Why We Built the CAC Score: A California-First Study of Online Casino Players

Why National Casino Ratings Fail California Players

If you live in California and you have ever tried to choose an online casino by reading a national review, you already know the quiet frustration that started this project. The headline ratings you find on most large review sites are built for a reader who does not exist: an average American gambler, stitched together from players in dozens of states with wildly different laws, banking habits, tax treatment, and expectations. That average reader is convenient for a publisher because it lets one score speak to everyone. It is useless to you, because the things that decide whether a casino is good for a Californian are exactly the things a national average sands away.

We built the CAC Score to fix that. The CAC Score is a rating read entirely through a California lens. Only California residents aged 21 and over count toward it, and the weights that drive the score reflect what California players told us they actually prioritise, not what an editor in another state assumes they should care about. When we say a casino earns a 98 or a 75, we mean that number describes the experience of a California player, measured against the concerns California players raised, verified against our own hands-on testing from inside the state. That is a narrower promise than a national rating makes, and a far more honest one.

The gap between a national rating and a California-specific one is not cosmetic. Consider three concrete examples. First, banking. A national review might celebrate a casino for accepting a long list of fiat payment methods, yet for many California players the methods that matter most are the ones that clear quickly and without friction, which in practice means cryptocurrency rails. A casino that pays a crypto withdrawal in under two hours but takes three days on a fiat wire deserves a very different score depending on which method a Californian is likely to use. Our data shows California players lean heavily toward fast-settling methods, so we weight payout speed and banking accordingly. Second, acceptance. Some offshore casinos quietly geo-restrict or throttle California players, or they accept the signup and then create friction at the cashier. A national score rarely captures this, because most of its respondents never tested the California pathway. We did. Third, device. California players skew strongly toward mobile play, which means a clumsy app or a site that breaks on a phone is not a minor demerit, it is a daily obstacle. A desktop-first reviewer will under-weight that. We do not.

None of this means national reviews are dishonest. It means they are answering a different question. The question they answer is, on balance, across the whole country, is this a reputable casino. The question we answer is narrower and, for you, more useful: for a California player, in spring 2026, weighing the things California players say matter, how good is this casino on a scale of 0 to 100. Those are not the same question, and pretending one score can serve both is the failure we set out to correct.

There is also a subtler failure worth naming, because it is the one that does the most damage. National ratings tend to be built on the experience of whichever players happen to dominate a publisher's audience, and that audience is usually concentrated in states where the regulatory and banking picture looks nothing like California's. When the resulting score is then presented to a California reader as if it were universal, the reader has no way to know which parts of it apply to them and which parts were calibrated to someone else's situation entirely. The score looks authoritative precisely because it hides its own assumptions. We think that is the opposite of useful. A rating should make its assumptions visible, and the most important assumption any casino rating can make is about who the player is. By fixing that assumption explicitly, a California resident aged 21 or over, we make every other number in the score interpretable. You always know exactly whose experience the 98 or the 75 is describing, because we told you at the outset.

We should also be candid that a California-lens score is harder to produce than a national one, which is part of why so few exist. To weight payout speed correctly for California players, we had to learn how California players actually move money, which meant asking them and then testing it ourselves rather than copying a generic banking checklist. To weight mobile usability correctly, we had to confirm just how mobile-first the California audience is, then test the products on the devices they actually use. To weight bonus value correctly, we had to measure not the headline offer but how clearly its real cost was disclosed to a California player at the moment of opting in. Each of those steps adds work, and each of them is work a national average lets a publisher skip. We did not want to skip it, because skipping it is exactly how you end up with a score that sounds confident and means very little to the person reading it.

This report is the first of four. In this part we explain why a California-lens score exists, where the project came from, and the research foundations it rests on. We review the academic literature on trust, satisfaction, service quality, and decision-making under risk that motivated our design. We map each of those theories onto a specific pillar of the score. We define exactly who we studied, state our research questions and hypotheses, and introduce the eight weighted pillars and the simple arithmetic that combines them into a single number. The later parts go deeper into methodology, results, and the casino-by-casino breakdown. Our goal throughout is not to impress you with jargon but to show our work, so that when you trust a CAC Score you know precisely what stands behind it.

