My memory—and the Beckett database had already confirmed what I wrote in yesterday’s column. Former Super Bowl III standout Randy Beverly never had a football card, I decided to do a quick online search just to make absolute
In writing about Randy Beverly yesterday, I decided to do a quick Internet search for any cards of him although the Beckett Database as well as my memory had already confirmed no cards were ever created for one of the stars of Super Bowl III. Imagine my surprise when I saw one cross-listed on Amazon and COMC. However, clicking the link revealed it was actually a 1989 Pro Set card of New England Patriots coach Raymond Berry, not Beverly.
Seeing that listing reminded me that there is now a year passed since COMC and Beckett went their separate ways and got me to thinking about some of the progress and the tribulations COMC has had in the past year since they began the long process of creating their own catalog, with crowd s
ourcing help from their many users.
Speaking from my Beckett experience with the Almanac of Baseball Cards and Collectibles I do realize what a long process the whole issue is. I was working primarily with just baseball and I would say about 90 percent of the database was already in shape when we moved to a more accessible system in 1996. There were several issues such as four different listings for the obscure but really cool 1898 Cameo Pepsin pins. To merge those and create one checklist was a much smaller issue than anything COMC has had to do lately.
While building their own catalog, COMC continues to have to work on their core business as well as always trying to think of new ways to monetize what they do. It’s all part of a young company’s growing pains and I’m sure it is all very laborious.
The Beverly-Berry listing is another example of an unintentional error caused by creating a system to ensure names are in the database as effortlessly as possible. Yes, the two letters for the first name and two letters for the last name (RA-BE) caused this error.
Nearly 20 years ago when I worked at Beckett we had a very similar data gathering issue. At the time we were using the three letters for the first name and three letter for the last name to gather information on players who were hot. We were inputting “Eli Mar” in trying to get information on the Cardinals young catching prospect, Eli Marrero. Instead we ended up gathering information on Royals infielder Felix Martinez because his name also contained those six letters in succession. Randy Beverly popped up because COMC used a two letter by two letter combination of ‘Ra’ and ‘Be’.
One thing we learned at Beckett was if we had a mistake in our data, it needed to be corrected immediately. COMC needs to follow the same principle and not leave mistakes such as having Tyler Colvin pop up as Ty Cobb taking nearly a year to fix. I realize correcting errors can be a long and tedious process but all known errors should take immediate priority. You cannot move to the next level until your important data is ready to go.
Having said all that, the progress COMC has made in their data base and checklisting effort in the past 12 months has been quite heroic and unlike the two previous big card databases (Beckett and Krause–now F&W), they are continuing to make an effort and pour resources into creating and improving the data they have. Beckett, is of course, continuing to add the popular mainstream sets they need to cover but they no longer have the resources or the time to add the more obscure or oddball sets into their system. Here is an example of a 2012 set which is not part of the Beckett data base.
To be honest, the effort to add any set to the Beckett database may not be financially important but as I’ve discussed before, it’s important to catalog these sets so their Marketplace users have a continually updated listing they could use. There are a lot of team and player collectors who need the information and have money to spend.
One advantage COMC does have is that by acting as essentially a consignment seller, they are actually handling cards so they can correct card company checklists if those lists turn out to be incorrect. COMC has now processed over ten million cards and they are continuing to get new cards into their system every day. But, as they are learning, and I know from experience, the process of ensuring accurate checklists and names is never ending and there will be bumps along the way.
Related Stories:
Ramblings: Don’t Kill Cataloguing
Ramblings: COMC’s New Partnership
Ramblings: Imagining the Hobby Without the Internet
Ramblings: COMC Hoping You Can Help Build Card Database
Rich Klein’s Ramblings: Sorting to a Science
Launch of The COMC Challenge: Help Complete Checklists
Ramblings: Past Not Always Better
