Category: Challenges & Insights: Generating a Price Guide of Trading Card Values

Recently I pulled a small subset of cards: the top-5 selling cards for each sports category (and the misc bin). So, that would be the top-5 selling cards (based on my dataset) in hockey, basketball, baseball, football and “the rest” (mostly, Marvel).

I haven’t much analysis — other than to say that, the cards are drawn from a dataset made up of a single consigner’s auctions (COMC). As such, the types of cards most likely to appear are skewed given the service’s price structure (e.g., it tends to cost $3.50 to 5.50 to list cards, so you won’t see high-volume cards that clear the market at a couple bucks).

So, without further ado…

Baseball Cards including Juan Soto; Ronald Acuna; Ken Griffey Jr; etc.

Perhaps most notably, while some cards boomed into July of 2020, all experienced steady declines thereafter.

Basketball Cards including Michael Jordan; Kevin Durant; Ja Morant; etc.

Pay close attention to the scaling on this graph. . . the y-axis peaks around $350, and so the lower valued cards that look like they have a shallow decline actually have a pretty brutal decline. Otherwise, I leave it to your inspection…

Football Cards including Patrick Mahomes; Justin Herbert; Barry Sanders; etc.

Perhaps more stability than the two graphics above… but, nonetheless, an uninspired environment. Curiously, I figured the hit to the market would be concentrated in the ultra-modern stuff, but looks like the Barry Sander’s card isn’t fairing so well either.

Hockey Cards including Cale Makar; the Hughes Brothers; Kaprizov and Lafreniere

Once again, hockey — like football — displays some relative stability, yet still not much to write home about. Incredible climb for Cale Makar, plateauing soon thereafter. Lafreniere’s value seemingly tracking his slow start.

Misc Cards including Magic; Stan Lee; Marcos Ambrose; Collin Morikawa; etc.

It was hard to pull a rounded sample here… I just wanted to pull the most frequently occurring cards, but ideally I would have covered golf; Pokemon; tennis; etc. Relatively flat trends, which I may need to look into more closely — suspiciously flat…

I am all for the explosion of parallels. They’re fun. But it also feels like the sports card makers are trying to pull the wool over our eyes. Consider the “1 of 1.” It is unique, which presumably means it is of scarcity that drives-up the price. A particular product might not only have cards numbered “1 of 1,” but also cards out of 10 and 100 and 1000. . . And. . . that’s all great. It is good fun to collect these cards. But when the number of parallels explode from a modest baseline — with many, many low numbered parallels — are they truly scarce? If a card company makes ten slightly distinct “1 of 1” cards, shouldn’t we just view them as out of 10? 

A great example comes with printing plates. They are technically unique. In hockey, their prices have always seems low for such rarity. I am sure some of that is just idiosyncrasy — for whatever reason, they just haven’t “caught on.” But, some of that is probably because every “one of a kind” printing plate is actually a “1 of 4.” Why? Because its hard to imagine the differentiation of the “cyan vs. magenta vs. yellow vs. black” plates is sufficient for collectors to really “buy-in” to a notion of uniqueness. 

If someone really buys into the notion that a unique card deserves greater value, however small the “differentiation factor,” then ask why no one is pressuring Panini (or Upper Deck or Topps) to have a random worker end of the production line randomly put doodles onto cards passing-by. 

The Comfort Zone: Manufacturing Scarcity

Of course, card companies are doing exactly what we would expect them to: increase sales. They tried, tested and (ultimately) confirmed that their sales improve upon creating dozens of parallel categories. Honestly, that’s fair enough.

But, the method dominates. As I see it, they are sticking to what they know…. they are sticking to what are comfortable with. . . which is to say, “manufacturing scarcity.” For us collectors, we need them to get uncomfortable. They need to innovate and elbow-grease their way into new hearts and minds.

As of now, they run the production lines, they own the printing presses… they are well positioned to do tweaks to existing products that ups their attractiveness-level. But, be warned! there is a fundamental contradiction at play: to create more scarce cards, the company actually has to print more scarce cards — thus watering down the specialness of getting a rare card, because now there are many other rare cards that could be pulled. Getting a rare card these days in not a rare event!

Going Beyond the Supply-Side: Manufacturing Demand

So, the problem: manufacturing perceptions of scarcity to drive sales can only go so far. In the end, these many unique cards (that are individually rare, but not as a class) need homes. There is only so much cash floating amongst existing collectors to bid these prices up. With the stock of rare cards growing, the collector’s pool of dollars gets diluted. And, so, we come to the main point!

The major players in the trading card industry must seek to drive-up demand for trading cards amongst new household, rather than focusing on ways to convince pre-existing hobbyists that to keep buying more of the many (seemingly) rare chase cards. In short, they have to drive the substance of scarcity, which is the demand-side of the equation. 

This is inherently uncomfortable. Card companies, for instance, are about making cool cards, not public outreach. But public outreach is what they have to do. A major driver of growing the hobby would be bringing youth and young adults into the fold. (At risk of sounding cynical, they are tomorrows money-makers.) If young people don’t pick up collecting, then card values will fall in the long-term, as the aging class of current collectors … well…. to be blunt… die-off. 

The “How” of Raising Demand

I’m just an obnoxious “armchair industrialist.” How to pull this off is not obvious; it will likely take many many prongs. But, a couple of thoughts…

One obvious starting point is ensuring a variety of products that are at once affordable and offer good value: this category of cards, targeted towards new collectors, can’t be an excuse for junk product, because then they won’t be coming back. It has to be fun to collect. Another option is to do giveaways for students. Most schools and universities are okay with “light-touch advertising” insofar as that simply means giving away freebies.  Likewise for organized sports teams/leagues.

It also means actively advertising. Likely, it is wise to pull-in folks that are already primed — such as handing out cards to families at sports games (generally an indicator of disposable income).  It might also mean including non-card prizes that those “lukewarm” collectors would appreciate chasing. (Not unlike the Tim Hortons hockey card promotion that takes place in Canada — which gives away gift cards, vehicles, chances to meet athletes, etc.) 

But, this is just a couple ideas, off-the-top-of-the-head. It is the industry’s brainstorm on which the steady appreciation (pray, not depreciation!) of millions (if not billions) worth of collectibles depends!

It’s been a roller-coster for the sports card industry (likewise for gaming cards). With the onset of COVID, card prices boomed. But, they have since cooled. With that said, not too many folks are trying to put down an exact number on how much prices went up (and how much, since, they have come down). So I will try to do a little of that today.

Average Price Movements by Type of Sport: Basketball, Baseball, Football and Hockey

I leveraged my database on card sale prices from COMC to come up with a few interesting findings. Using cards with 3 or more appearances in the data, I built an index of average card price movements; I controlled for the average price of individual cards, and just focused on tracking whether prices, on average, are going up or down (and by how much).

Let’s take a look at my first finding. Here, I just broke down price movements by broad category of sport: price fluctuations in baseball, basketball, football, hockey, and “all else” (which includes other sports — such as golf, tennis, soccer — but also non-sport — such as Magic, Pokemon, Marvel, Star Wars, etc.).

The data begins in early-March, with basketball cards at the beginning of their downturn. Notably, football cards kept going up in price for a while yet, before taking a dive also. Curiously, hockey and baseball demonstrated the most stability — limits on the upside, but the downside also.

The graph suggests basketball prices have roughly halved since their peak. Football managed to squeak-out another 30% before falling back to early-pandemic levels.

Price Movements by Price-Levels (Does “Top-Tier” Collectibles Hold Their Value Better?) {The Example of Basketball Cards}

We see the bubble for basketball cards deflating above.

When confronted with the challenge: “aren’t collectibles in a bubble?” many proponents of sports cards — particular, the new genre of “collectibles-as-investments” advisors — suggest that collectors should be okay if they focus on “the top of the market.” That is, buying high-end collectibles, rather than more low-to-mid-ranged product. We can theorize all day long about whether that ought to work or not… but we can also just dive into the data. So, taking the example of basketball cards (which have experienced the most notable deflationary trend amongst the major sports categories) here is my main finding… and some rather modest conclusions that I believe may be safely drawn. (Once again, I control for the mean value of each individual card in the dataset, then I track average movements in prices for each individual card.)

In the above, we see that the relatively expensive cards in the dataset (over $100) managed to keep their steam longer than the other price-brackets, but ultimately fell back down to their early-pandemic price-levels. In general, prices for basketball cards lost around 40% of their value for cards over $20 (they have lost around 50-60% of their value for cards under $20). While faring marginally better, an individual buying “high-end” product would not, on average, have fared well as an investor.

I don’t want to be a false-prophet, who promises much on thin evidence, so let me keep my conclusions to the obvious: even if your high-end investments might fare marginally better than your lower-end items, your immunity will still wane to downward market pressures. The bubble-bursting, it seems, will get to them too, in time.

Price Movements for New Releases: The Hype-Factor Appears all to Real {The Example of Hockey Cards, and Especially their E-Packs Format}

Finally, I split my dataset into “New Releases” and “Past Releases” to draw some modest conclusions about one final observation: it really seems that cards get a major boost just by virtue of being recent releases. But recent releases eventually become old news. So, what is the premium that collectors are paying to get the latest and greatest?

In the following, I look at hockey cards. The choice of hockey cards is intentional: my data is pulled from COMC, which has an agreement with Upper Deck E-Packs to allow cards to be transferred from E-Pack to the COMC platform for purpose of selling. As such, COMC is especially well-placed to list “hyped-recent-releases” rather quickly (relative the other sports).

Crucially, the trendiness demonstrate a serious premium to buy into a recent release. Products from the 2020/21 release years are shown to dive in value, whereas hockey cards from pre-2020 are shown to have held steady across the sample.

The size of the penalty is especially notable. Recent releases have lost approximately half their value since the beginning of the time-line, whereas older releases have held steady, and even made some marginal gains (again, I am looking at hockey as the subset of cards).

Final Notes: Be Wary those Over-Confident

Yes, some folks have done well by investing in sports cards. And perhaps you are especially astute too. But be careful. Maybe the are a good investment… I am not so smart as to tell the future. But some proponents of sports card investing seemingly do claim privilege over knowledge of the future. Ultimately, I want to make one key suggestion: the most important piece of data is missing.

A will write a future, in-depth, article on this very point: if cards are to be a steady, long-term, investment, then I very much suspect they need to be moving into the hands of long-term collectors — not short term flippers who can only drive short-term speculative bubbles. So, what is the first priority question that the industry should be surveying through a reputable pollster: “How much money do you [the interviewee] spend annually on collectibles that you intend to hold onto for the long-term?” The second priority question: “Given your economic outlook in the near-future, how much do you intend to spend in this coming year on collectibles to hold for the long-term?”

Making the (big) assumption that sport card companies will keep their supply-levels relatively constant, it would be tremendously useful to estimate how many dollars in demand exist, with which to buy-up the existing stock.

Once again, I’m just sharing a quick thought. I’ll aim for a more intensive article on the subject in the near future.

I’ll be the first to say it — despite pouring thousands of hours into this project, some (many!) card prices totally evade me. Recently a Wayne Gretzky rookie card sold for $3.75 million. My model pegged it at $175,000. So, I missed by a margin of 20:1. Ouch!

So, a moment of self reflection is merited:

First off, this is an evolving project. I am three months in on data collection (using my new methodology…). So, I can cast about for some excuses there: lots of cards just don’t have enough price points.

Still, a margin of 20:1 is pretty brutal, even with serious data limitations. So what else is going on?

