[AtlPirateFan]

I am a stats nerd. I'm the type who keeps full spreadsheets to record the entire history of a Madden franchise because I don't think the game keeps enough records. So being the data nut that I am I wanted to look at the numbers and see how the top 16 in 3A shape up by the numbers. These are just for informational purposes and are no way any prediction of end of season standings. So here goes

Top 3 Scoring teams by total points

1. Belfry - 535
   
2. Boyle Co - 511
   
3. Casey Co - 506
   

Top 3 Offenses by Total Yards Gained
[LIST=2]
[*]Boyle - 4969
[*]Casey - 4940
[*]Belfry - 4693
[/LIST]

Top 3 Rushing Offenses
[LIST=3]
[*]Belfry - 4026
[*]Western Hills - 3827
[*]Floyd Central - 3343
[/LIST]

Top 3 Passing Offenses
[LIST=4]
[*]Lex Cath - 2666
[*]Casey - 2386
[*]Caldwell - 2371[/LIST]

Top 3 Defenses by total points allowed
[LIST=5]
[*]Central - 114
[*]Etown - 114
[*]Corbin - 136
[/LIST]
Note: Central takes the top spot based on total yards allowed

Top 3 Rush Defenses
[LIST=6]
[*]Central - 521
[*]Corbin - 655
[*]Boyle - 773
[/LIST]

Top 3 Pass Defenses
[LIST=7]
[*]Belfry - 743
[*]LaRue - 756
[*]Waggener - 856
[/LIST]

Top 3 Average Points per Play
[LIST=8]
[*]Belfry - 1.37
[*]Etown - 1.20
[*]Boyle Co - 1.00
[/LIST]

Power Rankings All 16 teams
[LIST=9]
[*]Etown
[*]Boyle
[*]Belfry
[*]Corbin
[*]Casey
[*]Central
[*]Caldwell
[*]Western Hills
[*]Lex Cath
[*]Powell Co
[*]Floyd Central
[*]Waggener
[*]Bell Co
[*]Paducah Tilghman
[*]LaRue
[*]Garrard
[/LIST]
Note: Power Ranking uses the Yards Per Point calculation used by the NCAA and NFL and adds it to the Yards Per Play calculation used by a few college football statisticians and then uses the average of the two. Because SoS is also hard to determine each team was given a SoS modifier, for each team they beat that has a winning record they recieved 1 pt towards the SoS.


A few interesting notes:
- Belfry's offense is explosive. 535 points on 390 plays from scrimmage. What this stat doesn't tell is that most of the plays come in the 2nd half when then JV and Freshman are getting reps. One game they scored 53 points on 17 total plays. They average 12 yards per play, ahead of Boyle (9.75 ypp) and Western Hills (9.24 ypp)
- Centrals Defense is scary against the run. Of the 521 yards allowed 30% of them (164) came in their two losses this year.
- Surprisingly of the 16 Boyle is among the top 3 worst against the pass allowing 1673 Passing yards, second to only Casey Co who has allowed 1722
- Lex Cath can't stop the run allowing 3740 rushing yards this year in total their defense has allowed over 5000 yards of total offense.
- Boyle and Casey are your most balanced offenses while Caldwell is the most passing bias. Belfry leads the lopsided rushing teams but at least 7 others are run heavy.

Will post more once have analyzed more.

[64SUR]

Great looking stats. :rockon:

[jlow]

#16 Garrard beat #10Powell 38-0 to start the season. weird

[Tica]

are you using stats filed with the KHSAA by the schools?

[Glory_Days]

I love stats, but.........give me Tilghman & Garrard over Floyd Central, Powell or Bell Co all day.

[AtlPirateFan]

#16 Garrard beat #10Powell 38-0 to start the season. weird

I think Garrard would win that head to head 9 times out of 10. These stats are indicative of how I think they are ranked in any way. Garrard played a tougher schedule and had three winds over winning programs while Powell didn't beat a single winning program. Garrard falls on the power ranking simply because they were so inefficient moving the ball and scoring.

are you using stats filed with the KHSAA by the schools?

That was the base data. I cross referenced it against what other schools posted and then filled in gaps with news articles and recaps to complete the data set.

I love stats, but.........give me Tilghman & Garrard over Floyd Central, Powell or Bell Co all day.

