Monday, February 9, 2009

GAMES PLAYED BREAKDOWN

Hand-in-hand with Andy's attention-grabbing little bit on the league message board about injuries. . .

Back again to drop a little stat knowledge fo' dat ass (or doze asses, if you will). . .
Let's look at total Games Played this season, meaning the number of games that you have had players CONTRIBUTE to your team. . . If they sat on the bench the whole game, or they missed due to injury or because of their grandma's funeral, a playa doesn't count. If they played but got hurt in the first quarter, or only chipped in 2 points and 1 turnover before being unceremoniously benched for the rest of the game, the chump counts.

Here are the GP counts, through 15 weeks (in descending order of team rank in our standings):

Y-G-A - 676
Balding - 710
Del - 687
Nutcru - 671
Mao - 622
28 lyo - 605
Block O - 618
Brew-Ha - 642
Liberal - 613
Phi - 572


If you'll notice, the GP counts go down fairly well in line with the actual team standings. This might be expected, as you would think that some of the things that would lower your GP counts (injuries, poor day-to-day roster management) would go hand-in-hand with your team's ability to win games. But while a good GM tends to maximize the number of games available within their team, it should *not* be assumed that just because you have a lower GP count, you are a worse GM. Some of these factors, such as the uneven distribution of games played for each month across the whole NBA schedule, might weigh poorly against your team TO DATE, but will swing the other way in the months ahead. Other factors, such as injuries, are a combination of random poor luck and imprudent risk-taking in selection of players (ahem, Annan). A couple of anomalies to the overall curve would be Y-G-A overproducing on a per-game basis, Brew-Ha underproducing, and a subtle shuffle of teams in the 28 lyo/Block O/Liberal mix. Otherwise, everyone falls right in line.

One thing to note is that, at the extremes, this is an ABSURD difference in GP count-- Balding Ballers has 710 games over 15 weeks (47.33 games/week), while Phi Slamma Jamma's Injured All-Stars have merely 572 games over the same timeframe (38.13 games/week), which represents a 9.2 games/week difference, or 24% MORE games for Balding Ballers. That's like trying to match in 66 games the production that a normal player would get in 82 games, and that's damn difficult to do.

Now look at the GP counts AGAINST each of our teams-- the findings here are a little more interesting to me from an analytical standpoint, because it is entirely less in our control as individual owners. What I mean by that is this number looks at the total GP counts by all the teams that were FACING our team in any given week, and as a result, it just comes down to scheduling fortune, injury whims, and sloppy GM-smanship from our OPPONENTS on any given week. One would think that, over the course of the season, these GP counts would end up fairly consistent, and it would all pretty much wash out. However, AS OF RIGHT NOW, that does not appear to be the case. . .

Again, in descending order of team rank in our standings:

Y-G-A - 657
Balding - 642
Del - 652
Nutcru - 638
Mao - 600
28 lyo - 617
Block O - 637
Brew-Ha - 665
Liberal - 638
Phi - 626

A quick look confirms that the numbers overall are more "cunched" than the normal GP counts. At the extremes, you have Mao's cupcake schedule of only 600 games, and Brew-Ha's tougher schedule of 665 games. This 65-game disparity is still pretty substantial, representing a 4.33 games/week difference in average opponent, or just over 10% (the average number of games played/week for all teams, for the season, is 42.77). So that goes a long way to explain how Brew-Ha could have such difficulty living up to his own GP numbers, and it sheds some new light on Mao's season performance to date-- is his team a legit playoff contender, or are they inordinately overperforming because of the wimpy competition they have "lucked into" facing? Another thing to note is that Y-G-A has the 2nd-most GP-Against, which may be surprising (and is further proof of my dominance). In fact, all of the top 4 teams in the league (Y-G-A, Balding, Del, and Nutcru) represent the top half of the GP-Against rankings, which may suggest that when the average team faces us, they MAKE SURE to trade out their rosters as much as possible, to maximize the number of games played and give themselves the best shot at pulling out a win.

Finally, if you combine BOTH lists, and you subtract the GP-Against from the GP count, you have the GP-Difference for all teams in the league. . .

Y-G-A - (+19)
Balding - (+68)
Del - (+35)
Nutcru - (+33)
Mao - (+22)
28 lyo - (-12)
Block O - (-19)
Brew-Ha - (-23)
Liberal - (-25)
Phi - (-54)

Surprising or not, you'll notice that this list goes almost EXACTLY in descending order according to our team's total rankings this year. The one exception, of course, is mine own Y-G-A, further demonstrating statistical dominance by coming out on the top of the ranks despite only the 5th-best GP-Difference.

Y'all betta watch out.

No comments: