clock menu more-arrow no yes

Filed under:

Judging the Kings' Schedule Using Win Probability

New, comments

When the full 2010-11 NBA schedule was released earlier this week, it became clear the Kings would have a pretty accomodating slate early on but huge difficulty from mid-December on. (Sort of like 2009-10.)

But just how easy is the first part, and just how difficult does it get?

I used Bill James' log-5 prediction method, as adapted to NBA use by current Rockets analyst Ed Kupfer, to map out the weakest and toughest stretches of the Kings' '10-11 schedule. There are some inherent difficulties in doing this now, one of which being several teams changed dramatically this summer. For now, I've used last season's record as the variable for opponent strength; clearly, Cleveland won't be as good as they were last year, and Miami will be better. But season-to-season quality tends to hold pretty well in the NBA, so big picture this should look just about right. (Later, when Basketball Prospectus' new book comes out, I'll update using Kevin Pelton's team projections.)

As the key notes, the purple line represents the game-to-game win probabilities, based on opponent record and home court advantage. The red line is a moving 5-game average of the Kings' win probability. That helps us see trends in the schedule's difficulty. One more important note: I have assigned the Kings' ability (or winning percentage, in this case) as that of a 32-50 team. That's my official prediction as of August 13. Put it in the book, and ... well, you know.


After the jump, some interesting (to me) notes about what log-5 says right now.

* Based on my 32-50 prediction, log-5 has the Kings' best record at ... 8-7. At no point in log-5's projection do the Kings get more than one game over .500. Mind you, this system currently assumes Cleveland is a 60+-win team, which clearly isn't the case. If you adjust that downward below 42 wins, the Kings end up two games over .500 at 8-6.

* The last time the Kings are .500 is at 10-10, on Dec. 8.

* The log-5 system projects an 8-8 record in October and November, a 6-7 record in December, a 5-11 record in January, a 4-9 record in February, a 6-10 record in March and a 3-5 record in April. So the Kings would start 14-15 (.482) and then go 18-35 (.339) the rest of the way.

Now remember: this assumes a 32-50 record. You may think that's pessimistic. You may think that's optimistic. Just for kicks, if you think the Kings have the ability of a .500 team, the team peaks at 21-17 on Jan. 15 before coming back to Earth (where "back to Earth" is ".500 ball"). If you think the Kings have the ability of a 25-57 club (the team's record last season), then the team peaks at the gaudy record of 6-6. This tool isn't to be used to predict the Kings record -- that's an input, the prediction. This just shows you in action how the early schedule is quite easier than the January-February section.