ANNOUNCING "CELTICS ADVANCED STATS 018": AN ONGOING FEATURE OF THE BOARD

[Sam]

We all know that there are many basketball stats flying all over the place. Very often, those people who know where to find appropriate stats have an advantage when it comes to analyzing game results, monitoring the progress of the team, or supporting their posts on a message board.

We think there are two types of basketball stats: (1) “available” stats such as points, rebounds and assists (one just looks them up) and (2) “derived” stats (either one calculates them by combining two or more available stats via some sort of formula OR one keeps the stats while watching a game).

Within a couple of weeks, we’ll be posting a list of sources of available stats (in addition to many other useful links). Moreover, starting with the first game of the season, we’re planning on providing a number of derived Celtics stats, both game-by-game and based on a cumulative seasonal trend. To facilitate your analysis, the seasonal trend data will be shown in total, versus eastern teams, versus western teams, for Celtics winning efforts, and for Celtics losing efforts.

Here's how it will work

• I have a list of “derived stats” that are not usually reported but which I’ve often found meaningful in analyzing games and identifying trends in the team’s performance. I’ll post them game-by-game and on a cumulative basis.

• Here are the stats I’m going to be responsible for:

√ “Enabling Chemistry”: Ratio of Celtics assists to made Celtics field goals

√ “Work Ethic”: Percentage of Celtics made field goals that are two-pointers

√ “Defensive Alertness”: Ratio of the sum of Celtics successful gambles (steals plus blocks) to unsuccessful gambles (personal fouls committed by the Celts)

√ “Offensive rebounding efficiency” (percentage of the rebounds coming off the Celtics’ offensive boards that are captured by the Celtics)

√ “Defensive rebounding efficiency” (percentage of the rebounds coming off the Celtics’ defensive boards that are captured by the Celtics)

√ "Offensive Sharing" (percentage of Celtics players in double figures)

Beat has agreed to provide me with available stats so that I can calculate:

√ “Lunchpail Effectiveness” (points in the paint as a percentage of total points).

√ “Transition Effectiveness” (fast break points as a percentage of total points)

√ "Opportunism" (total shots—both field goal attempts and freebies—as a ratio of total possessions as measured by shots plus
turnovers)

• If anyone would like to suggest any other stats to add to the list (please make them “derived” stats…not available stats that anyone can easily look up) please feel free to make the suggestion. There’s just one catch: If you suggest it, you get to be responsible for providing it to me after each game (by private message within 12 hours after the end of a given game). And you'll have to commit to doing it every game or having another person fill in for you.

We hope this feature will be interesting to people and will provide a nontraditional resource. I’m particularly looking forward to seeing which (if any) of these “derived stats” may be most influential in making the difference wins and losses.

Sam

[beat]

Sam

I will certainly give you the updates after the game ASAP reguarding pts in the paint and fast break hoops.

One other area that might be interesting is a turnover to fg/fs (total) attempt ratio. Obviously the better one takes care of the ball there will be more shot attempts.

Although it would be impossible to do for any length of time I'd be curiuos on average how far the shot clock runs down before a shot is hoisted and at what intervals is the most success achieved. 24-16 sec 15-8 or 7-1 left on the clock.


beat

[Sam]

Beat,

I can add your first suggestion without any additional research because it involves two stats that I'll be using in other calculations. Can you think of some snazzy nickname for the stat? (I've been trying to jazz up each one with a descriptive nickname that also suggests the unique value of the stat).

You may recall that, at least a couple of years ago, I went through a several-game span during which I posted the percentage of Celtics shots taken with fewer than 10 seconds, 10-14 seconds, 15-19 seconds, and 20-24 seconds left on the clock; and I also posted the shooting percentages for each time segment. I thought it was very useful, and it would be conducive to something called a "Pace Efficiency" stat.

The problem is that keeping that stat requires intense concentration, to the extent that the statkeeper can't take a second off to watch any other phases of the game and can barely squeeze bathroom time out of timeouts. I had to video each game and run it back because invariably something would get past me.

I'd love to include that stat because it's my theory that the most effective shooting takes place before 15 seconds have elapsed on the shot clock. But I'm not prepared to invest the energy except on an occasional basis. If you can enlist the aid of several people to share that duty on some sort of rotation basis, I'd be delighted to include the stat and even to do it on opening night.

Just let me know.

Sam

[beat]

Sam

The problem is that keeping that stat requires intense concentration, to the extent that the statkeeper can't take a second off to watch any other phases of the game and can barely squeeze bathroom time out of timeouts.


And no drinking either!!

As for a snappy name the "TURNOVER to TRY" ratio is all I can come up with.

Also should we consider just shots as is or "weigh" them a little differently. Meaning that foul shots count 1 and all fga count 2. If a team shoots a lot of fouls they get less fg attempts off. Just thinking, perhaps too much.

beat

[Sam]

Beat,

We can just divide the free throws by 2.2, which will come close enough to equating field goal attempts and free throw attempts. (I used more than 2.0 because around one-fifth of all field goal attempts have 3-point potential.)

Sam

[beat]

Sam

Fine with me!

Hope you have batteries for your calculator........and your mind !

beat

[swedeinestonia]

I think that if you would look at fg% during different periods of the shot clock it would only make sense if there was a higher % earlier on.

