Packard Forum Index » General

Dear All,

I had previously plotted the owner's registry by year. I have now reproduced this plot by series, which is far more accurate. This also allows me to plot the number of cars produced by Packard on the same graph. A couple of observations:

1) Prior to the 54th series, the number of cars registered at PInfo corresponds remarkably well with the number produced. The exceptions are:

3rd -5th series are under-represented by 55-89%
10th series is over-represented by 310%
11th series is over represented by 400%
17th series is over represented by 60%
26th series is over represented by 56%
54 and 55th series are over represented by approx 250%
56th series is over represented by nearly 800%

2) If the average representation is correct, then .4% of vehicles produced, on average, are registered at PInfo. This analysis excludes the 54-56th series data.

I have attached a pdf of the plot as well as the data file (as an .xls file).


Happy pondering!

Tom

ps. BigKev, if you can have the number of cars registered for a given series/year show up for that search, it would make updating and editing this data much easier.

Attach file:

---

  (88.39 KB)

Size: 122.83 KB; Hits: 53
Size: 69.50 KB; Hits: 45

Here is a plot showing the relationship between cars registered and cars produced. Each point is labelled by series.

Tom

Attach file:

---

  (61.20 KB)

Nice job. This falls right along the lines of what I hoped could be generated from the Owner Registry.

As the vehicle count grows, the possibilities for graphical display are only limited by software and imagination.

Keep up the good work!

Thomas,

you might be interested to read this thread.
Given that the members of PackardInfo are a representative sample of Packard owners your graphs seem to contradict or at least not to support another, more linear hypothesis ("attrition rate").

Again, keep up the good work!

I have looked a little more closely at the data. I decided to see what happens if I examine the data in three chunks:

1st chunk: series prior to the launch of the 120
2nd chunk: remaining pre-war series
3rd chunk: post-war

I based this division on the following:

a) production volume, and the audience of customers, changed dramatically after the introduction of the 120. The change in volume and affordability as both used and new vehicles, I hypothesize, will effect the proportion of production that has ended up as collector cars.

b) The post war market was significantly different from the pre-war market in both volume and consumer tastes

c) post-war cars may have a higher chance at surviving due to age and the war-time scrap efforts.

Observe the plots below.

Attach file:

---

  (41.01 KB)


  (41.10 KB)


  (34.88 KB)

Quote:

Guscha,

The data suggest that for the 12th through 22nd series, car survival does not reach zero, but reaches some relatively consistent proportion of the total produced. For example, after a period of attrition, approximately 1% of a years production remains in collectors hands.

For that time frame, PInfo registered cars very closely track 0.4% of production. So, if, as has been suggested, that pre-war production survival is around 1%, then 40% of the surviving cars are represented in the Owner's Registry.

The data for post war is particularly striking, and really argues against a general attrition rate being responsible for what cars survive today. There are probably plenty of cars in the registry for that time frame for statistical validity (unlike the 3rd-11th series), but there is absolutely no correlation between production and cars registered. Given that it has been over 55 years since the last 'true' Packard, I don't think this is a result of the 'attrition' hypothesis. It is also clearn not a result of some set percentage of vehicle production surviving.

Some possible hypotheses for post-war data are:

1) collectors of these vehicles are more active online, and so more likely to register their cars.

2) Packards have cachet, but Packard's with a Packard V-8 are freaking irresistible.

3) Many more living car collectors could have been conceived in a Packard V-8 than in other production years, so those years have a greater nostalgia value.

Tom

I enjoyed looking over the distibution of plot points in your first scatter graph, but these new graphs are a bit over my head. Not sure if it would help my perspective, but should the plot points be identified by their Series in these three?

While the Owner Registry only represents members of PackardInfo who are willing to share the details of their Packards with us, that's really no different than any other attempt - except, perhaps, for access to and willingness to use the Internet. I'm still hopeful that, as we get more prewar content for the online archive, our Registry will become a more representative sample as it continues to grow (with a bit of help from Mal).

Fooling around with the data, and I decided to plot the percentage of cars that Packard manufactured that are currently registered with Packard Info. The most highly represented series is the 56th, with almost 0.35% of all the cars Packard made that year now residing with members of Packard Info! Prewar, the highest proportion of production represented goes to the 10th and 11th series.

Below is the plot.

Attach file:

---

  (70.51 KB)

Tom, interesting graph. I wonder how the number of Packards registered with the PAC compare with those registered here? It would take a bit of time and effort to do the counting, but may be worth it (for the curious minded of us). Just thinking out loud.

(o{I}o)

Tom, I could not pass up an opportunity to count something. I counted 424 1956s (Clippers and Packards) registered with the PAC. How does this compare with the number of 1956s registered here? Of course there will be duplications. I would think that the PAC registered 1956s would be larger than the ones registered here, and may be a better indicator of percent surviving.

(o{I}o)