Well, secondly, I gotta say there is a major challenge in estimating card prices because there are, seemingly, to be two distinct “data generating processes” going on — with a very ill-defined line cutting off cards from being in one set versus the other.

Very loosely speaking, there are some cards that have everything going from them: for the vintage cards, you might think of the uber high grade cards (from PSA and Beckett) of the top players in any given sport — like Gretzky or Howe, Michael Jordan, Mickey Mantel, Jerry Rice or Joe Montana, etc. It is my belief that these cards obey Exponential laws… tiny upticks in a card’s rarity blows-up the value like crazy. It feels as if every “big dollar” collector wants these cards and will spend what it takes to get them. A PSA 10 isn’t just worth 10% or 50% more than a PSA 9. It might be worth 1000% more.

Then there are the cards of players who are extraordinary but fall just short of being superheroes. Let’s get real… these guys and girls are amazing… they represent the top 0.001% of sports talent in the world — but… they aren’t quite in the top 0.0000001%. And, because of that, their cards go up incrementally in value. A linear relationship might exist between their rookie card as a PSA 9 vs. PSA 10 — maybe one is worth $100 and the other $150. More, but not crazily so. Perhaps the RPA of a 20 goal scorer in hockey might be worth 50% more than a 30 goal scorer — which, of course, feels right… a player that producers 50% more should be represented by a card worth 50% more. (And, yet, this relation falls apart when we are talking about 50 goal scorers. Their value might be 500% that of the 30 goal scorer.)

What’s to be Done?

So… the problem having been stated… and a plausible theory explaining the puzzle having been presented… what can be done?

Well, first-off, anyone doing data science for purposes of prediction is going to have to make trade-offs between doing data-driven results vs. parametric modeling.

Letting data drive your model can be great, especially because the estimates you produce are gonna be accurate — they are, after all, pretty much just an averaging of the prices that you observed for that particular card… BUT this method is only gonna let you predict for what you’ve already got. Obviously, that sorta sucks — if two cards are highly similar, why not use data from one to inform prediction of the other?

Letting theory drive your model can be great, especially because it lets you predict the value of cards that you do not have data for, but only if your theory (or theories) are… in reality… the primary one’s that drive results. Moreover, if you need multiple theories to explain what is going on with some subsets of cards, but not others, then your scope conditions need to be well-defined (which is one of the major challenges here)

So, in my case, I may have to sacrifice range for accuracy — to focus more on what I do have rather than extrapolation. As a theoretical sort of guy, this isn’t wholly satisfying, but the game’s not over…

IF I can think through how to parameterize my model to accurately divide cards into the exponential vs linear data generating process… THEN I am the playoff team that was down 3 games to 1, but who has suddenly tied the series back up.

Some of this I have already done. Some I am working on. Some is still yet to be realized. But, overall, I feel pretty good that’s where we’re headed.

In future posts, I’ll get into some of the gritty details of what’s driving my current models and how I intend to evolve them.

When it comes to trading cards (sports, gaming and otherwise), this isn’t the only price guide out there. Not by a long shot. And, yet, I think you should keep checking in with us. Maybe its a lack of humility, but here are our concrete advantages as I see them:

The Expert & the Data

(1) Data driven results from actual auction results crush bias. If you aren’t presenting data-driven prices, you are doing expert opinions. Now, to be certain, expert opinions are great to have. (Although, ideally, resources should be clear about the criteria and data that experts gather to inform their estimates.) But there are two disadvantages. First, experts are also experts because they love the hobby. That’s doesn’t have to be a bad thing, but it is hard to not let that affection creep into evaluations. By sticking with hard stats, you avoid this. Second, expert opinions can be swayed by asking prices and ideologies (i.e., the potential use of untested assumptions about what cards A, B and C should be worth, given my theory that X, Y and Z are the real drivers of value). But asking prices are not market prices. An item can sit in a shop forever if priced too high. And theories are just theories until they are not falsified with empirical tests. Auctions, whatever their faults, solve a lot of problems on these fronts: what is a card actually proven to be worth, given a reasonable time horizon to turn it over?

