What caught my eye is how poorly the Knicks fared in Krolik's high level analysis. You don't have to go much further back than the last week of games to know the Knicks haven't exactly been a model for any kind of offense lately, running or otherwise, but putting aside the gallows humor, some of the specifics of John's analysis weren't consistent with my observation of the Knicks all season. So I thought I would look at some numbers of my own to see if maybe I had been missing something.

Before I go on, I do need to provide some qualifiers. Knick fans and Cleveland fans, whether they like it or not, have been positioned as adversaries in too many online discussions because of the whole issue of Lebron's free agency, which conjures up a whole bucket of issues around the self-image of the cities (and their respective fans). I like John's blog precisely because he's so level-headed given the heated rhetoric surrounding all things Lebron, but such is the pervasive noxiousness of 2010 Lebron speculation that even he can't resist getting in a (slight) dig at New York based on the conclusions provided by his data. 

I'm not responding because I feel any need to support New York as a free agent destination or defend the suitability of the coach's system for certain superstars; I'm responding strictly because some of the issues John has raised are interesting to think about. I'll let others continue to speculate endlessly on all things Lebron. (For what it's worth, my personal position is that Lebron is a low probability opportunity for the Knicks, and to the degree I think about what to do with the cap space that opens up this summer, I mainly concern myself with scenarios for trades and the acquisition of second tier free agents who can be acquired for less than max money).

Now that we're done with that, let's look at what John did in his blog post. He decided to use the rankings of NBA teams for pace and fastbreak points to get a very basic snapshot of what he calls "running efficiency". Right off the bat, I saw a problem with this approach, mainly because the Knicks are dead last in fastbreak points. Given the Knicks are one of the fastest paced teams in the league (even with their decision to slow down the offense in December in the wake of a very slow start), such low production of fastbreak points generates the conclusion that the Knicks are getting terrible returns on their fast pace of play -- which is indeed what Cavs: the Blog's first table concludes (it's worthwhile to sneak a peek at the first table in the blog post if you haven't already done so).

My first issue with this mode of analysis is that fastbreak points aren't well defined (I've seen them defined as baskets scored within a range of 6 to 8 seconds with no clock stoppage, presumably not including scores after an offensive rebound), and there isn't a lot of information out there about the nature of fastbreak possessions, nor is there much historical tracking. The best discussion of fastbreak points I could find is a 2 year old article on 82games.com -- which found some interesting trends, but the primary predictive value for fastbreak points  seems to come from fastbreak *differential*, or net fastbreak points (the difference between your team's fastbreak points and the opponent's). 

The other issue I have with using fastbreak points as a loose barometer for efficiency is that a team like the Knicks may not score on many quick possessions that meet the definition of "fastbreak",  but they may *still* shoot early in the shot clock or soon after transition opportunities generated by rebounds and turnovers -- in other words, they still generate points commensurate with their pace, and counting only fastbreak points understates what they do. (Here's the current ranking of teams based on fastbreak points)

I decided to assemble John's table with a different statistic in place of fastbreak points -- not because it's a better alternative, but because it would provide another snapshot that might be a useful contrast. Specifically, 82games.com collects information on percentage of shot attempts taken by teams at certain points in the shot clock (0-10 seconds, 11-15 seconds, etc), as well as shooting percentage on those shot attempts, and points generated. (you can see an example here - look at the table titled "Shot Clock Usage")  I ranked the teams based on the points they generated on shot attempts within the first ten seconds of the shot clock. Here's a table that summarizes where the teams stand on points produced on "quick shots":

	Points on Shots <10secs
GSW	41.5
PHO	39.5
MEM	36.9
PHI	36.5
ATL	36.4
DEN	36.2
NYK	36.1
MIN	36
LAL	35.2
SAC	35
HOU	34.7
IND	34.5
NOH	34.2
ORL	33.8
OKC	33.6
MIL	33.6
CHI	33.4
WAS	31.9
LAC	31.8
UTH	31.5
BOS	31.3
TOR	31
DAL	30.9
SAS	29.9
DET	29.1
MIA	28.8
NJN	28.4
CLE	27.6
CHA	26.6
POR	24.3

Right away you can see the Knicks suddenly vault to the top ten in the league if you expand the definition of fast paced offense to encompass shots taken early in the shot clock. I understand this poses a different set of problems, since not everyone will consider a shot attempt at, say, the 9-10 second mark to be a "transition" or "quick" shot attempt. But shot attempts in the first 10 seconds seem to represent the teams fairly based on their pace of play in most cases, and definitely encompass fastbreak points. For example, the first incarnation of the iconic Seven Seconds or Less Phoenix Suns team coached by D'Antoni took a remarkable 48% of their shots within the first ten seconds, and Golden State Warriors teams the last few years have generally been in the 44-46% range. (Last year's Knicks were at 45%, while this year's group is at 38%)

