What is weighting and when do you use it? I haven't heard from any Lindbergh fans about Monday's Greater Metro Boy's Basketball ratings where Eureka, who has lost three close games to Lindbergh, is rated slightly better than the Flyers. Why would I do that? Not because I am their unofficial statistician and have a bias towards the Wildcats. No, it has to do with season performance, as much as head-to-head competition. You can find all of my rating blogs HERE
This is where statistically modeling team ratings becomes somewhat judgemental and it isn't just Lindbergh and Eureka ratings. There are approximately 525 girl's teams and 540 boy's teams and there are instances similar to the Flyer-Wildcat scenario in every class of schools. If I just used raw ratings (the average of every game outcome) and did not put some weight on who defeated whom, the model would not be as good. So I have to take into account both seasonal performance (consistency against all opponents) and head-to-head results to build the best Braggin Rights Ratings.
I decided to compare Eureka, Lindbergh, Lafayette, and Marquette because they all play each other and the winner outcomes are quite muddled, including the three defeats Lindbergh handed Eureka. This deep analysis also helps me identify if I have over- or under-rated a team during the season. I have been tracking the top ten or twelve teams in the state to identify the "real" number one team in Missouri (read THIS) and I did the same analysis for the teams listed above.
Let me explain the layout. I entered the team schedule (date, opponent, score and record) in the first five columns. Column six is the spread (difference in score), column seven is the game rating. The game rating is the sum of the opponent's rating and the spread. The tighter the game average rating shows how consistent a team has been all season. Finally, the last column is the opponent rating which averages out for a Strength of Schedule rating. At the top, I list the raw weighting which is just the average of game ratings, and the Gramps Braggin Rights rating which I have posted that also has some weighting built into it as I build the rating model.
Clear as mud, right? As Mark Twain said, "There are lies, damn lies and statistics" and I try to stick to just the last one but it isn't always easy because you can manipulate statistics in many ways. Just ask any politician or corporate CEO.
I will start with Marquette, the lowest-rated of the four. Marquette started the season with high hopes, riding the talents of the talented, lithe 6'8" Power forward Andrew Young. Unfortunately, after teams started scouting the high-scoring forward, they all found defensive strategies to defend Young and let the rest of the team pick up the scoring. It has not been pretty for Marquette fans. The rest of the team hasn't done enough to take pressure off Young and he has been hounded by double and triple teams. Tough on any star player and Young hasn't bulked up enough to power through those double and triple teams. Look for Marquette to finish strong and get back to a 0.500 season. By the way, the beat Lindbergh by sixteen but are rated six points lower. Why? They have had some bad losses and it has drug their rating down.
Lafayette, under new head coach, Bryan Keim, has got the Lancers moving in the right direction after last season's debacle with parent and player revolts. Lafayette has been a very consistent team and had a resounding victory over Eureka in December in the HWY 109 championship and split two games with Lindbergh, beating the Flyers by 16 in the return match on the Lancer's home court. Recent defeats to Parkway West and Hazelwood Central show that they have a ways to go to be top-shelf but Keim is building a good culture in a program that hasn't always had that. Lafayette should go 4-2 and a respectable 16-10 for the season going into Districts. Because of their consistency, better record and recent dominant win, they are higher rated than Lindbergh.
Lindbergh has had an up and down season, with an 11-8 record. They play a very patient offense that would not hold up well if a shot clock was introduced. Some of the offensive possessions are two minutes long and they more likely end up with points rather than a turnover. If you take the three Eureka wins out, they are a five hundred team with a raw rating of 121. Because of those victories, I have their weighted average at 125.9. They are one of the most consistent teams but they also have some unexpected losses to Parkway North, Marquette and Lafayette that drug down their raw rating. The three Eureka games were all close with the last two decided by a couple of points. I have to give them credit for those three victories but the rest of their season keeps them below Eureka.
Sorry Flyers.
Eureka started the season gang-busters out at the Gene Steigenhorst Hillsboro tournament and was undefeated before getting beaten badly by Lafayette. Maybe they were reading my blog and didn't show up mentally. LOL. When they come to play, and they play with that intensity throughout, they are tough to beat and they defend as well as any team in the Metro area. The losses to Lindbergh were all nail-biters, decided on last shots or in the final minutes. They could have won at least two of those games. Losses to Lafayette and Jennings will turn a coach's hair prematurely grey. The loss to Parkway West might have been expected but not by 15. They have the opportunity to finish 18-8 with only Webster Groves in the way of running the table. If they show up and grind from the tip-off they might even compete with the Statesmen. Their raw rating is 127 and I have them weighted at 126.8, slightly above both Lafayette and Lindbergh, teams that gave them four losses. They could easily be 17-4 right now. They have been really good and really bad, thus their consistency rating is worse than the others. Let's hope they decide to be really intense the rest of the way.
I don't know if that explains my ratings any better but you can see that rating 1000+ boy's and girl's teams is not easy. I hear from a lot of people that point out they beat two or three teams rated above them and, if they are polite, and they provide good evidence, I can resolve the model issue with better weighting. Sometimes, it is like Lindbergh... they beat a very good three times but they don't show the same consistency in other games and are just a five hundred team with three upsets. It's tough for coaches of 16 and 17 year-olds, being able to get the best out of each player every game. I don't envy what they try to do and, and besides, I already have enough grey hairs.
Thanks to those of you who wished me well on my pacemaker upgrade. I had it inserted Monday and came home in time to score a JV game. Feeling great. Enjoy the ratings. I will post boys and girls class ratings soon.
Comments
Post a Comment