Autograph Numbers for Chamber of Secrets

[Guest]

[allender]

Updated 11 May 2007. I've seen breakdowns now for 292 boxes, or 4.87% of the packaging run. I separated the signers below by the number of "clusters" in case types, in case it really means anything. There have been 13 case clusters (by both autographs and other high chase)  out of 23 cases, but I suspect the total number of clusters is 15 because I see 5 each for two groupings of cel cards and only 3 clusters for the final cel threesome.

# reported - approximate number made - signer
    ---[3]---
  9   -   220   -   Shirley Henderson as Moaning Myrtle (R)
    ---[5]---
  9   -   220   -   Fiona Shaw as Aunt Petunia
11   -   315   -   Gemma Jones as Madame Pomfrey
    ---[6]---
11   -   325   -   Chris Rankin as Percy Weasley (R)
13   -   325   -   Harry Melling as Dudley Dursley (R)
14   -   325   -   Kenneth Branagh as Professor Lockhart (R)
    ---[8]---
17   -   440   -   Miriam Margolyes as Professor Sprout (R)
18   -   440   -   Julian Glover as Aragog
19   -   440   -   Richard Griffiths as Uncle Vernon
    --[10]---
21   -   540   -   Bonnie Wright as Ginny Weasley (R)

The totals are adjusted to 3600. There should be only 3000 inserted in 6000 boxes, but the "overage" numbers for damaged/replacement cards for costumes (the total of the x/yyy numbers above 6000, which would be the total at one-per-box) is 11.1%, props is 22.22%, and film cels is 20%. (For instance, CFC cards total 9 times 240 or 2160, 20% more than the 1800 we expect to be in boxes at 3:case.) So I used 120% to estimate the total numbers signed.

[allender]

We have one more potential case "cluster", and now all chase but P6 have been reported in one case type or another. But it's tough to decide something's really a case type until you get at least three that are the same. (If you have 6 perfect clusters for 600 boxes, then the distribution of each card would have to be 100, 200, 300, 400, 500, or 600. So there have to be some "miscellaneous" cases.) And maybe some of our reports are from the miscellaneous bin.

A - 2 cases - Lockhart, Myrtle, Pomfrey, Vernon, Ginny with CFC 1-3-9
B - 2 cases - Aragog, Petunia, Pomfrey, Sprout, Ginny with CFC 4-5-8
C - 2 cases - Dudley, Percy, Sprout, Vernon, Ginny with CFC 4-5-8
D - 1 case  - Aragog, Lockhart, Myrtle, Sprout, Vernon with CFC 2-6-7
E - 1 case  - Dudley, Percy, Lockhart, Petunia, Ginny with CFC 1-3-9

But now we have each CFC card represented in two clusters except for 2-6-7, and a sixth main cluster with those cels would make things symmetric.

If you use the box breakdowns that aren't part of cases and the number of completed ebay auctions as a guide (dangerous), we would need more Aragog, Petunia, Pomfrey, Sprout, and Ginny.  That would make Dudley, Myrtle, and Percy the scarcest of the autos at about 200-240 each. That's the purpose of all this early math, trying to guess the market for the scarcest items.

[blackcat20]

Not sure if this helps, but it looks like the shared Aussie case is the same as one of Jackies...I pulled Aragog, Ginny and Pomfrey from my boxes (with the same associated costume cards) which would most likely make it Case type B; but I guess we have to see what the other girls pull from their boxes

[5594mike]

wow allender you have a great eye for card cluster and a very good calculater, i salute you mate

[allender]

Well, I try to keep this all on one page with updates. However, in this case I figured I needed to leave the old message there as part of my application for Professor of Divination.

Yesterday, I speculated that it looked like a sixth cluster should have CFC 2-6-7, and that the number of completed ebay sales hinted (dangerously) that we seemed to have been short of reports for autos by Aragog, Petunia, Pomfrey, Sprout, and Ginny. Those might be part of an unobserved sixth main cluster of cases.

So pinksamstar came through and reported a new case breakdown with exactly those eight cards. 

We already had a cluster with those five autographs. So I can't say yet whether this cluster should be called "F" or "B2". Memorable Moments seemed to have 6 autograph clusters, but two of those had subclusters of different costumes/props.

There don't seem to be enough C1-C6-C9-C15 or P6-P7-P8 yet, and probably a bit too many C16-C17 and P9 for them to be included in every case: 115 copies of P9 can't cover every case of two "100-case" clusters. Maybe some of them are displaced by P6, which hasn't shown up in any posted breakdown yet.

But here's the working assumptions.

A - 2 cases - Lockhart, Myrtle, Pomfrey, Vernon, Ginny with CFC 1-3-9 and P8
B - 2 cases - Aragog, Petunia, Pomfrey, Sprout, Ginny with CFC 4-5-8 and C10/P3
C - 2 cases - Dudley, Percy, Sprout, Vernon, Ginny with CFC 4-5-8 and C13/P7
D - 1 case  - Aragog, Lockhart, Myrtle, Sprout, Vernon with CFC 2-6-7 and C9
E - 1 case  - Dudley, Percy, Lockhart, Petunia, Ginny with CFC 1-3-9 and C1/C6/C15
F (or B2) - 1 case  - Aragog, Petunia, Pomfrey, Sprout, Ginny with CFC 2-6-7 and C10/P3

Because of the "short" props and costumes, I now speculate that there's a seventh and final cluster (or several subclusters), which would include:
  C1-C4-C6-C9-C11-C12-C14-C15 and two from C3-C7-C8
  P6-P8 and one from P3-P4-P7

A seventh cluster (or naming "F" a subcluster) would break the current symmetry of the CFC cards. Let's hope we get breakdowns from some of the people who bought five or more cases.