Back to John's original table, and adapting it for "quick shots" rather than fastbreak points. Below is a revised table of...let's call it "quick shot" efficiency for teams. To explain again, I used John's same methodology and took the ranking of a team based on their Pace, and subtracted their ranking based on points produced on "quick shots" -- shots taken in the first ten seconds. Thus if a team is the 20th slowest team in the league, but produces the 2nd most points based on "quick shots", they score as a +18. Similarly, if a team is 2nd in Pace but only 20th in points produced on "quick shots", they are a -18. Here's the table (10SOL stands for 10 seconds or less):

As with John's table, Atlanta and Philadelphia are still way beyond most of the other teams, and New Orleans fares well too. However, teams like Charlotte and Dallas and Cleveland sink to a middle tier with very little separation, and the Knicks are no longer at the bottom, but part of  a big group of 16 teams between +2 and -2. I would conjecture that teams like Charlotte and Dallas are good at producing fastbreak points relative to their pace, but once they slow down even slightly, they tend to be most productive later in the shot clock. One other data point that struck me: Toronto's bottom ranking despite being one of the leaders in overall offensive efficiency. Presumably they do very well at producing points later in the shot clock, much less so in the first ten seconds.

Another (geeky) point that should be made is that using ranking differentials to produce this table is probably overstating differences a little between the teams. But I feel comfortable making the point that while the Knicks are a middling offense at best and were wisely slowed down by the coach because they don't have the personnel to be a strong running team... they're not quite the train wreck depicted in Krolik's original post. Especially if you take a more expansive definition of what a fast paced offense entails. In fact, they do pretty well relative to the rest of the league on "quick shots"...they're the only sub-.500 team along with the Warriors in the top ten in shooting percentage in this area:

Cleveland's shooting percentage is very good, but they're 29th in the league in percentage of shot attempts taken in the first ten seconds, given their preference to run at a slower pace. It's logical the Knicks fare well in this category (and in points produced), given their tendency to take trailing threes in transition (Gallinari most effectively). David Lee has also shown an increased tendency over the last 20 games to take a 15 to 20 foot jumper as the trail man on a "soft" fastbreak opportunity.

John made some other interesting points regarding other factors that characterize teams that generate points in transition: 1) generating turnovers and 2) giving up some defensive rebounding in order to optimize the number of transition opportunities. I'm not able to give these categories the kind of scrutiny I did to shooting and points produced earlier, but  I admit that at first I was a bit skeptical of what could be produced with this approach, and found myself pleasantly surprised at the correlation he unearthed between slower paced teams and better defensive rebounding numbers. John also puts forth the point that the philosophy of some transition teams may indeed necessitate some sacrifices in defensive rebounding. 

I remain somewhat unconvinced, though, that teams like the Knicks and Nets, for example, should get credit for having a transition-oriented philosophy based on poor defensive rebounding numbers -- they're just bad rebounding teams. I've watched the Knicks enough to know that they don't consciously leak out players to create break opportunities any more than any other team (they're not remotely in the Warriors' league), and if anything, their shaky rebounding stems more from size issues and talent, and also their propensity to do a lot of switching on defense and play a variety of zones depending on the matchup.

There's much more that could be said about the Knicks' strengths and challenges in making the D'Antoni system work effectively, but I'll save those for another post. Suffice it to say that I agree with Truehoop that you can't really judge the effectiveness of his system with the personnel he has, and that the coach *has* made some concessions to his talent -- he's slowed the pace down and no longer has the Knicks running at every opportunity, he's implemented different defensive wrinkles using Jared Jeffries as a defensive quarterback, and shortened his rotation. The elements of his system that do remain are the focus on crisp ball movement, and taking quick shots at opportune times -- with a few opponents, the pace is pushed when it's deemed to be advantageous (such as the recent games against the Timberwolves).  

These concessions may not be enough to produce a playoff team, and when they don't work the results can be unsightly. (Witness all the wailing on Knick forums that the players are "just chucking", now that the shooting seems to have gone south in the current slump). But these results don't imply the failure or misapplication of a fast paced philosophy. They do speak to the realities of a limited tactical hand when the ultimate goal is cap space and a fresh start after this season.