The Pi-Rate Ratings

June 20, 2017

The PiRate Ratings Are Hard At Work

Welcome back gridiron fans.  The PiRates are back at sea working hard on getting their treasure for the 2017-18 season.

There will be a few tweaks to the formulas this season as we refine our ratings to make them as accurate as possible, and we are excited about it.

Tweak #1: Our first tweak involves increasing the alteration of each college football teams’ rate of adjustment as it applies to depth concerns.  If a team is stacked two-deep with talent, they should be able to endure a long, hard season.  But, if the team is only good in their starting lineup, and they lack the depth, they will weaken as the season goes on.  Consider two teams, State and Tech.  State and Tech are about equal in talent in their starting lineups, and on August 25, they are exactly the same in power rating.  For the sake of argument, let’s say they both look like 9-3 teams in a power 5 conference and have identical ratings of 118.0.

In the month of September, State and Tech both go 4-0 over similar competition, and their statistics show that they are still basically the same in talent and results.  Both teams ratings go up, but State’s go up a tad more than Tech’s, due to the depth issue.  Through four games, it won’t be all that much, and now on the First Monday in October, State is now 122.5, and Tech is 121.8.

In October, State and Tech both go 3-1 to sit at 7-1 with four games to go.  Both teams have lost some key players by now, but State has plugged in able second-teamers to fill the spots, while Tech has been forced to go with talent not up to Power 5 conference standards.  State’s wins have been a little easier, and their loss was much closer than Tech’s wins and loss.  Now, State has a rating of 123.1, while Tech has fallen a bit to 119.5.

In November, State finishes strong, going 3-1 to finish the regular season at 10-2.  Tech falters going 1-3 to finish 8-4.

We see this every year.  We believe we can factor in this depth issue into our ratings before the season begins, and we can adjust the rate of adjustment during the season as a team sees key players leave the lineup.  It also works in reverse; a team may have a blessing of several new underclassmen beginning to contribute.

Look at Georgia in 1980.  Hershel Walker was a true freshman.  In the second half of the season opener against Tennessee, Coach Vince Dooley inserted Walker into the game with the Bulldogs trailing by double-digits, and Walker turned the game around in his first 10 carries.  Georgia won the game, and Walker led the Bulldogs to an undefeated National Championship.  Had this new twist of our ratings been in effect then, The Bulldogs might have improved by an unprecedented 10 points between game one and game two.  As it was then, it took several weeks until Georgia’s rating really reflected their power, all because of one freshman.

This can just as easily happen to an average team that loses a couple of key players.  There have been instances when a team has lost a dozen key players by the end of October.  If it is Alabama, the Crimson Tide might be five points weaker than they would have been with the dozen players still healthy.  But, if it is Iowa State, the Cyclones might be 15 points weaker with the loss of these dozen players.

Going back to the unexpected bonanza, what if a team has five or six redshirt freshmen that have displaced upperclassmen in the starting lineup?  Obviously, this team must now be better talent-wise if starters have been replaced  (unless the coach has thrown in the towel on the season and is looking to the future).  What started as a so-so season may continue as a better season because the surprising advancement of the freshmen has made this team a touchdown better than it was earlier in the season.

The PiRate Ratings will adjust for this during the season by increasing or decreasing the adjustment rate of the teams after each Saturday.  Because the PiRate Ratings always show a total average of 100.0, that means teams might lose ground in the ratings after a good game, because other teams now earn more bonus for their play on the field.  It will be a work in progress, but in the end, we believe it will lead to more accurate ratings.

 

Tweak #2: Our basketball ratings have always been Four Factors-based.  We have three different algorithms using the Four Factors Data.  Football also has its factors.  In fact, in football, there are five factors.  They are, in order of importance: Explosiveness, Efficiency, Field Position, Finishing Drives, and Turnovers.   You can find excellent content online about these factors–some explain in detail like a thesis, while others give you a quick outline.

For our purposes, we have changed how we use the box scores of the games in our updating formula.  We will look more at Explosiveness and Efficiency when we update the teams’ ratings every week.  Additionally, we will keep an unpublished running score on how consistent teams are in being able to move the ball and stop the movement of the other team.  Which leads us to our big breakthrough for 2017-18.

Big Breakthrough

Have you ever noticed that over the course of an era, a team may celebrate a national championship or conference championship when they appear to not be as good as prior teams or soon see future teams that look better but do not win the championship?  You have probably seen that a really good team goes 10-2 or 11-1 and demolishes most of their opponents but suffered a terrible upset.  Then, that team runs the table but wins most of their games by nice amounts but by no means blowouts.

Think of a team that wins 55-17 and 38-10 or something similar for most of the season and then falls 31-27 to a mediocre team.  This team leads the nation in total offense or scoring defense, but they always fall a game short of the accomplishment.  But, then along comes a season where this school wins 31-20, 27-14 or something similar and runs the table.  They finish well down in total offense and near the top in scoring defense, but they go undefeated.

We have seen this happen multiple times in the last 30 years.  The dominating team is not as consistent as the team that went undefeated, and in three out of 10 games, the less consistent power may be two touchdowns better than the undefeated champion.  However, 70% of the time, the undefeated champion will be better than the dominating team.

The PiRates have tried to assign value or lack of value to the consistency of a team, but this is something that takes a good sampling of games.  This adjustment will be used in November after all teams have played 2/3 of their regular season schedule.  By then, we should have a grasp of what teams are staying consistent, and what teams are all over the map.  We will not adjust their ratings, because you never know if your inconsistent team will deviate 14 points above their rating or 14 points below.  What we will do is search for consistency when we select our Money Line plays.  We might be crazy, because we are coming off two consecutive profitable Money Line seasons (just for fun and not real wagering), so we might be cutting off our noses to spite our faces.

Enjoy your summer.  Football season will be here before you know it.  The PiRates are actually ahead of schedule this year with their college football updates, and we should have enough time to expand our preseason coverage this year.

Advertisements

Leave a Comment »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Create a free website or blog at WordPress.com.

%d bloggers like this: