March 2023: The order that looked perfect on screen
In March 2023, I approved a custom print run for our indoor recreation venue group. We were launching a weekly poker night and a casual MTG league. The plan: 1,200 custom printed poker cards, 5,000 custom MTG tokens, 300 poker cut cards, and two new poker table sets. I had been handling custom game orders for ice-games for nine years, so I thought I knew the drill.
My initial misjudgment? I assumed if the digital proof looked sharp and the printer confirmed 300 DPI, the rest was just paper. A custom poker card is a custom poker card. A custom MTG token is just a smaller card. I was wrong on both counts.
That order cost us $3,800 in reprints, expedited freight, and one very awkward opening-night apology.
The process: proofs, Pantone, and a false sense of safety
The artwork looked great. We had a deep navy back with a white border, a red accent on the MTG tokens, and a matte finish requested for the cut cards. The printer sent a PDF proof. I zoomed in. Edges were clean. The Pantone 286 C blue matched our brand guide. I approved it.
Here is the thing: a PDF proof does not tell you how a card shuffles. It does not tell you how a sleeve fits. It does not tell you whether a cut card will slide out of a dealer's hand or get mixed into the deck. I had been treating print like a visual purchase. Commercial card printing is a handling purchase.
We also ordered two poker table sets. Those arrived first, and they were fine. Speed cloth, padded rails, chip trays. No issue. That made me overconfident. If the tables were good, the cards would be good. Right? Wrong.
The turning point: opening night and 47 small failures
The first red flag came from a regular player. She tried to sleeve a custom MTG token. It would not fit. Not even close. We had specified standard poker size because the printer's order form used that as the default. But Magic: The Gathering cards are not poker size. They are 2.5 by 3.5 inches, or 63.5 by 88.9 millimeters. Our tokens were slightly off. On a table, nobody noticed. In a sleeve, everyone noticed.
Then the poker cut cards started showing up inside the decks. We had ordered them with the same dimensions and coating as the playing cards. Bad idea. A cut card needs to be visually distinct and easier to handle. Ours blended in. After two hours of play, the dealer was sorting cut cards out of the deck every few hands.
The custom poker cards themselves? The colors were close, but the stock was too light. We had chosen a 270 gsm cover stock because it felt substantial in a sample pack. It is a heavy business-card weight, not a playing-card weight. After a few shuffles, the corners started to nick. By week three, regular players were asking for new decks.
People think custom printed cards cost more because of ink or color. Actually, the cost often comes from setup, waste, and low-volume runs. The real killer is a reprint. A reprint adds setup again, plus freight, plus the cost of having your opening night look amateur.
What it cost
Let me put numbers on it, because numbers are what finally changed our process.
- 1,200 custom printed poker cards, reprinted on proper playing-card stock: $1,650
- 5,000 custom MTG tokens, reprinted at true 2.5 x 3.5 inches: $900
- 300 poker cut cards, reprinted in plastic with a contrasting edge: $380
- Expedited freight: $420
- Lost opening-night revenue and staff overtime: about $450
Total: roughly $3,800. That is not a catastrophic number for a business. But it is a stupid number. It was all avoidable with a $60 physical proof and a 20-minute checklist.
The checklist I created after the third rejection
After the third rejection in Q1 2024, I created our custom card pre-check list. We have caught 47 potential errors with it in the past 18 months. Here is the short version.
1. Match the card to the game, not the printer default
Custom MTG tokens need exact Magic dimensions and corner radius. Poker cards need poker or bridge size, and you should know which one your players expect. Cut cards are not playing cards. They need to be thicker, often plastic, and visually different.
2. Order a physical proof before the full run
A PDF proof is not enough. Ask for a printed sample on the actual stock. Shuffle it. Sleeve it. Deal it. Leave it on the poker table for an hour. If the printer cannot provide a physical proof, that is a signal.
3. Check resolution at final size
Industry standard for commercial offset printing is 300 DPI at final size. A 3000 x 2000 pixel image at 300 DPI gives you a maximum width of 10 inches. For a 2.5-inch token, your file needs to be built for that final size, not just look sharp on a monitor.
4. Lock color with Pantone or accept drift
If you have a brand color, use Pantone. But know that Pantone colors may not have exact CMYK equivalents. Pantone 286 C converts to roughly C:100 M:66 Y:0 K:2, and the printed result can still shift by substrate and press calibration. For brand-critical colors, tolerance should be Delta E < 2. Above Delta E 4, most people will notice.
5. Verify paper weight and finish
Paper weight is not a vibe. It is a spec. A 270 gsm cover stock is not the same as a playing-card stock. If you are printing your own deck of cards for a prototype, lighter stock may be fine. For commercial play, confirm the stock and coating. Glossy looks nice in photos. Matte handles fingerprints better. Linen finish changes shuffle feel.
6. Plan the poker table set together with the cards
Your poker table set, cut cards, chip trays, and card size should be chosen as one system. A cut card that matches the deck is a bug. A token that does not fit a sleeve is a bug. A table with narrow chip trays and oversized chips is a bug. Buy the system, not the pieces.
What I learned about print your own deck of cards
Print your own deck of cards sounds simple. For a prototype, it is. For a commercial venue, it is a minefield of specs. The industry has evolved. Five years ago, a local print shop could run 500 decks and everyone accepted the variation. Today, B2B buyers expect online proofing, Pantone matching, and consistent reorders. That is good. But the fundamentals have not changed. You still need a physical sample. You still need to test the card in the context where it will be used.
At ice-games, we now treat custom printed poker cards, custom MTG tokens, poker cut cards, and poker table sets as one category: playable print. If it touches a player's hands, it needs a play test. Not a screen test.
So if you are about to approve a custom card order, stop. Ask for the physical proof. Check the corner radius. Sleeve the token. Shuffle the deck. Then sign off. That is it. Simple. Not easy. But simple.