How the Project Began: From a Palo Alto Panel to a Statewide Study

The work that became the CAC Score did not start as a rating system. It started as a question among a small group of players in Palo Alto who were tired of guessing. We began with a modest local panel: a handful of California players, all aged 21 or over, who agreed to log their real experiences with online casinos and to compare notes honestly. We asked them simple things. How long did your withdrawal actually take, not how long the casino promised. Did the bonus terms match what you expected when you opted in. When something went wrong, did support help you or stall you. The answers were inconsistent in a way that told us the published ratings were not capturing what mattered.

That small Palo Alto panel is the root of everything that followed, and it is why our organization carries the name it does. As the panel grew, we realised two things at once. First, the patterns we were seeing were not unique to the Bay Area. Players in Southern California, the Central Valley, Sacramento, the Central Coast, and the far North State were describing similar frustrations, but with regional differences worth measuring. Second, a panel of friends comparing notes, however honest, is not a study. To say anything defensible about California players as a group, we needed a real sample, real instruments, and real statistics.

So we expanded. We took the Palo-Alto-rooted approach and built it out into field sites across the state, recruiting California residents 21 and over from every major region, then bringing their responses back into a single, statistically weighted picture. That California-wide, Palo-Alto-originated effort is the reason the organization is named CA Casinos, Palo Alto Organization, and it is the reason the rating is called the CAC Score. The C and the A stand for California; the C stands for casinos; the Palo Alto in our name marks where the work began. We kept the name because we never want to lose sight of the fact that this started with a few real California players asking honest questions, and the discipline we added later, the sampling, the surveys, the statistics, exists to make those honest questions answerable at scale.

There is a practical consequence of that origin worth stating plainly. Because the score grew out of California players' lived experience and was then formalised into a statewide study, every design choice is anchored to California. We did not import a generic rating template and relabel it. We started with what California players said, measured it, and only then built the weighting around it. When you read later that the California Player Survey carries the single largest weight in the score, that is not a marketing flourish. It is a direct expression of where the project came from and whom it is for.

Two Ratings, One California Lens

On our reviews you will see two numbers, and it is worth clarifying the relationship between them up front so they are never confused. The first is the CAC Score out of 100. This is the California-specific, data-driven rating produced by the study described in this series. It is the number this entire report is about. The second is a star rating out of 5, which is our broader editorial verdict on a casino as a US-facing operator. The two are related but not identical. As a rough guide, the stars approximate the CAC Score divided by 20, so a casino near 90 on the CAC Score sits near a 4.5-star editorial verdict. When the two ever diverge, the CAC Score is the one to trust for a California-specific decision, because it is the one built and weighted for California players. The star rating is the wider-angle view; the CAC Score is the California close-up.

What Prior Research Tells Us: A Literature Review

A score is only as good as the thinking behind its design. Before we wrote a single survey question, we grounded the work in the established research on how people form trust, judge satisfaction, evaluate service quality, decide under risk, and recommend products to others. We are not the first to study any of these things. Decades of peer-reviewed work give us validated ways to think about each one, and we leaned on that work deliberately so that our pillars would rest on theory rather than on our own untested intuitions. This section walks through the research that shaped the CAC Score and explains, in each case, why it mattered for an online casino audience in California.

