28 Upsets last night? Or 12? (FHSAA vs. MaxPreps)

[Wellington]

Of the 100 or so games played last night it would appear that there were 28 upsets...based on FHSAA seeding (a lower seed beating a higher seed). But if we use a more reliable measurement...MaxPreps' rating...we see that there were only about 12 upsets. I'm new to Florida, so maybe someone can justify the use of FHSAA data? Including RPI? Dunbar losing decisively looks especially bad for the FHSAA...who have them ranked #5 in the entire state (ahead of Aquinas, Jesuit, Venice, Gibbons, Central, etc.) One more fun fact before I go: Northwestern is ranked #85 in the state (using RPI) but they are ranked #78 in the entire country by MaxPreps. If you can make sense of that then you are smarter than me. Lol. 

[THIS_IS_DILLARD]

Funny thing maxpreps have Dillard ranked #75 in the entire nation 

 

but not even top 30 in Florida smh , behind teams whose not even playing right now smh 

[Ray Icaza]

11 hours ago, Wellington said:

Of the 100 or so games played last night it would appear that there were 28 upsets...based on FHSAA seeding (a lower seed beating a higher seed). But if we use a more reliable measurement...MaxPreps' rating...we see that there were only about 12 upsets. I'm new to Florida, so maybe someone can justify the use of FHSAA data? Including RPI? Dunbar losing decisively looks especially bad for the FHSAA...who have them ranked #5 in the entire state (ahead of Aquinas, Jesuit, Venice, Gibbons, Central, etc.) One more fun fact before I go: Northwestern is ranked #85 in the state (using RPI) but they are ranked #78 in the entire country by MaxPreps. If you can make sense of that then you are smarter than me. Lol. 

Thanks for your work in compiling this data.  If it is accurate it certainly does throw some shade on this RPI formula as I predicted a couple of days ago regarding won-loss record during initial playoff round.  Good pollsters like Joe Pinkos and others generally are better than 80% accuracy (usually mid 80's) in their predictions week in and week out; these numbers are like 70% which aren't terrible so need obvious tweaking.  Better yet, apply them to a smaller sample of wildcard teams; ie...  2 per region rather than 4.

[JDHoss]

Like other states (such as VA and TN, where I'm from) I'd guess Florida rankings are made by people who have no way of actually watching everyone play, and they have to rely heavily on score comparison and strength of schedule. 

[Ray Icaza]

1 hour ago, JDHoss said:

Like other states (such as VA and TN, where I'm from) I'd guess Florida rankings are made by people who have no way of actually watching everyone play, and they have to rely heavily on score comparison and strength of schedule. 

Others have pointed out that score comparison might be a way to tweak the RPI formula.  Currently it weighs equally for you or against you when you lose to a team by 1 or 50;  a win is a win and a loss is a loss with no weight given to that stat.

[Perspective]

Also, you have to remember that, within each region, district champs are seeded 1 through 4 based on their respective RPI's and then the four wildcards are seeded 5 through 8 based on their respective RPI's.   Often times (more often than not?), the 5 seed (whose only loss may have been to a really strong team that won their district) has a higher RPI than the number 4 seed who happened to win a relatively weak district.   That's why I said last week, somewhat tongue in cheek, that unless I really knew the teams, I'd pick the 5 seed over the 4 seed in most games.  It's kind of like picking 9 seeds over 8 seeds or even 11 seeds over 6 seeds in the NCAA basketball tournament.   From my standpoint, a 5 seed (especially a 5 seed with a higher RPI) beating a 4 seed really doesn't constitute an upset. 

 

[DarterBlue2]

One easy fix is to give the district champs an automatic playoff berth, but not automatically seed them 1-4. Instead, they should be seeded based on their RPI at season end. So, the four champs could be anywhere from 1-8 based on their results. 

[Perspective]

2 hours ago, DarterBlue2 said:

One easy fix is to give the district champs an automatic playoff berth, but not automatically seed them 1-4. Instead, they should be seeded based on their RPI at season end. So, the four champs could be anywhere from 1-8 based on their results. 

Not a bad idea, Darter.   But until the FHSAA comes up with a more scientific formula for determining the RPI of all the teams, I'm OK with the making the four district champs in each region the top 4 seeds (in whatever order makes the most sense).   A district champion shouldn't have to play their playoff opener on the road, especially to a team from their own district that they've already beaten once - just because some mathematical formula says that other team has a higher number. 

Best recent example I can give you to support that argument would be the Lakeland/Lake Gibson scenario.   Lake Gibson beat Lakeland in the regular season to win their district.  That earned them a 4-seed, where they got the privilege of playing the 'best' wildcard team, i.e., the team that had the highest RPI among the wildcard teams, i.e., the team that had the highest RPI in the region, i.e., Lakeland.   Lake Gibson got to host that game and ended up beating Lakeland for the second time to knock the Dreadnaughts out of the playoffs.  

