Research · Data
What 78 Verdicts Look Like When You Count Them
We counted our own table. Not to prove that it works, but to show its shape, its biases and its limits — and to make the whole thing inspectable.
Reviewed 14 November 2026 · How this page is checked
Most yes/no tarot pages give you a table and no account of it. Where did the verdicts come from? How many are positive? Is the deck balanced, or does it lean yes? We publish the table as CSV and JSON, so the honest next step is to count it. This page is the result.
The data describes this site's editorial system, not the tarot tradition. There is no historical consensus on yes/no assignments, and a different table will produce different counts. That is the point of publishing the numbers: you can compare systems instead of arguing about vibes.
The headline numbers
Verdicts across the 78-card deck
The deck leans yes: 43 of 78 cards (55.1%) are positive, against 12 no (15.4%) and 23 maybe (29.5%). That is not a neutral mirror of life. It reflects a choice: most cards describe situations that can be worked with, and only the sharpest ones are a flat no. A table with more no cards would feel harsher; a table with fewer would feel like a fortune cookie.
Major arcana are less decisive
Split the deck by arcana and the pattern changes. The 22 majors contain 9 yes, 4 no and 9 maybe — a three-way split with 40.9% maybe. The 56 minors contain 34 yes, 8 no and 14 maybe, with 60.7% yes. The majors are big stories: Death, Judgement, the Wheel, the Hanged Man. It is hard to call them simply yes or no without knowing the question, so the table marks more of them as conditional.
This is a design choice, and it has a cost. If you ask a yes/no question and draw a major, you are more likely to get a maybe than if you draw a minor. That is worth knowing before you treat a maybe as evasion. The card may be telling you that the question is bigger than the binary you brought.
Suit by suit
| Suit | Yes | No | Maybe | Yes share |
|---|---|---|---|---|
| Wands | 9 | 0 | 5 | |
| Cups | 10 | 2 | 2 | |
| Swords | 5 | 5 | 4 | |
| Pentacles | 10 | 1 | 3 |
Wands is the most optimistic suit: nine yes, five maybe, no no cards. Cups is close behind at ten yes, two no and two maybe. Pentacles is ten yes, one no and three maybe. Swords is the outlier: five yes, five no and four maybe. The suit of conflict is the only one where no is as common as yes.
That is a useful check on the table's internal logic. If Swords had come out as positive as Cups, the verdicts would be ignoring the suit's imagery. The numbers do not prove the table is right, but they show it is not random.
Strength is not accuracy
Every card also has a strength from 1 to 3. Strength 3 means the card answers directly; strength 2 means the answer is clear but conditioned; strength 1 means the card is soft or descriptive. Across the deck, 26 cards are strength 3, 39 are strength 2 and 13 are strength 1.
| Strength | Cards | Yes | No | Maybe |
|---|---|---|---|---|
| 3 · direct | 26 | 21 | 5 | 0 |
| 2 · clear | 39 | 19 | 7 | 13 |
| 1 · soft | 13 | 3 | 0 | 10 |
Two patterns stand out. First, no strength-3 card is a maybe: the strongest answers are all yes or no. Second, every strength-1 card is either yes or maybe — there is no soft no. The average strength tells the same story: yes and no both average 2.42, while maybe averages 1.57. Conditional answers are deliberately weaker, because they depend on a condition the querent has to resolve.
None of this is a probability. A strength-3 no does not mean a 90% chance of no; it means the card's answer is unambiguous within this system. Confusing strength with accuracy is the most common mistake people make with a table like this.
Where our table differs
Comparison is the test of an editorial system. The iFate comparison documents four cards where their public table and ours disagree: the Chariot (they say no, we say yes), the Hanged Man (yes / maybe), the Hierophant (yes / maybe) and the High Priestess (no / maybe). None of these is a scandal. They are choices about how much agency a card implies.
The disagreements are also a warning against treating any table as ancient authority. If two modern sites can differ on four majors, the yes/no layer is clearly a modern editorial convention. Ours is published with conditions and a method so it can be argued with rather than mystified.
What the data cannot tell you
- It cannot tell you whether a reading will be accurate. It counts verdicts, not outcomes.
- It cannot tell you which suit matters for your question. That is a reading skill, not a statistic.
- It cannot settle the historical question. The table is modern, and we say so.
- It cannot make a small sample large. The suit counts are 14 cards each; differences of one or two cards are noise.
What it can do is make the system inspectable. You can download the CSV, filter it, compare it with another table, or check whether a claim about “the tarot” is really a claim about one publisher's choices. That is a modest but real form of information gain.
Use the data
The table is published under CC BY 4.0. The CSV has one row per card with the verdict, strength, condition, timing and URL. The JSON adds the version, review date and counts. If you cite it, link to the verdict table and note the review date.
For the rules behind the numbers, see the methodology. For the historical context, see the history page. For why a reading can feel accurate regardless of the table, see the psychology page.
Method
Counts were generated from the same data that builds /verdicts.csv,
/verdicts.json and the 78 card pages. No card was reclassified for this page. The
percentages are rounded to one decimal place. The review date is shown at the top of the page;
the table has not changed its verdicts since launch.
How many cards are yes, no and maybe?
Of the 78 cards in this table, 43 are yes, 12 are no and 23 are maybe. That is 55.1% yes, 15.4% no and 29.5% maybe. The counts are fixed by the editorial table, not sampled from readings, so they describe the system rather than any user outcome.
Which suit is the most negative?
Swords. Five of its 14 cards are no, five are yes and four are maybe — the only suit where no matches yes. Wands has no no cards at all, which is a good example of how a suit theme (action, momentum) shapes a table. The sample is small, so treat the pattern as a description of this table, not a law of tarot.
Are major arcana cards more positive?
No. The majors are more balanced and more ambiguous: 9 yes, 4 no and 9 maybe out of 22. The minors lean yes more strongly, largely because the suit cards describe ordinary situations that can be acted on. The majors describe chapters, which is why nearly half of them are conditional.
Does strength predict accuracy?
No. Strength describes how directly a card answers the question in this system: 3 is direct, 2 is clear but conditioned, 1 is soft. It is not a probability and it is not a confidence score. A strength-3 card is not more likely to be right about your future; it is simply less ambiguous as an answer.
Put it to the deck
Ask your question and draw a card
One clear question, one card, one answer. The draw is free and takes about four seconds.