I would think that an experienced player would consider the shot clock to be a period time during which you will want to take your best shot. If you have a 75% look early on you will take that shot but if you have a 30% look you will pass and hope that a better look will appear.

So it would take a good look for somebody to shot early in where as later in you start settling for lesser shots since the chances of getting a better one diminishes.

Some stats people would interpret that straight off that it is BETTER to shoot early for some reason, not that is just a result of a situation, they think they can cause the situation by shooting more earlier.

Some of it can of course be accounted for fast breaks and the fact that the defense has not settled in yet.

So high fg% for early periods is a good thing in the way that it means that the players know how to value the looks well.

[swedeinestonia]

This might be "duh, no sh*t" thing for you guys though

[Sam]

Swede,

I think it's possible to identify a certain pattern for a team in terms of when, in the shot clock, they tend to shoot when in the halfcourt game as well as how frequently they run the break. Certainly there will be exceptions. But I'm talking about the normal pattern for a given team.

A few years ago, 82games.com published a report on the 2003-04 NBA season, breaking out four shot clock categories within which a team could shoot:

0-10 seconds (which they called "quick possessions": breaks, putbacks, turnovers, etc.)
11-15 seconds ("early resolution")
16-20 seconds ("structured possession")
21-24 seconds ("late possession": often rushing

For each NBA team, they showed (1) the percentage of total shots that fell in each category and (2) the shooting percentages for shots taken in each category.

I did a rough job of estimating the overall results for each category (because they didn't provide totals for all teams combined). Basically, I took the highest and lowest percentages for each category and calculated the midpoint between them. (Someone may prefer to spend the time on obtaining the exact figures.) Here's what I arrived at:

0-10 sec. represented 41% of the shots and an avg. shooting % of .483
11-15 sec. represented 23% of the shots and an avg. shooting % of .430
16-20 sec. represented 23% of the shots and an avg. shooting % of .419
21+ sec. represented 13% of the shots and an avg. shooting % of .388

So the lesson might seem to be that shooting more quickly yields a higher shooting percentage. But the "0-10" shooting figures are obviously padded by point blank offensive rebounds (and probably by playground, low-defense garbage time too). And it figures that the "rushed shot" category would be lowest.

But what I find most interesting is that, in the two categories that probably represent the most strategically resolved halfcourt situations (each comprising 23% of the shots), shooting within 11-15 seconds seems to yield better results than taking 16-20 seconds. My theory is that, by the 15-second mark, the defense is better set and the basic structure of the offensive strategy has become more obvious to the defense.

Which leads me to wonder why those two shot clock segments aren't skewed more toward shooting within 11-15 seconds. I'm certain that's what happened in the more formative years of the league. In fact, in the 60s, I'd bet the distribution of shots would have been more like:

0-10 seconds 50%
11-15 seconds 30%
16-20 seconds 15%
21+ seconds 5%

In the Russell Years history I wrote, I refer to those years as featuring "volume shooting." Which means worrying less about shooting percentages and more about hoisting a lot of shots—especially when the defense was less set. And shooting percentages were lower in those days, partly because of the tighter rims but also undoubtedly because of the quicker shooting mentality.

Moreover, I think the advent of the three-point shot has slowed down today's shooting sequence because, in general, it takes longer to set up a three-point shooter.

I'll leave it to each individual to decide for yourself which style would represent the more effective strategy (more shots at a lower percentage or fewer shots at a higher percentage including threes). And which would present the more exciting spectator experience.

The URL for the 82games.com report is:
http://www.82games.com/clock.htm

I wish they would update those stats, but I can't find any record of it.

Sam

[swedeinestonia]

I do not think that shooting early necessarily gives the advantage. I am thinking that is more of the fact that you are always trying to get the best shot within 24 seconds.

If you happen to have a totally open look you will take it and if not then wait until later. The further down the shot clock you get the more likely you would be to think "there will not be a better shot" and settle for worse and worse shots.

I do not think the lesson to learn is to try and shoot as early as possible, but to try and put pressure on the defense early to get good and "more" looks.

[Sam]

Swede,

I didn't mean that rushed shots are good, and I've gone back and changed a couple of words in my text to clarify that it's not necessarily the case. Thanks for identifying some sloppy writing on my part.

You're absolutely right. The key is to put pressure on the defense to get early and good looks. The key word is "pressure." It's not enough just to look for early, good looks. There should be some aggressiveness, starting with pushing the ball up the court. I don't necessarily mean fast-breaking all the time. I'm referring to getting it into the forecourt with as much time as possible to get the halfcourt offense going.

It drives me nuts when a Celtic walks the ball into the forecourt. Yes, there will be times when the guy may be at a 1-to-5 disadvantage and he needs to slow it down. But, as a rule, I like to see the ball push.

And quite frankly, as a rule (not always), I'd prefer to see an occasional rushed early shot than a deliberate, standaround offense that too often leads to desperation heaves. And that goes for end-of-quarter shots too. I hate it when they pound the ball into the floor while the defense gets set and a bailout shot becomes a great likelihood. Push the ball and get a good shot regardless of the clock, and then get back and play good defense and let the other team try the desperation shot.

Sam

[Sponsored content]