Winning a district championship ought to give a team at least one home playoff game.  

Just my two cents. 

[DarterBlue2]

15 minutes ago, Perspective said:

Not a bad idea, Darter.   But until the FHSAA comes up with a more scientific formula for determining the RPI of all the teams, I'm OK with the making the four district champs in each region the top 4 seeds (in whatever order makes the most sense).   A district champion shouldn't have to play their playoff opener on the road, especially to a team from their own district that they've already beaten once - just because some mathematical formula says that other team has a higher number. 

Best recent example I can give you to support that argument would be the Lakeland/Lake Gibson scenario.   Lake Gibson beat Lakeland in the regular season to win their district.  That earned them a 4-seed, where they got the privilege of playing the 'best' wildcard team, i.e., the team that had the highest RPI among the wildcard teams, i.e., the team that had the highest RPI in the region, i.e., Lakeland.   Lake Gibson got to host that game and ended up beating Lakeland for the second time to knock the Dreadnaughts out of the playoffs.  

Winning a district championship ought to give a team at least one home playoff game.  

Just my two cents. 

I understand your perspective (no pun intended). And, if the districts within each region were relatively comparable, I would support the position that a district champ should host at least their first playoff game. But within regions, there are some very lopsided ones. An example which is the subject matter of a thread herein is 7A, Region 3.Another example would be 8A, Region 2, where Vero Beach traveled to Palm Beach County to beat district champs, Forest Hills, a team that would likely have struggled to beat Freedom or Ocoee here in Orlando. As such, I am not convinced that all district champs deserve to host their first game, but I do feel that they should be given a playoff berth as it is not their fault (usually) that they ended up in a very weak district.

Although I was a skeptic when it came to using an RPI to determine playoff spots, I have gradually come to the opinion that on balance it is better than what we had before. I do agree with you that the calculation of the RPI could be refined and improved. 

[Wellington]

 

Again, I do not claim to be an expert on FL rankings. But a quick look at RPI reveals a flaw in their algorithm. Too much value is placed on record. In 7A, undefeated Tampa Bay Tech is ranked #1...ahead of 4 nationally-ranked teams (Aquinas, Bucholz, Edgewater and Dillard). Unbeaten does not mean unbeatable. (Mater Dei in CA being a possible exception). In the case of Tampa Bay Tech...it’s more likely that they haven’t lost yet. Here are some other disparities:

6A: I’ve already mentioned undefeated Dunbar having the 5th highest RPI in the state...and then losing badly in Round 1. And there are 3 more undefeated teams -- Vanguard, Naples, Auburndale -- all with a lower strength-of-schedule and a higher RPI than nationally-ranked Jones and Northwestern.

5A: One-loss Merritt Island ranked #2...ahead of Miami Central and American Heritage

4A: undefeated Villages Charter ranked higher than Cardinal Gibbons and Bolles

3A: One-loss Florida Christian ranked higher than FSU High and Clearwater Central

2A: undefeated Avante Garde ranked higher than Trinity Christian and Champagnat Catholic

Personally, I don’t care about a team’s record. I care about the opposition they’ve faced. Having said that...if Naples beats Northwestern this weekend then I may have to revise my own algorithm. Lol.

[Ray Icaza]

11 hours ago, Wellington said:

 

Again, I do not claim to be an expert on FL rankings. But a quick look at RPI reveals a flaw in their algorithm. Too much value is placed on record. In 7A, undefeated Tampa Bay Tech is ranked #1...ahead of 4 nationally-ranked teams (Aquinas, Bucholz, Edgewater and Dillard). Unbeaten does not mean unbeatable. (Mater Dei in CA being a possible exception). In the case of Tampa Bay Tech...it’s more likely that they haven’t lost yet. Here are some other disparities:

6A: I’ve already mentioned undefeated Dunbar having the 5th highest RPI in the state...and then losing badly in Round 1. And there are 3 more undefeated teams -- Vanguard, Naples, Auburndale -- all with a lower strength-of-schedule and a higher RPI than nationally-ranked Jones and Northwestern.

5A: One-loss Merritt Island ranked #2...ahead of Miami Central and American Heritage

4A: undefeated Villages Charter ranked higher than Cardinal Gibbons and Bolles

3A: One-loss Florida Christian ranked higher than FSU High and Clearwater Central

2A: undefeated Avante Garde ranked higher than Trinity Christian and Champagnat Catholic

Personally, I don’t care about a team’s record. I care about the opposition they’ve faced. Having said that...if Naples beats Northwestern this weekend then I may have to revise my own algorithm. Lol.

These are some very good examples as I think most would agree with your conclusion that the lower RPI teams on "THIS" list would defeat the higher "RANKED".  Overall, I do agree with Darter that it has played out better than expected but still has room for improvement.  I've been following an undefeated Auburndale this season and also agree with you they will have their hands full with a tough Jones team from our area; they haven't played anyone of that caliber.