Tough Night for ScoreCzar in 1A

[FilmDoesn'tLie]

As much as I enjoy ScoreCzar's Northwest Hoops & Football computer poll, a couple predictions fell short tonight.

LC/SA -
Mt. Baker/PT +
Eat/UPrep -

Key issues where Freeman seems to be a bit better are the inclusion of non default values in Strength Of Schedule ratings for opponents outside WA/OR/ID and no lower end maximum. These would have probably bumped up LC's rating who had tourney losses to good CA teams. CPCs out of state tourney wins would probably have bumped up Eatonville's strength of opponent schedules as they have two CPC losses. I have not been able to key in on UPreps mysteriously high ranking, other than some of the teams they and SA play probably max out their low rating on the strength of schedule scale which inflated theirs and SA's ranking.

I have updated my week 10 poll thread with knocked out and regional locks, where UPrep became the highest ranked team sit out of regionals. Pumas are a good team that have to play their way through a competitive tri-district with little warm up other than Seattle Academy.

 

[Fan_of_Hoop]

Uprep also lost to Charles Wright -- a team that finished 6th in the Nisqually. Now I will give you that was early in the season, but has Uprep really played enough games that stretch them to have improved significantly over the course of the season? I am doubtful.

But that does bring me to my biggest complaint with scoreCzar. Being a numbers geek I really like their formula, but there is nothing that calculates where teams finish the season -- every game has the same value. So two teams that played the exact same teams, and each team finished 10-10 with exactly the same scores against those teams would have the same rating. But suppose one team won their first 10, while the second team lost their first ten, then they reversed for the last half of the season? A rational person would obviously pick the team who won the last ten to win a head to head matchup. They are the hot team (maybe coaching, or injuries, etc). Maybe if scoreCzar added some weight to the most recent games over the earlier games it would improve their system. Not trying to throw rocks, just sayin....

FOH

 

[Warrant3]

I enjoyed reading your breakdown of the different rankings, although I have to admit I don't completely understand the way ScoreCzar gets it's numbers even after the explanation. The ScoreCzar is a good resource, but certainly not the know all/end all when it comes to predictions. Speaking of UPrep, they had a rough night. They had a plethora of turnovers.

 

[FilmDoesn'tLie]

I started this post because I had included this computer ranking in my earlier ranking threads and felt I owed it to this community to point out where something I represented did not pan out as what may have been expected. My observations as I made that series of polls/ranking posts left me with some hunches of both the key issues I reference above and that the polls were also influenced by this computer ranking. Let's face it, the sportswriters polled do not always have a lot of 1A data to work from, and if you are like me, you observe teams that are putting quality coaching, skills and effort on the floor who are not perhaps getting the recognition compared to others. It is hard to cut through the noise and without a strong newspaper industry, regional local preps coverage at the 1A level is not what it once was.

I don't want this to turn into either:
1) something kind of like a "Why does the BCS calculate so unfairly?" thread. ScoreCzar is a computer poll which is free of some biases that go into polls, but has other potential deficiencies. ScoreCzar went into a lot of detail in his blog on some the how's and why's of his particular poll. He is very active on twitter and via email if you have trouble like me posting comments on his site.
2) Bashing UPrep while they are down. Have I thought all year long, they were overrated in his rankings? Yes. I think this inflated SA as a result. But having seen a lot of these Puma players come up, they have some great skills and will remain a strong 1A team that we will see back in the playoffs in the future.
Ok to keep complaining about the Seattle Times not including 1A in their rankings, but only if you do give them credit for stepping up their 1A coverage recently.

Here are some great ScoreCzar positives:
He had lot's of accuracy for his 1A football this year, including score predictions.
If you are not using his site to help you make your March Madness brackets, you are missing a great resource he has for college hoops.
He is super responsive to questions. I will contact him directly with my observations so he can tweak if appropriate.
Despite the fact he lives in District4, I have seen nothing, zip nada, zero to indicate any bias at all in his calculations.

We don't need to go deeply into the Freeman poll, whose failings mainly result from incomplete results data. Though they did have the Pumas at 23rd. Garbage in, garbage out. Still a fun way to compare top teams nationally and across divisions.

When it all comes down to it, at this time of year the rankings don't matter, just the results. The argument for the best team this year will come down to who gets to bring home the hardware.




ScoreCzar