Including Inference in Prediction

(2) Statistical inference lets us understand what hasn’t yet happened. There are a lot of rare cards out there. Many have sold. Many have not. Some have sold but, like, 10 years ago. Price guides that don’t do statistical inference, but want to do data, can only give us historical averages of just those cards that they have data for. But, by making some assumptions (that are hopefully reasonable), we can say a lot about cards for which we do not, yet, have any sales data.

It is for this reason that I argue inferences are crucially important to building a price guide. Probability-wise, the more enticing a card’s value, the less likely there is any public records of its sales price. After all, the most interesting cards that you probably want to know about (because they are super rare!) are exactly those you probably don’t have strong market data on (why? because they are super rare!)

Inference allows us to take characteristics of cards that did sell, and make reasonable statements about what that implies for other cards (that have never sold) given how similar or different they are from it. So, let’s say we have prices for a bunch of Wayne Gretzky autographs, but we want to know the value of Connor McDavid autographs… or Michael Jordan, or Mike Trout, or Peyton Manning, etc.). Well, we can ask what what multiplier on value exists for two comparable cards for each player, and then apply it to autographs. Obviously, a good model will have more than that going on to conduct its estimations… but that hopefully gives an initial idea of how these models work. This inference also allows us to give updated estimates that reflect the current times. So, say you do know what a particular card previously sold for, long, long ago… well, without inference, that doesn’t mean you know what it is worth today. But, again, here inference comes through for us. We can use trends in time, about the hobby in general, and about characteristics particular to an individual card (e.g., has the player’s values, in general, been going up or down? Have rookie cards been getting a greater or lesser premium of late? ), to put down a reasonable guess on its value today.

SCVs Upshot

Some fellows are doing the first, and others are doing the second, but not many are making a point of doing both. That’s what I hope to offer… Also, the whole thing being free don’t hurt either.

Thank you for visiting my website, which launched in October, 2020, as part of a broader project to track the value of sports collectibles.

I hope, in time, to share a broad array of data and models with which to evaluate the value of your own collection. For now, I am doing my best to just get this website started.

That means a few things. Most importantly, this is a project that evolves. Both in quality of projections — as I learn from the data, I can better model my predictions (on the value of trading cards) — and of breadth — as I have time to collect more and more and more data, I am able to offer more estimates for a greater of variety of cards, both within and across categories (baseball, basketball, hockey, football, Magic, Pokemon… and on and on and on….)

Our Major “Value-Added”

For now, let me stress the website’s major motivation, its value-added: I want to share what I hope to do better, or at least different, than the other guys out there — many of whom, for the record, I have a lot of respect for.

First, I want to focus my project on auction prices, so as to not muck-up my estimates of collectibles’ values with the millions of unreasonable asking prices that are out there. Second, I care about how estimates are produced. In my next writing, I will share with you how (i) statistical methods matter — by freeing us of our potential biases (whether as expert appraisers or hobbyists) in appraising cards we have attachments to — and how (ii) statistical inference can allow us to learn a lot about the world, with relatively little data. We don’t need to observe the value of every single card in the universe to make reasonable estimates about what those cards might be worth.

And that brings me full circle to a closing comment: we’re evolving. Its hard to make good estimates. But, as we learn what characteristics of a card best predicts its value (be it type of sport: baseball or basketball; be it the player featured: LeBron James or Tom Brady; be it the series: Topps or Upper Deck or Panini; or be it the value-added features: whether that’s memorabilia (jersey, patch, stick, bat, glove, ball…), autograph, rookie card status, etc.

So, please stick with me. Check-in (often, if you’d be so kind), and let’s learn something about this hobby together.

Best,

K.S. @ sportcardvalues.com