Testing Marginal Win Score using Missed Player Seasons

One method I like to use to test whether Marginal Win Score calculations are somewhere within the realm of reality is to look for seasons of “substitution”, in other words seasons when the given player is no longer playing for the given team.   I then compare the MWS numbers from the previous season and see how they compare to the difference in pythagorean wins after the player departed or joined the team compared with the effected team’s pythagorean win total the season before.    Below are some test examples using the Win Resumes I produced in the prior three posts.  For reference on the player’s mentioned, please skim down through the past three posts.  The resumes are clearly marked and easy to find.

Kareem leaves the Bucks and joins the Lakers

The season before Kareem left the Bucks, he played poorly, mainly due to the aftereffects of an eye injury that gave birth to his famous goggles.  In that 1974-75 season, I credit Kareem with producing 5.4 “Wins above .500%” (meaning his wins minus his losses divided by two, hereafter referred to as “W5 wins”).  That season the Bucks finished with 42 phythagorean wins.  In the following season without Kareem, the Bucks phythagorean win total dropped to 37 wins, just as MWS would expect.

In the season before the Lakers acquired Kareem, the team produced 30 pythagorean wins.  In Kareem’s debut season with the Lakers I credit him with producing 12.1 W5 wins.  In that debut season, the Lakers pythagorean win total improved to 40 pythagorean wins, in the neighborhood of what MWS would expect.

Jordan plays Baseball

In the season before Michael Jordan joined the Chicago White Sox, I credit him with producing 9.8 W5 wins.  That season the Bulls won the NBA crown and produced 58 pythagorean wins.  The next season the Bulls produced 55 wins, but they were lucky because their pythagorean total was 50 wins, which was reflected the following season when the team won only 47.  That is precisely the amount MWS would have expected upon Jordan’s loss.

Bird misses the 1989 season with injury

Near the end of his career, Larry Bird missed about the entire 1988-89 season.  The season prior to that MWS credited Bird with producing 11.4 W5 wins, in a season in which the Celtics recorded 56 pythagorean wins.  The following season without Bird the Celtics produced 44 pythagorean wins.

Magic contracts HIV

Magic Johnson contracted HIV, as I recall, in the preseason prior to the 1992-93 season.  That season, without Magic, the Lakers pythagorean win total dropped by 18 games.  Now, MWS does not credit Magic with producing 18 W5s, it credits him with 12.1 W5s.  But, it credits him with 18.1 straight wins.   It could have been that the Lakers were simply unable to find adequate replacement for Magic on such short notice.

Barkley is traded from Philly

In Barkley’s last season with Philadelphia, 1991-92, MWS credits Barkley with 7.8 W5s and the Sixers produced 37 pyth wins.  The following season, the Sixers produced 26 pyth wins, what you would expect.  However, MWS did not work at all on the other end of the trade.  The Suns had 57 phythagorean wins in Barkley’s first season, but they had 56 phythagorean wins the season before.  This is odd. 

Garnett leaves Minnesota

2006-07 was Kevin Garnett’s last in Minnesota.  That season MWS credits him with producing 8.6 W5s.  The team recorded 30 pythagorean wins.  The following season the team recorded 22 pythagorean wins.

Conclusion

So MWS appears to reflect some form of reality in most situations.  Of course, there are other situations where it doesn’t at all.  For instance, I thought the Cleveland Cavaliers would drop a number of games, but I did not see them completely cratering as they have.  But I think there are a series of explanations for that, but I won’t get into them here. 

Advertisements

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


%d bloggers like this: