Wednesday, December 14, 2011

any number can play

(today's entry powered by a song from the only post-more betterness NUFAN album i can still stomach.)

2010-2011 Utah Jazz Projected Stats (rookies listed with projected 2011-2012 stats)

2011-2012 Utah Jazz Projected Stats (rookies listed with college stats)
2010-2011 Dallas Mavericks Stats (unedited)

2010-2011 Dallas Mavericks Stats (rookies listed with college stats)

This blog's namesake tweeted this morning with a question:

, i would like to know what Jazz fans are expecting this year in terms of wins. out of 66 games, how many wins?

Among the various answers were a few suggesting that the jazz simply don't have the roster to make much of an impact. I thought such negativity was, while probably grounded in an affinity for reality, a little big of a drag given all the hype coming out of training camp so far. So I thought I'd do a little number crunching. I should preface what you're about to read with a couple of things: (1) my big statistical assumption isn't very scientific in its premises, and (2) I don't have a great brain for statistics anyway.

My thought was to compare the production numbers (points, rebounds, assists, steals, blocks and minutes per game of each player, along with age and size) of the currently posted Jazz roster against the Dallas Mavericks championship-winning team from last year. I didn't include Elson or Fes, but I did include Kirilenko rather than speculating that Josh Howard joins the team.

What you see above are four matrices/images representing individual averages of each player of the team; two of the images are for the 2011-2012 Jazz roster as it stands right now, and two are for the 2010-2011 Mavs. In order to even attempt a realistic comparison, there is one key point you need to keep in mind. The Jazz, as they stand now, are made up of four rookies (if you exclude Trey Gilder's 7 games played last year, which I have). I couldn't very well include their college stats without making a crazy impact on the team's averages. So you see, for the Jazz, one spreadsheet labeled "WITH MODIFIED ROOKIE STATS" that represents what I anticipate each of the rookies to bring to the table this year in combination with the actual 2010-2011 numbers each player put up (except where otherwise noted); the other spreadsheet, "WITH ROOKIE COLLEGE STATS," shows the numbers each of the four rookies put up in their senior year of college along with the otherwise unedited '10-'11 stats (again, except where otherwise noted.

The two Dallas spreadsheets are my attempt at finding a reference for the Jazz's numbers. The first Dallas image, "ACTUAL STATS," shows the real numbers the Mavs players put up last year. The second, "WITH ROOKIE COLLEGE STATS," shows how last year's Mavs would have stacked up had Dallas' only rookie, Dominique Jones, put up his college numbers.

In short, the only spreadsheets that I'll actually be comparing are the Jazz "MODIFIED" matrix and the Dallas "ACTUAL" matrix. But how do I adjust for rookie performance, you ask? Good question.

For the sake of comparison, I’m going to start by using the same ratio of productivity change from Dominique Jones. Jones’ college averages are represented in the Dallas matrix labeled “WITH ROOKIE COLLEGE STATS,” and his actual numbers are in the Dallas matrix labeled “WITH ACTUAL STATS.” Jones, like all other rookies in these spreadsheets, has his numbers contained within a red box.

In order to explain the ratios I’m using for the Jazz rookies, let’s compare Jones’ college averages with his rookie year production straight up:

Jones scored 12.17% as many points per game with Dallas as he had in college, pulled in 25.93% as many boards, dished 31.43% as many assists, stole 20% of passes, blocked half the shots, and played 20.55% as many minutes in his rookie year with Big D as he had throughout college. For the sake of comparison, let’s run the same numbers on Derrick Favors:

I realize that I’d need a much bigger sample to compare these two sets of numbers. But simply to provide something of a control, in order to make a very rough, finger-in-the-wind estimate, I’ve averaged the two percentage drops in each category and applied them to this year’s Jazz rookies (excluding Kanter, as I’m just using his Euroleague numbers, and including Trey Gilder, who might as well be a rookie). So, the average drops are as follows:

So, if we extrapolate this (almost useless) set of percentages against, say, Alec Burks’ college numbers, we wind up with the numbers you see in the Jazz’s “WITH MODIFIED ROOKIE STATS” matrix:

So that's my very unprofessional attempt at accounting for rookies. Take it for what it's worth.

What does all of this mean? The most ego-inflating numbers to compare are the average total points per game scored by last year's champs (142.4) versus the projected Jazz total points per game (151.6). But that's not a very useful yardstick to judge the potential of this year's Jazz team.

The average Mavericks player last year scored 7.28 points in 19.04 minutes of play, 0.382 points per minute. At five players for 48 minutes, that puts the Mavs at 91.76 points per game last year. This number doesn't reflect the Mavs actual regular season average of 100.2 points per game because not all players included in my mock-up roster were on the Mavs team for the entire season (at least, that's my best educated guess on the discrepancy).

The average Jazzman, after adjusting for my guess as to rookie performance, put up 8.92 points in 21.99 minutes of play. That's .406 points per minute played, considerably higher than the respective Mavericks number. If each Utah player was playing with those kinds of numbers for a full 48, that's 97.44 points per game--almost six points per game higher than last year's league champs.

The picture gets more complex when looking at individual box scores.
  • REBOUNDS: The average Jazz player last season (after the rookie index) had 0.155 rebounds per playing minute, for a team total of about 37.12 per game. The average Mavericks player had 0.168 rebounds per playing minute, for a team total of about 40.38 per game.
  • ASSISTS: The average Jazz player last season (after the rookie index) had 0.091 assists per playing minute, for a team total of about 21.94 total per game. The average Mavericks player had 0.089 assists per playing minute, for a team total of about 21.35 per game.
  • STEALS: The average Jazz player last season (after the rookie index) had 0.031 steals per playing minute, for a team total of about 7.47 total per game. The average Mavericks player had 0.030 steals per playing minute, for a team total of about 7.21 per game.
  • BLOCKS: The average Jazz player last season (after the rookie index) had 0.022 blocks per playing minute, for a team total of about 5.29 total per game. The average Mavericks player had 0.018 blocks per playing minute, for a team total of about 4.41 per game.
The Jazz only fall behind the Mavericks in rebounds; all other categories favor the Jazz. And the assist margin is pretty low, considering these numbers don't include any numbers from Deron Williams.

My biggest take-away? Even if you can think of some reason why the Mavs numbers here are lower than they should be, you could modify the Jazz's numbers to reflect only the final few weeks of the season for Gordon and Favors to reflect their increased productivity late in the season. Or you could plug only the first-half-of-the-season numbers for last year's players; do you recall that at this time last year, the Jazz were third in the West? I suppose my conclusion is that the Jazz obviously don't have the kind of perennial all-star firepower you get with guys like Dirk and The Jet. But the numbers don't lie: this year's Jazz team has the potential to knock off last year's best squad.

No comments:

Post a Comment