40
5) Leland 33 vs 4) Christopher 35
6) Gonzales 33 vs 3) Carlmont 35.33
7) Aragon 30 vs 2) Milpitas 37.5
Division 5
8) Mt. Pleasant 16.5 vs 1) Willow Glen 27.5
5) Saratoga 22.83 vs 4) Del Mar 24.5
6) Santa Clara 20.83 vs 3) Piedmont Hills 25
7) Soledad 19 vs 2) Monte Vista Christian 26.5
Can you post a link to where CalPreps has that policy? I dont recall reading it on CalPreps and have only heard that from you, so be good to understand if this is speculation or fact. Not doubting you, just trying to understand the source.
This season's Bellarmine team was likely its worst in 50+ years. From season to season, you will see a lot of variance, so looking at Bells historically would be a bit of a mismatch, but certainly not this season. If they improve, they would be in a much higher divisional bracket.
Unfortunately, there are only a handful of undefeated teams left, so I can't show you a specific remaining example. There is no link that spells this out, but I can point you to this case.
Using Carmel as an example, by the end of the semifinals their rating was about 28. Since we obviously can't go back in time to show this, you can see this post by PALbooster who states that they were rated 9th in the CCS as of October 26th (see the last paragraph of the opening post). If you look at the current CCS list on Calpreps, #9 would put them somewhere around a rating of 27-29.
https://norcalpreps.forums.rivals.c...align-a-leagues-change-play-off-format.20194/
However, if you look at the current "trend" for Carmel, it shows what the value was for each of their games. What game even suggests a rating above 25? That would be Salinas in week 0. So, if almost all of the rest of their games were somewhere in the 15-18 range, how did the rating get up to 28? Same reason... Salinas, whose rating was about 26 at the time (and now, as their final rating). Further, if you had looked at their trend at the time, it would have said "Sharp Downward". Well, how could it have been a sharp downward trend if their rating kept rising to 28? I'll let you answer that.
http://calpreps.com/cgi-bin/2018/trend.pl?school=Carmel_(CA)
In the Wilcox case, the gap between their true rating and their actual was so big, that they were kept at 0.2 higher than the highest team they had defeated, Valley Christian. 0.2 is the minimum that an undefeated team can be over someone they defeated.
If you're still not certain, but would like to further investigate, I suggest contacting Calpreps directly at football@calpreps.com. Other posters strongly familiar with Calpreps like @FBAddict can corroborate the above information, as well.
One other thing about using Calpreps is that now teams have an incentive to run up the score a bit.
You're up by 12 late in the 4th quarter and you have the ball? Don't take a knee at the goal line... punch it in to go up by 19. This year 15 was the minimum point value given for a win (i.e., even if you beat a 25-rated team only by 2, you would get a minimum game rating of 40). That's 4 extra points.
You're up late by 25 in the 4th with the ball? Again, punch it in to go get the maximum gate rating (this year, that was 30). That's 5 extra points.
4 and 5 points don't sound like much, but over the course of a season, they can really add up to a bigger overall rating and can make the difference between a home playoff game and one on the road, or even whether you make the playoffs at all.
You're up by 12 late in the 4th quarter and you have the ball? Don't take a knee at the goal line... punch it in to go up by 19. This year 15 was the minimum point value given for a win (i.e., even if you beat a 25-rated team only by 2, you would get a minimum game rating of 40). That's 4 extra points.
You're up late by 25 in the 4th with the ball? Again, punch it in to go get the maximum gate rating (this year, that was 30). That's 5 extra points.
4 and 5 points don't sound like much, but over the course of a season, they can really add up to a bigger overall rating and can make the difference between a home playoff game and one on the road, or even whether you make the playoffs at all.