Trust as the Foundation

Nothing else in online gambling matters if a player does not trust the operator to hold and return their money. That makes trust the natural starting point, and the most influential model of how trust forms is the integrative model proposed by Mayer, Davis, and Schoorman (1995). Their work frames trust as a willingness to be vulnerable to another party based on the expectation that the other party will act in a certain way, and it identifies ability, benevolence, and integrity as the core attributes a trustor evaluates. For a California casino player, those three attributes translate cleanly. Ability is whether the casino is competent and properly licensed to run real-money games and process payouts. Benevolence is whether the casino appears to act in the player's interest rather than purely extracting value from them. Integrity is whether the casino's stated rules and its actual behaviour line up, especially at the cashier. We built our Trust and Licensing pillar directly on this model, because Mayer and colleagues give us a structured way to ask whether a player believes the operator is able, fair-minded, and honest, rather than asking a vague single question about whether a casino feels trustworthy.

Satisfaction and the Confirmation of Expectations

Trust gets a player in the door. Satisfaction decides whether they stay and whether they recommend the casino to anyone else. The dominant theory here is Oliver's Expectation-Confirmation Theory (1980), which holds that satisfaction is not an absolute judgement but a comparison: people form expectations before an experience, then compare the actual experience against those expectations. When the experience meets or exceeds expectations, satisfaction rises; when it falls short, dissatisfaction follows, even if the objective experience was acceptable. This is profoundly relevant to casinos, where so much disappointment comes not from a service being bad in absolute terms but from it being worse than what was promised. A withdrawal that takes three days is fine if the casino said three days, and infuriating if the casino implied same-day. Oliver's framework is why we ask players not just what happened but how it compared to what they were told to expect, and it is why our overall satisfaction and would-recommend measures sit on top of the pillar structure rather than inside any single pillar.

Service Quality and the SERVQUAL Tradition

When a player needs help, the quality of the service they receive is its own dimension of the experience. The most validated approach to measuring perceived service quality is SERVQUAL, introduced by Parasuraman, Zeithaml, and Berry (1988). SERVQUAL conceives of service quality as the gap between what customers expect and what they perceive they received, measured across dimensions such as reliability, responsiveness, assurance, empathy, and tangibles. For an online casino, the most pointed of these is responsiveness: when a player contacts live chat or email with a problem, how quickly and how competently is it resolved. We built our Customer Support pillar on the SERVQUAL tradition, focusing on the responsiveness and assurance dimensions that players raise most often. SERVQUAL also reinforces a lesson that runs through this entire study, which is that quality is perceived relative to expectation, echoing Oliver, so we measure support not in a vacuum but against what players reasonably expected the casino to deliver.

Technology Acceptance and the Mobile Experience

California players, as our data will show, play overwhelmingly on phones. That makes the usability of the app or mobile site a first-order concern, not an afterthought, and the literature on how people accept and adopt technology gives us the right lens. Davis's Technology Acceptance Model (1989) argues that two perceptions drive whether people use a technology: perceived usefulness and perceived ease of use. Applied to a casino app, perceived usefulness is whether the app lets a player do what they came to do, deposit, play, and withdraw, without obstruction, and perceived ease of use is whether the interface is intuitive enough that they can do it without friction. We built our Mobile and Responsible Gambling pillar on TAM, because it gives us a validated way to ask whether the mobile product is both useful and easy, rather than relying on a subjective gut feeling about whether an app is nice. We fold responsible-gambling tools into this pillar deliberately, because for a mobile-first audience those tools are only effective if they are easy to find and use, which is a direct ease-of-use question in TAM terms.

Decision-Making Under Risk: Prospect Theory and Heuristics

Bonuses are where psychology and arithmetic collide. A welcome offer is, in the end, a gamble wrapped in marketing, and players evaluate it under uncertainty. The foundational work here is Kahneman and Tversky's Prospect Theory (1979), which showed that people do not evaluate outcomes rationally against expected value; instead they weigh gains and losses relative to a reference point, overweight small probabilities, and feel losses more sharply than equivalent gains. Their earlier work on judgment under uncertainty (Tversky and Kahneman, 1974) catalogued the heuristics and biases that distort these judgements. For casino bonuses, this is everything. A headline bonus amount feels like a gain and is weighted heavily; the wagering requirement attached to it is a probabilistic cost that players systematically underweight, especially when it is buried or framed obscurely. We built our Bonuses and Value pillar on this literature, focusing not just on the size of an offer but on how clearly its real terms, above all the wagering requirement, are disclosed before a player opts in. The theory predicts that opaque framing will mislead, so we measure clarity directly.