I don't disagree at all. As I stated above the power ranking isn't meant to be a top 16 list but just how they stack up by the numbers this season. As like most seasons I think PT is better than their record, not a premiere team in 3A but better than their record suggests. Obviously SoS is hard to calculate in and while ETown by the numbers is the best team in the list I would rank them at probably 4th or 5th if I were doing a list by my gut as I think Belfry, Boyle and Corbin are better and Central may be even if their offense is lethargic.

[Hound05]

That’s some very good data sets to determine the PR. I’m with ya on ETown. After throwing in SOS into the fold I think they’re around 5ish. But that’s strictly going by metrics. They could win the whole damn thing, this is just a good measuring stick in order to compare top teams. But as a OP kinda said, some of the data is skewed, as always is, in HS football. A couple running clocks-while the JV are in- coupled with some garbage time touchdown by opponents could take a team from the top to 4th or 5th. Especially with 4 of the teams being as close as they are. Still very impressive, better than Cantrell’s crap. Good job and good luck y’all

[Tica]

The fact that Powell is rated where it is would suggest there is yet a bit of data refinement to be done. Perhaps less offensive weighting.

[Hound05]

This is objective measuring. Subjective ones are the AP poll. You can’t look at any team in here and say they should be moved up or down, that’s not the basis of his data.

[OutlawJoseyWales]

The fact that Powell is rated where it is would suggest there is yet a bit of data refinement to be done. Perhaps less offensive weighting.

I agree. Powell is awful. Sorry.

[Tica]

It is not objective measuring. It is measuring inclusive of some data and exclusive of other data and there also was an arbitrary assignment of some values. Plus you are assuming the public data is accurate, which we know at least some of it is not.

I applaud the effort. But power ratings do not necessarily translate into game day predictability.

[AtlPirateFan]

It is not objective measuring. It is measuring inclusive of some data and exclusive of other data and there also was an arbitrary assignment of some values. Plus you are assuming the public data is accurate, which we know at least some of it is not.

I applaud the effort. But power ratings do not necessarily translate into game day predictability.

Game day predictability was never the goal. The data accuracy is a legitimate concern, there were some places where reconciling data posted by one school wasn't exactly matched up to data posted by another school but for the most part it was within a percentage of error that was acceptable. PT was the biggest offender and their posted stats on KHSAA especially defensively were way off. It took time and had to be compiled game by game but the numbers should be close. Also I am not sure what data is being excluded. Defensive data is being included.

Also as Hound05 stated is that it's meant to be objective based on on field performance alone. 30 years from now when people look at Powell County's 8-4 record they are going to see Powell had a solid year and a District Title. No one will remember they won 8 games on a crappy schedule.

I adjusted the SoS weight a bit more using CalPreps SoS ratings. It made things better but gives Lex Cath too much boost in my opinion. If I weren't using numbers to build the list I would slip Lex Cath below ETwon and could be happy with the list as a Top 16 remaining. The Power Ranking was only included as a fun aside and is not attempt to say "this is the pecking order"

Current list:
Boyle
Lexington Catholic - Heavily benefits from SoS
Belfry
Central
Corbin
Waggener
Etown
Garrard Co
Caldwell
Paducah Tilghman
Western Hills
Bell
Casey
Floyd Central
LaRue
Powell

[Hound05]

It is not objective measuring. It is measuring inclusive of some data and exclusive of other data and there also was an arbitrary assignment of some values. Plus you are assuming the public data is accurate, which we know at least some of it is not.

I applaud the effort. But power ratings do not necessarily translate into game day predictability.

It’s NOT meant for game predictability. That’s the whole point of it. Yea the data used is a little questionable because “official” records in HS games aren’t good as gold. It’s a platform to compare and further analyze somewhat comparable teams performance on the field. Just like it is for college FB/BBall power rankings.

[Jackson Purchase]

I love it Altpiratefan. You must really like 3A football to take all that time developing your algorithm. Probably any of the top 5 could win it all. But when they take the field, get punched in the mouth. It'll be the team that gets back up.....and punches back even harder.

[Malcolm Ex]

Casey ahead of central? Nah, I am not seeing that. 3A bout to find out Nothing has changed at State Champ High.

[AtlPirateFan]

Casey ahead of central? Nah, I am not seeing that. 3A bout to find out Nothing has changed at State Champ High.

The original power ranking was based on a poorly crafted SoS. The new one listed two posts above is more accurate using a better SoS model. That being said as has been said in this thread many times the power ranking isn't a who can beat who list, it is an analysis based on widely used and accepted algorithms to determine on field tangible performance. It is not an indicator of talent level. The new list is pretty accurate and that is my fault for not using a better SoS calculation in the original list.