I take collation pretty seriously, mostly because I enjoy knowing what I'm going to get once I know the top card (or sticker, as the case may be). For this box, if Tom Seaver was first out of the pack, your four others would be in this order: Jerry Remy, Dusty Baker, Mike Schmidt, and Vida Blue. Similarly, if you pulled Paul Molitor first, there was nearly a 100% chance that the third sticker would be Manny Trillo, the fourth sticker would be Bob Horner, and the second and fifth stickers would be random.

Even the seemingly random-seeded stickers weren't seeded all that randomly. For instance, Andre Dawson was locked in the 1 hole, John Castino and Fred Lynn both 2's, John Mayberry a 3, Pete Rose Highlight a 4, and Cecil Cooper, Cliff Johnson, Dave Collins, and at least 18 others at 5.

There were "Rogue" stickers as well, or those that didn't appear in a consistent slot (Keith Hernandez, Rollie Fingers League Leaders). And there, hidden amongst the doubles, triples, rogues and locked rows were actual single print stickers – 25 of them to be exact, including all five of the stickers from Pack 1. Amazing.

Someone asked me today what my goal of this project was. I answered that I wanted to be able to hypothetically assemble an uncut sheet of stickers without knowing exactly where each sticker would have fit. And while that would be a neat exercise to actually do, the more I thought about individual box collation, pack cycles, and pack pockets, the more I realized that to make generalizations based a very small sample is at best unwise and at worst just plain stupid.

I remember a few years ago when I started A Pack A Day, I ripped a box of 1989-90 Hoops Series One. I found that not only was the David Robinson draft-day rookie short print not actually short-printed, but I found that if you got it in a pack, you also got Larry Bird (I ended up getting about 4 of each from that one box). But while my experience tells me that the Robinson wasn't actually a short print, why is it always listed as a short print? Using my one-box example to make a generalization doesn't really work.

There are reasons why card companies serial-number cards nowadays. For one thing, it creates a sense of a limited supply. For another thing, it completely destroys the idea of a general collation. Just because you get autographed relic card A doesn't mean you'll also get commons B, C, and D (well, unless the auto relics are considered Rogues, then that opens up whole new possibilities...). Of course, this is not true of regular card products. I'd bet that Topps 2010 has just as poor collation as Topps 1986.

In any event, if you can view a Google doc, you can take a look at my documented collation from my box of 1982 Topps Stickers.

Colored-coded Collation