Recommendation and the Net Promoter Logic

One of the most practical questions you can ask any customer is whether they would recommend a product to someone like them. Reichheld (2003) argued that a single question, how likely are you to recommend this to a friend or colleague, is among the strongest available predictors of loyalty and growth, and formalised it as the Net Promoter Score. We adapted this directly, asking each respondent how likely they are to recommend a casino to another California player on a 0 to 10 scale, then computing a net-recommend figure. This gives us a clean, validated summary measure that complements the pillar structure: the pillars tell us where a casino does well and poorly, and the net-recommend figure tells us the bottom-line verdict of players who actually used it.

Why Players Sometimes Defend Bad Choices

Two further pieces of theory helped us interpret what players told us, even though they do not map onto a single pillar. Festinger's theory of cognitive dissonance (1957) explains why a player who has sunk time and money into a casino may rationalise a poor experience rather than admit a mistake, which is a real source of bias in self-reported satisfaction. We kept this in mind when interpreting unusually positive reports from heavily invested players, and it is part of why we cross-check self-reported payout times against our own hands-on testing rather than taking them at face value.

Responsible Gambling as a Health Concern

Finally, because we are rating a product that can cause harm, we grounded the responsible-gambling component of our work in the public-health literature rather than treating it as a compliance checkbox. Griffiths (2005) offers a components model of addiction within a biopsychosocial framework, identifying features such as salience, mood modification, tolerance, withdrawal, conflict, and relapse. This informed how we think about responsible-gambling tools: the point of deposit limits, self-exclusion, reality checks, and easy access to support is to interrupt exactly the behavioural patterns Griffiths describes. A casino that makes those tools easy to find scores better on our Mobile and Responsible Gambling pillar, and the literature is why we treat that as a genuine quality dimension rather than a courtesy.

The Research That Shaped Our Methods

Beyond the theories that shaped the pillars, a second body of work shaped how we collected and analysed data. Likert (1932) gave us the five-point agreement scale that forms the quantitative core of our instrument, a format players understand intuitively and that produces data we can analyse rigorously. Goodman (1961) is the foundational treatment of snowball sampling, and it is precisely because Goodman is clear that snowball sampling is a non-probability method that we use it only as a supplement to reach hard-to-find players, never as the backbone. Cochran (1977) is the standard reference for the sampling techniques, including stratification and finite population correction, that underpin our probability design. Dillman, Smyth, and Christian (2014) provided the tailored-design principles for building mixed-mode surveys that minimise error and non-response. Cronbach (1951) gave us coefficient alpha, the statistic we use to confirm that the multi-item batteries within each pillar actually hang together and measure a single underlying construct. Together these works are why we can claim our numbers are defensible rather than merely plausible.

Theoretical Frameworks Mapped to the CAC Pillars

The literature above is not decoration. Each framework we relied on corresponds to a specific, named component of the CAC Score, and we think the cleanest way to show that is to lay the mapping out directly. The table below connects each established framework to the CAC pillar or measure it informs. We want a reader to be able to trace any pillar back to the research tradition that justifies it, because that traceability is the difference between a score grounded in theory and a score invented to look authoritative.

Reading down that table is, in effect, reading the blueprint of the score. Trust sits at the top because trust is the precondition for everything else. Satisfaction and recommendation sit across the whole structure because they are downstream summary judgements rather than single attributes. Support, mobile usability, and bonus value each draw on the specific tradition that studies them most rigorously. And the entire apparatus of how we sampled and measured is anchored to the methods literature, so the design choices are not arbitrary. When we introduce the eight weighted pillars shortly, you will be able to see each one resting on a row of this table.

