AUTHOR’S NOTE: Lately, online pundits who are pushing a psycho-sociological agenda have been all over the topic of “bias” in human relations: confirmation bias, cognitive bias, affinity bias, attribution bias, conformity bias, gender bias, etc. It has me thinking that maybe tarot needs another label: “prediction bias.”

There is already a framework for it in two assumptions: the idea that there are cards that typically say “yes” or “no,” and others that are noncommittal, amended by the notion that there are no “bad” cards and any of them can be repurposed in a more positive way to support the concept of “empowerment.” The odd premise that “it’s all good” and there are no irredeemable negatives is itself a form of “optimism bias” that I find unconvincing.

I decided to take my long-standing list of “Yes-leaning,” “No-leaning” and “Maybe” cards (note the accidental bias-avoidance there) and outfit them with new headers that show what their bias would be if they appear as the “outcome” (aka “answer”) card in a reading. Although the card titles are from the Waite-Smith (RWS) deck, this binning is based on Aleister Crowley’s definitions for the Thoth cards. (You can ignore the elemental symbols for each card; they have no bearing on the bias determination.) Those who use the RWS tarot will have other opinions about this arrangement and, because not all questions or topics are amenable to value judgment, trust in any such list for reading purposes should be approached cautiously.

The chance of receiving a “Maybe” answer from the full 78-card population is exactly one-third (33%), while the remaining ratio is slanted toward likely satisfaction (41%) compared to an unsatisfactory outcome (26%). If we split the group of “Maybe” cards down the middle and add 13 to the “Likely” column and 13 to the “Unlikely” column, we come up with a 58% chance of receiving a positive nod that, given the fact that people are generally programmed for success by societal expectations, does not seem like an unreasonable distribution for a tarot reading

.

The disposition will obviously change when we limit the target to a random five-card or ten-card pull from the total of 78 because there are literally millions of possible permutations, but the math for that is beyond my scope here. Another way to assess it would be to look for a preponderance of any of the three categories in the spread as a whole and weigh that against the “bias” of the outcome card, drawing conclusions from their cooperative, neutral or contentious dynamic. For example, a favorable outcome card could be weakened by four unfavorable cards preceding it, to the point that there is nothing left to cheer about by the time we get to the end except crossing the finish line.