So does Calpreps manually manipulate the rankings to place those undefeated teams higher? Or does their algorithm just lend itself to what you are talking about? I mean it does make sense in a way. Carmel did play Salinas and was clearly the better team that night. Then after that other then their win over Palma Salinas was just beating schools that were not as good as Carmel (including a loss to a school rated a decent amount lower then Carmel in Hollister) until they beat Aptos. I mean did Salinas beating up on Milpitas, Seaside, Gilroy, Christopher, and Alvarez really warrant them being ranked over a team that beat them by two touchdowns at the Pit?
It seems Ned Freeman removed the changes made to his system over the years. At least I can't find it now and I thought it was a part of his "How Ratings Work" page. From memory, Ned's first change was to weight playoff games more than regular season games. Ned's 2nd change was to historically weight sections and states to increase accuracy of inter-section or interstate games. The 3rd change was to lock an undefeated team's rating to always be above the teams it defeated until at last it lost a game. The fourth and last change that I'm aware of was to synchronize all rating of all years of data so that a user could know that a rating of 2018 could be cross comparable to a rating in 2015 (or any other year). This was actually when Ned introduced this last dynamic shift in his ratings that he explained all his changes.
I found the link http://www.calpreps.com/cgi-bin/2018/ratings_fixes.pl. But I cannot find the documentation to the point we are discussing.
Last edited:
Are we really going to do this again? Calpreps recognized what I recognized right away, but took 3 or 4 others to either recognize (or perhaps just admit at the end of the regular season). Salinas was better than Carmel by about week 4. Salinas beating Palma, who had just beaten Menlo-Atherton in conjunction with Carmel's extremely weak subsequent schedule would have been enough to drop the Padres below. Carmel struggling against Monte Vista Christian and Alisal should have been a bright red flag, but apparently some people on the peninsula are color-blind. Calpreps evaluates the entire season, which consists of 10+ games for most teams, not just one.
I can't quite tell if you guys are just trying to jab me or if you really are unable to conceptualize how an entire season works.
As soon as the Padres lost to TKA, they dropped 8 points below the Cowboys. That's about right. 26 for Salinas and 18 for Carmel is about right for each team if you were to look at their entire schedule.
The algorithm automatically keeps an undefeated team higher. Keep in mind that the system processes results for over 14000 schools. You'd never be able to do that manually.
I guess I should have used Wilcox as my example so you wouldn’t get emotional and sidetrack the conversation. Lesson learned.
That is my point.. "Worst Team in 50 years" should not have it followed by 'top seed heading into playoffs" Just something I think they need to consider. The old system made clear distinctions between A and non A leagues and this would obviously blur that line..
Like I stated early, wasnt doubting just hadnt seen it anywhere...
But with that being said, we have to understand that CalPreps is a reflection of how a team has performed to date, not how good/bad the teams are. For example, if Folsom lost all their star players due to freak practice injury, CalPreps wouldnt change their rating before the game, nor take it into account for the game. I guess what I'm saying is that I have no problem with an undefeated team being rated above any team they beat. Wilcox beating VC should have kept them rated above until they lost regardless of whether VC seemed to improve during the year and would likely beat them in a rematch. CalPreps is not a ranking system, it is a rating system of the team's yearly performance. Not sure how their project a matchup algorithm works but teams with lower ratings can project to win against a higher rated opponent. Look at 2017 DLS vs 2017 Folsom. DLS is predicted winner on neutral field, even thought Folsom rated higher.
No system is perfect, but CalPreps is much better than the CCS Point system they were using exclusively prior.
You are confusing worst team in 50 years for Bellarmine, not in the CCS. Even with their poor year, they were still in the 25th spot for CalPreps CCS. Considering that Div 4 would start with #25 best team in CCS, why shouldnt they be the #1 seed? If your point is that the top 5 teams of CCS should move on, than that would require a very different playoff system where you would spread the top 5 teams across the 5 Divisions. Unless the CIF changes their system to go back to Enrollment based playoffs/bowls, the CCS should follow with competitive game model. I personally would love to see a state playoff system for teams Under 1000 and Over 1000 players instead of the bowl games...