Who We Studied: Defining the California Audience

A study is only meaningful if you know exactly who it describes. Our population is precisely defined: California residents aged 21 or over who play, or intend to play, online casino games. Every word of that definition is load-bearing. California residency is what makes the score a California-lens rating; we did not include players from other states, because their experience would dilute the very specificity we are trying to deliver. The age floor of 21 reflects the population we are willing to study and serve. And the inclusion of those who intend to play, alongside current players, lets us capture the concerns of newcomers, including what almost stopped them from signing up, which turns out to be some of the most useful information in the whole study.

From that population we drew a sample of N = 4,217 verified respondents. That is roughly a one percent sampling fraction of the estimated active California online-casino audience, which is a substantial fraction for a study of this kind and is what allows us to report tight margins of error. We will detail the sampling design and its mathematics in Part 2, but the headline is that the sample was built on a stratified random backbone, supplemented by snowball referral for hard-to-reach segments, and then re-weighted so that its regional and demographic composition matches California. The overall margin of error on the probability portion is plus or minus 1.51 percent at the 95 percent confidence level. We also gathered qualitative depth: open-ended survey responses from across the sample plus sixty follow-up interviews, coded thematically. Throughout, participation was voluntary and consented, residency and age were verified before inclusion, and no personally identifying information is published. We report aggregates only.

Research Questions and Hypotheses

We did not go fishing in the data. Before fieldwork, we set out the questions we wanted to answer and the hypotheses we expected to test, so that our analysis would be disciplined rather than opportunistic. The table below states them plainly. Each research question is paired with the hypothesis we carried into the study, and you will see in Parts 2 and 3 how the data bore on each one.

These six questions structure the entire series. RQ1 is the question that justifies the weighting itself: if California players did not prioritise trust and payouts, our weights would be wrong. RQ2 and RQ3 are the inferential heart of the payout analysis, and we will report the t-test and ANOVA results that bear on them in Part 3. RQ4 connects the pillars to the bottom-line net-recommend logic from Reichheld. RQ5 puts Prospect Theory to the test against real bonus sentiment. And RQ6 is the reliability check, without which we would not be entitled to treat each cluster of survey items as a coherent pillar at all. Stating these in advance is part of how we keep ourselves honest.

The Eight Weighted Pillars

We can now introduce the structure the whole score is built on. The CAC Score out of 100 is a weighted sum of eight pillars. Each pillar captures one dimension of the casino experience, grounded in the frameworks above, and each carries a weight that reflects how much California players told us it matters. The weights sum to 100, so the score is directly interpretable as points out of a possible 100. The table below lists the pillars and their weights, ordered from heaviest to lightest, and the donut chart that follows shows the same information visually.

The weights tell a story about what California players value, and we want to interpret them rather than leave them as a list. The single heaviest weight, 20 percent, goes to the California Player Survey itself. That is deliberate and it is the clearest expression of our origin: the lived experience of real California players, gathered through the survey, is the most important input, more important than any single attribute we measure on their behalf. Trust and Licensing follows at 18 percent, consistent with the literature that places trust at the foundation of the relationship, and consistent with what players told us when asked what they care about most. Payout Speed and Banking takes 15 percent, reflecting how central the cashier experience is to a California player's judgement, and Bonuses and Value takes 14 percent, capturing the offer side of the equation where Prospect Theory predicts so much confusion lives. Game Selection at 13 percent matters but sits below the trust and money factors, which is exactly the ordering our first hypothesis predicted. Security and Fairness, Customer Support, and Mobile and Responsible Gambling carry 8, 7, and 5 percent respectively. Their lower weights are not a statement that they are unimportant; they reflect that, in the survey, players treated them as expected baselines that mattered most when they failed rather than as the primary basis for choosing a casino.

