In casino marketing, Average Daily Theoretical (ADT) is the standard metric for player value. It seems simple enough: Take a player’s theoretical win (theo) and divide it by the number of rated days played. The higher the ADT, the more valuable the player.
Right?
Wrong.
ADT is one of the most misleading metrics in casino marketing. If your casino still relies on it, you’re likely overlooking hidden VIPs, misjudging player value (both overvaluing and undervaluing), and creating unnecessary volatility in your offers. This can leave players confused about why their rewards fluctuate.
The ADT trap: Low trips, high volatility
ADT assumes every recorded play session accurately reflects a player’s worth. In reality, ADT is highly volatile, stabilizing only after multiple rated days. The fewer trips a player has, the more inaccurate and misleading their ADT becomes.
So, how many players does this affect?
That’s the alarming part, or exciting, depending on how much you want to grow your database. Typically, 60–80% of a casino player database has fewer than three rated days in a given marketing period. That means the majority of your database is being evaluated with a wildly unreliable metric.
No matter the casino—big or small, local or regional—the pattern remains the same. When analyzing full database draws, roughly two-thirds of players have three or fewer trips.
Why ADT fails: The house wins twice
On paper, theoretical win (theo) seems straightforward:
Total Bet × House Advantage
Or broken down further:
(Average Bet × Spins per Minute × Minutes Played) × House Advantage
But here’s the problem—who controls “minutes played”? Not the player. The random number generator (RNG) of the slot machine does.
We’ve analyzed this a lot, and most slot machines are mean—quick losses are the most common experience. If a player walks in, gets steamrolled by the machine in minutes, and leaves, their theo will be low. That means:
- The player has a terrible gaming experience.
- They get a terrible offer in the mail.
It’s a double whammy: ADT is often more a reflection of short-term RNG variance than actual player value.
That’s why ADT only stabilizes with multiple trips and more data. When we use ADT for marketing decisions, we’re not just misjudging players—we’re marketing to our own randomness.
So, when we send out offers based on ADT, are we rewarding true player value or just the luck (or lack thereof) of the last session?
If we’re serious about retaining VIPs and growing revenue, we need to stop investing in the RNG outcome—and start investing in the player.
Players: Their own worst enemies (unintentionally)
Most players are unaware of how flawed our valuation methods are, yet they often sabotage their own ratings without realizing it.
Consider this real-life example:
A player approached me at a casino, frustrated that his offers had suddenly dropped. He insisted he was a $2,000 per-trip player. Historically, his offers reflected that but then, they vanished.
I reviewed his play history and immediately saw the problem.
He was a low-frequency player, with only three trips in the last six months. On two trips, his play was consistent with a $2,000 budget. But on the third trip, he made one crucial mistake—he briefly played a Megabucks machine for just 10 minutes while waiting for his wife to return from the restroom before attending a show.
That one low-theo session became one-third of his ADT calculation, drastically pulling his value down. As a result, his offers collapsed overnight.
Did the player change? No.
We changed his rating—punishing him for simply using his player’s card.
My advice to him? Never rate at a casino unless you’re giving a full trip.
But is it really the player’s responsibility to work around our flawed valuation system?
Skewed data = skewed marketing
Another time, I applied a predictive model instead of ADT and discovered a VIP player hiding in plain sight.
This table games player had just two trips in 90 days, with total losses of $8,000. But because a pit boss had mistakenly recorded her loss over a single two-minute session, the system calculated an absurdly low theo, placing her in the lowest rewards tier.
As a result, she stopped receiving offers entirely.
Once we applied a predictive algorithm, rather than relying on ADT, her rating skyrocketed overnight—from no offers to top-tier VIP status. She was sent the casino’s highest offer, and when she returned, she played her usual $4,000 drop.
This isn’t a rare occurrence. It happens all the time.
“But we use ADW!”: Better, but still flawed
Recognizing ADT’s flaws, some casinos have adopted Average Daily Worth (ADW)—a metric that takes the better of ADT or a percentage of the player’s actual loss, divided by total rated days.
On the surface, ADW seems to be a major improvement. It can help correct cases like the table games VIP we just mentioned.
But ADW has problems, too:
- We only calculate total end-of-day losses.
- ADW sums up a player’s final loss over a trip. But this is volatile—a player may expose a significant portion of their wallet early but recover later, creating a misleadingly low loss figure.
- A better approach would examine the depth of loss exposure—how much a player was truly at risk at any given point. If they get lucky and walk away with a profit, that doesn’t mean they aren’t a high-value player.
- The “loss percentage” varies widely.
- Many casinos use a 40% loss assumption, but this doesn’t work for all game types.
- For example, table games typically have drop holds of 15–20%, meaning ADW may significantly overvalue these players while still undervaluing slot players.
- The loss percentage isn’t universal—and misapplying it creates more distortions.
Bottom line: ADW is better than ADT but still far from perfect.
Predicting value, not punishing players
So, what’s the fix?
Stop relying on averages. Start using predictive modeling.
Both ADT and ADW fail because casino play is not normally distributed. It’s heavily skewed—players have good days and bad days, and bad days are more frequent. These non-normal distributions wreak havoc on player valuation.
Instead of blindly using ADT or ADW, we should analyze player-controlled behaviors such as:
- Game selection: What types of games do they gravitate toward?
- Average bet size: Are they consistently wagering at a high level?
- Speed of play: Are they casual players or rapid bettors?
By comparing low-frequency players to known high-frequency VIPs, we can accurately predict their true value and adjust marketing strategies accordingly.
The industry must change—or keep losing big players
Casinos operate under the Pareto principle—80% of revenue comes from 20% of players. That means missing even a few high-value players could cost millions over time.
Every casino still clinging to ADT (or even ADW) is leaving money on the table—and actively offending players who should be nurtured.
The industry must adapt, evolve, and embrace predictive modeling—because the next VIP is already in your database. You’re just not seeing them.
I’ve asked players why they continue visiting casinos despite these flaws in player valuation, and their response shocked me:
“It’s the same everywhere.”
But what if you were the first casino to get it right?
Conclusion
Relying on outdated, misleading metrics like ADT is a mistake that costs casinos significant revenue and loyal customers.
To win back lost VIPs, casinos must shift to predictive analytics that recognize true player value—before they take their business elsewhere.
Are you ready to see who your real VIPs are?
This article was contributed by Michael Minniear, Data Scientist at Power Promotions.