This is true, but if all of the remaining results for Folsom no longer lived up to their rating, it would drop unless one of the teams they beat continue to prop them up. The results on the field over the course of the year is what matters.
With regards to the 2017 example, the prediction algorithm is different than the rating. I know that doesn't make sense, but Calpreps uses two algorithms. My guess is that the prediction model utilizes margin of victory a great deal more, or at least has fewer limitations as for the rating.
My issue with the CCS using Calpreps in conjunction with their own system isn't severe. I do like the change overall and the example I gave would not impact a great number of teams. It's just that one of the first models of its applications by PALbooster showed Carmel as borderline D-I/II with the WCAL schools because of what I described. I don't think anyone wants to see that.
There are going to be winners and losers in every system, but as long as it benefits the great majority, I'm cool with it. I think this system will do that. As long as everyone knows the rules going into the season, whatever happens is pretty much up to them.
All 4 proposals as detailed above have made it through the AD Advisory Committee and League Commissioners committee. It requires two more reviews - the next by the Executive Committee in mid-January 2019 and the final by the Board of Managers in late January or possibly. if needed, April 2019.
I get the equity thing and everyone deserves a trophy these days idea. It just sounds wrong to send your #1, #9, #17, #25, #33 best teams on to a regional. I’m sure there will be another proposal next year.
There, you finally got to prove it. Salinas was better than Carmel. And yeah, I’m jabbing you Cal!
My point is if you're mediocre in league I don't like a system that rewards you by lessening the competition. Bellarmine (not to stir it up, just an example) couldn't hang this season so instead they go into a division void of any and all competition that they're accustomed to facing. I get that hey if they're better and they win but isn't that what it was like when 2 & 3 win teams were being crowned CCS champs. Every system has it's flaws for sure.. I just think they need to come up with a better plan with the top divisions (NCS has 4 team bracket at top, why not have 2 of those) and have some stopgaps in place so that some A league schools can't drop down and conquer.
Fair enough...
It's not merely beating teams that are better than the one that beat you, the rating is a compilation of all of the games played. Whether we're talking about Carmel or Wilcox, the same situation applies.
The teams that Carmel and Wilcox played after their early big win were not as good as those Salinas and Valley Christian played, respectively. While the Padres and Chargers both got some of those wins by blowout, not all of them were (e.g., MVC for Carmel and Los Gatos for Wilcox). In normal years, those games would have been a weight on the rating, but because of the policy at Calpreps, they weren't. Conversely, the wins over teams like Palma and Aptos for Salinas and Pleasant Valley and Serra for Valley Christian were big boosts.
Again, if you look at the "trend" pages for each team, you can see what value the computer applies to each game. Most of Wilcox's later games were in the teens and 20s, while Valley Christian had multiple in the 50+ range. That normally would have caused a switch if, for instance, the Chargers had just beaten a team "as good" as VC. But, because it was VC, the policy kept them higher. The trend pages also give an indication of how the movement should be going. In both Carmel's and Wilcox's case, it said "Sharply downward" despite the fact that both of their ratings were actually rising... rising because of Salinas and Valley Christian.
I view the policy more of a convention. Most published pollsters will follow suit with this type of logic, but I think there needs to be an expiration date. Week 0 is not even week 7, much less 14. So much happens in the meantime and some teams just develop better than others. Many times, it's not their fault (i.e., teams with more players going both ways probably won't develop like those with separate units due to depth). Other times, they can contribute to it (i.e., better non-league opponents challenge you more to improve).
The fallacy is looking at each week as the benchmark for the following, but not really responding to data that indicate a change. Cal-Hi did this in their prediction of the Lawndale-San Joaquin Memorial game, stating that since the Cardinals "showed they could beat Sierra Canyon", they were considered the favorite. However, that LHS-SCHS game took place at the beginning of the season. I believe I'd be hard-pressed to find anyone who said that they thought that the Lawndale that played last Saturday could beat the Sierra Canyon that played later in the evening.