How the Pillars Combine: The Weighted CAC Formula

The arithmetic that turns eight pillar scores into a single CAC Score is intentionally simple, because a score people are asked to trust should be transparent. We score each pillar on a common scale, multiply each pillar score by its weight, and add them together. Because the weights sum to one, the result lands cleanly on a 0 to 100 scale. The formula below states this exactly.

In words, the CAC Score is the sum, across all eight pillars, of each pillar's weight multiplied by that pillar's score, with the weights constrained to add up to one. There is no hidden adjustment, no secret editorial thumb on the scale. If you knew a casino's eight pillar scores and the published weights, you could reproduce its CAC Score yourself. That reproducibility is the point. A rating that cannot be reconstructed from its inputs is asking for trust it has not earned, and after building a model of trust on Mayer's notion of integrity, we were not willing to fail that test ourselves. The weights you see are the weights we use, and the formula above is the whole of the combination logic.

What the Scores Look Like in Practice

It is one thing to describe a method and another to see what it produces. The chart below shows the final CAC Scores for the fifteen casinos we reviewed in this round, ordered from highest to lowest. We include it here, at the end of Part 1, so that the abstract apparatus we have just described, the population, the pillars, the weights, the formula, lands on something concrete before we dive into methodology and results in the following parts.

A few things in this picture are worth noticing now, even though the full explanation waits for Part 3. The top of the table clusters tightly between 95 and 98, with Ignition at 98, BetOnline at 97, All Star Slots at 96, and Super Slots at 95. These are the casinos that pair fast crypto payouts with strong trust signals and clean mobile products, which is precisely the combination our weighting rewards. There is then a visible step down into the upper 80s and low 80s, where casinos remain solid, our Very Good tier, but give up ground on payout speed, support responsiveness, or bonus clarity. At the bottom sit Shazam, BetWhale, and VoltageBet in the mid-to-high 70s, still in our Good tier but clearly behind on the factors California players weight most heavily. We use a simple tier language throughout: 90 to 100 is Excellent, 80 to 89 is Very Good, 70 to 79 is Good, and anything below 70 is Fair or lower. As a rough cross-reference to the star rating you will also see on our reviews, a CAC Score divided by 20 approximates the stars, so a 98 sits near five stars and a 75 near three and three-quarters.

What the chart does not show, and what the rest of this series exists to demonstrate, is why these numbers are credible. Anyone can publish a bar chart. The reason you can trust this one is the work underneath it: a defined California population, a stratified random sample of 4,217 verified residents 21 and over, a survey instrument grounded in decades of validated theory, reliability confirmed by Cronbach's alpha, inferential tests behind the payout and recommendation claims, and a transparent weighting formula you could reproduce yourself. That is the difference between a ranking and a study, and it is the standard we hold ourselves to.

Where This Series Goes Next

This first part has explained why a California-lens score needs to exist, where the project came from, and the theoretical and methodological foundations it rests on. We have reviewed the research on trust, satisfaction, service quality, technology acceptance, decision-making under risk, recommendation, dissonance, and gambling-related harm, and we have shown how each strand maps onto a specific pillar of the CAC Score. We have defined our population of California players 21 and over, stated our research questions and hypotheses, introduced the eight weighted pillars, and set out the simple weighted-sum formula that produces every score. The bar chart above is the output; everything before it is the reason the output can be trusted.

The remaining three parts build directly on this foundation. Part 2 details the methodology in full: the stratified random sampling backbone, the snowball supplement and why it is treated separately, the sample-size and margin-of-error mathematics, the survey instrument, eligibility and consent, and the validity, reliability, and limitations of the design. Part 3 presents the results: the descriptive statistics for each pillar, the inferential tests behind our payout and recommendation findings, and the qualitative themes drawn from open-ended responses and depth interviews. Part 4 is the casino-by-casino walkthrough, showing how each of the fifteen casinos earned its CAC Score against our hands-on testing. To see the live methodology summary and the current scores, visit our CAC Score hub at /cac-score/, where this series sits alongside the rest of our California-first research.