Still, I'll post this and wait to see if some cases show up that are exactly like this. Divination isn't nearly as convincing if you wait until after the events before you predict them!   

**NEWS FLASH**
I re-analyzed the breakdowns for Memorable Moments, and it looks like the numbers come out almost perfectly if you assume that there were five basic autograph clusters and two sub-clusters for each autograph cluster. Of course, they might find a different arrangement was convenient for COS.

[allender]

Second news flash, it's all breaking down. Even Trelawney can get one right every once in a while.

We've got more case patterns already than we saw for MM or TSS. Maybe a lot more breakdowns will say that there are fewer "major" patterns and that some of the clusters are only a single case, which would allow us to say there was a "miscellaneous" cluster of cards that didn't fit into the even numbers. But it's going to be back to estimating autograph numbers "the hard way", from the total autos reported in all of the breaks. I updated at the top to reflect cases and boxes that cover 2.2% of the packaging run.

[Trey]

Not sure if this helps, but it looks like the shared Aussie case is the same as one of Jackies...I pulled Aragog, Ginny and Pomfrey from my boxes (with the same associated costume cards) which would most likely make it Case type B; but I guess we have to see what the other girls pull from their boxes

..and I got the Petunia auto...and the props/costume breakdowns are the same so it definately looks like we got a 'B' case.

[Sharon]

I finally got around to posting my case breakdown and I think I may have another auto variation case.  The autos were

Vernon, Percy, Ginny, Aragog and Pomfrey

Very similar to several clusters, just a bit different.

[allender]

I finally got around to posting my case breakdown and I think I may have another auto variation case.  The autos were

Vernon, Percy, Ginny, Aragog and Pomfrey

Very similar to several clusters, just a bit different.

Yeah, and your distribution of the other chase cards was different, too. It started out following the "pattern" because there were only four combinations in the first 7 cases, and we also had a report of 3-for-3 identical cases that wasn't posted in the breakdowns. But since then, every case has been different. My one bout of successful divination seems to have been a fluke.

All I can really say for sure is that the costumes that are numbered to 640 or 665 will probably show up in every case. ... And each case (so far) has given one of only three triplets of cel cards. We have six of CFC4-5-8 now but only three each of CFC1-3-9 and CFC 2-6-7, and I'd cringe less about the autograph totals if these were balanced better.

Oh well, if it turns out these things aren't very predictable, then I'll save some work for the next set. The number of breakdown posts is declining, with an average of nearly 10% of the run for the previous 4 sets but only 2.5% for COS up until now.

[Sharon]

The number of breakdown posts is declining, with an average of nearly 10% of the run for the previous 4 sets but only 2.5% for COS up until now.

Now that is interesting.  I wonder if it is because people just don't get around to posting the breakdowns or if the 'regulars' are just not buying as many.  I suspect the latter.  Oh dear well maybe too many sets, too close together with too many repeats is starting to take it's toll.  Artbox were warned.

[allender]

Now that is interesting.  I wonder if it is because people just don't get around to posting the breakdowns or if the 'regulars' are just not buying as many.  I suspect the latter.

It's a bit of both, because the former is certainly true too. Not a huge increment to the earlier stats, but I posted an update for 16 February.

[allender]

Ruminations as of April 1st. akimira posted what looked to be one of the missing "clusters", the only posted case-break with a P6.

Obviously things aren't as severe in "same cases" as some previous sets, but there are 13 unique case distributions out of 23 cases, with only 6 of the 23 different from any other posted case. So (on average) Artbox has made the cases more random, which doesn't satisfy you quite as much if you're the unlucky person who broke three cases, all the same.

I see five case types each with CFC-267 and CFC-458, and only three posted with CFC-139. I assume that all of the CFC are equally packaged, and then some simple math can explain a total of 15 clusters with about 41-48 for every single high-chase card per cluster. For example, C3, C7, and C8 are numbered abover the total number of cases (600) and there's one in every case (43-44 in each of 15 clusters). P2 is in four case clusters (# to 190, 47 in each cluster without counting holdbacks for replacements).

If this holds up, then the two "missing" clusters both have CFC-139 and C3-4-6-7-8-11-12-14. Each of the two holds three from P1-2-4-6-7-10 and two from C1-13-15-17.

As for autographs, it depends a bit on what we haven't seen so far. But Myrtle is in 3 of 13 clusters, Petunia-Percy-Pomfrey-Dudley-Lockhart in 5 of 6, Sprout-Aragog-Vernon in 8 each, and Ginny in 10.  230 of the 270 reported boxes are from case breaks, but adding the 40 from non-case box breaks doesn't change the order for estimated scarcity. Fiona Shaw might be a bit scarcer than the cluster calculations hint.

[allender]

Updated 11 May 2007, covering 292 boxes or 4.87% of the packaging run.