## Different method, same win prediction for Bucks

I used a different method to project the number of wins the Milwaukee Bucks would produce next season, but came up with almost exactly the same projection:  a little over 46 wins.

The previous method I used was based on individual players, my projection for each player’s position and playing time, and each player’s Marginal Win Score average over the past three seasons.

This time I considered only the player’s personal Win Score average over the last three seasons, the average number of minutes he played the last three seasons, and the Bucks average Opponent Team Win Score under Scott Skiles.

Basically, what I’m saying is no one was positioned, and no one’s personal defense was considered.  Instead I just assumed Skiles would put out a team that played about the same defense he demanded last season.  Then I simply computed the players personal productivity, expressed as the player’s average Win Score points per minute, multiplied that by the player’s average playing time per season, added all of the player’s totals up, and then divided that number by the total roster’s average number of minutes per season and then multiplied that total number by 241.2 to get a projected Team Win Score.  It sounds complicated but its actually really simple.  In simple terms, I just assumed that the spillover in total minutes would get trimmed down evenly among the players in proportion to their usual season minute averages.  With that assumption in hand, all I did was translate the team’s per minute average into a per game average (per game with overtime minutes, which is usually around 241 minutes).

Doing all that, I came up with a “Team Win Score average” projected to be 42.80, which is just slightly above last season’s NBA Team Win Score average.

To project Opponent Win Score average, which is the flip side of Team Win Score and loosely represents defense, I did two things.  First, I looked at the Bucks Oppo Win Score averages for the two seasons under Scott Skiles.  In Skiles first season the number was 42.85.  In his second season it was 41.50, an improvement of a little over 5% from season to season.  If you assume the same improvement, next season’s average will be around 39.46.

But personnel makes a difference.  So I looked at the Oppo Win Score the Bucks posted after the All-Star break last season.  That was 39.23, not too different from the projected number.  So I used 39.23 to project wins.

42.80 minus 39.23 divided by ten produces a “representative” Marginal Win Score per 48 of +0.36, which translates into a winning percentage of .563%, which over 82 games is equal to 46.1 wins.

Last summer when I used the same projection method I came up with 38.9 wins, which was way off the team’s actual win total, but was almost exactly the win total they were headed for during most of the season prior to acquiring Salmons (on, Feb. 17, the day before the Bucks acquired Salmons, Coolstandings had the team projected to win 39.7 games) (Note:  the link does not go right to Feb. 17, but you can do it manually in the left hand column.)

So I’m solidifying my 46 win projection.  And I can say with almost exact certainty that if the Bucks are going to win more than that number of games it almost certainly will be because of better than expected defense.  The Bucks are so chock full of veterans with long track records of production its hard to imagine a Team Win Score average that would be substantially better than what I projected.  (In fact their are so many older players the reverse looms as a small possibility).

Can the team improve on the defensive end enough to produce wins above 46 games?  That is possible, although when I analyzed each of the players the Bucks have added to their roster I found that none stands out as a defender.

With all that said, veteran teams do tend to play better defense, and the Bucks have certainly gone “veteran”.  So they’ve got that going for them.  Plus, while Skiles has improved the Bucks defense dramatically, the team has yet to come near the Oppo Win Score averages his teams posted in Chicago and Phoenix.   So dramatic improvement is not out of the question.

### 8 Responses to “Different method, same win prediction for Bucks”

1. Jerbil Says:

You project the Bucks to win the same number of games this year as they won last year. If so then there will be a lot of bitterly disappointed fans.

You are guessing the Bucks will be better on defense (for the year). -So your calculations must have them worse on offense.

Given the contracts already signed, not a lot Hammond can do to improve the situation?

• tywill33 Says:

What a difference a year makes. Last season I, and one other, were the only ones who had them north of 30 wins. Now predicting 46 wins is disappointing.

I had a feeling that was coming.

But remember, things happen during the season. And, things could go +/- 3 to 4 games.

And there’s two other wild cards. One, if Andrew Bogut can finally play a “full” season, that would make a difference.

Two, if Brandon Jennings substantially improves his scoring efficiency, that woud also make a difference.

But history says those things can’t be taken as given.

• palamida Says:

and you can add one more: Trading Redd’s expiring contract at the deadline (hopefully) for a highly productive veteran!

• tywill33 Says:

It sickens me what a hole he puts in the team’s salary structure. 18 million dollars just pissed away, literally pissed away.

• Jerbil Says:

Here’s a point I was hinting at: You assume Milwaukee will be better on defense this coming season. + Hammond made moves to solidify and improve the offense. How can the Bucks not be expected to win more games this year???

There are several elements to a possible explanation (Stackhouse, Thomas, et c.), but isn’t this the main thing?: Salmons played miraculously. (I.e. he played way above his career level for his 30 regular season games with the Bucks.) And your methodology assumes he and everyone else will return to their averages on offense.

Note that Hammond’s thinking goes along the same line. -He only claimed his FA moves were designed to keep the Bucks from going backwards this season.

But anyone with a spark of optimism has got to be dreaming…

2. Daybreak Doppler: A Series Sweep & Training Camp Coming « PocketDoppler.com - A Wisconsin Sports Blog Says:

[…] Coutside Analyst goes into detail predicting wins for the Bucks next season. […]

3. 2010 Miami Heat Projection: 63.3 wins « Courtside Analyst Says:

[…] Courtside Analyst An analytical basketball blog written for basketball fans who are not statistical analysts « Different method, same win prediction for Bucks […]

4. Fear the Deer? and Forecasting the NBA | The Wages of Wins Journal Says:

[…] seem to agree with respect to the Bucks.  Ty Willihnganz projected the Bucks twice (see HERE and HERE) and each time concluded the Bucks can expect to win about 46 games in […]