THE BEST PREDICTIVE TOOL FOR 2023-2024 - GRAMP'S RATINGS

I have taken readers through several predictive formulae that leave something to be desired.  The best predictive tool is the Pythagorean Win Percentage which is calculated by taking the points scored minus the points allowed to some magical power.  Unfortunately, you have to play the games to calculate PWP and the season is over.  You know how successful your season was already.

I said in my last blog that I had a better tool and you can do this yourself.  Take your team's schedule for this coming 2023-2024 season and look at Gramp's ratings for this past season.  For each opponent, write down their rating at the end of last year.  Then subtract their rating from your rating.  If the number is positive, give yourself a win.  If the result is negative, chalk up a loss.  Once you have done this for your entire schedule, count up the W's and L's.  That will be close to your record this coming year.

Now Grama says, "Don't do that!"  "Kids can change and teams can grow.  The team's rating might be much higher this year because they worked hard over the summer".  Yep, Grama is always right.  This is just an approximation and players and coaches can make a difference from year to year.  And if you don't like the analysis, come up with some strategies to make a difference in all those spreads that are four or five-point negative.  With the right strategy and a little luck, you can change those L's into W's and change a losing season into a winning season.

Here's how it works:

I will use Lafayette County as an example.  Lafayette Country was fourth in the state class three state tournament.  Here's a summary of last year's season.  They had a 28-4 record with two of those losses in the state semi-final and third-place game.  I took their rating of 125 from 2021-2022 and compared them to their opponents rating from 2021-2022.  I looked at the actual spread for each game and the predicted spread based on those 2021-2022 ratings.  In many cases, the spreads are quite similar.  I added up the negative predicted spreads and it came to four losses and 28 wins.  Now the team did not lose to all of the teams rated above them or win all the games they were rated higher, but the record came out the same.... 28-4.

 I did the same analysis for 50 teams, including the final four teams from all six boys' classes and some middle-of-the-road teams with losing records.  All of those charts can be found HERE.

If you want to try this out for your team for 2023-2024, click this LI NKfor a spreadsheet form.

You can find last year's team ratings in alphabetical order HERE.

I then summarized all of these predictions in a table below:

For the most part, Gramp's ratings from the year before helped predict the 2023-2024 records accurately for all but a few teams.  Winnetonka, Marquette, Lee's Summit and a few other teams were outliers but many of the 50 teams analyzed were predictable.

When I plotted the actual 2022-2023 winning percentage vs. Predicted Winning %, the scatter plot shows an R-Square value of about 0.85 which suggests a very good fit for predicting the coming year with Gramps rating from last year.

The bottom line is that teams will have to prepare for and play the games this winter.  Grama is right.  You might go through this exercise and not like what you see, but you can change that with extra effort and a good game strategy.  Prove Gramps wrong again.  Grama does every day.

If you want to play with your own schedule, enjoy trying out Gramp's season predictions for your team.

Good luck with your season.