The Not-So-Simple Yes-or-No Spread


Final Note

This is much more statistically credible than it was, thanks to excellent input from Etene. The spread template attached to the original post has been updated accordingly. The upshot of the changes is that:

The incremental positive and negative adjustments are much smaller than they were for each two-card comparison in order to keep a statistically significant number of answers within the 0-to-100 range.

There is a stricter descending order of importance to the offsetting values

There is additional negative reinforcement to counter the excessive positive bias of the Elemental Dignity system (positives have also been proportionally reined in)

I've added a "Best Case/Worst Case Analysis" table that shows the bounding scenarios for maximum up-and-down adjustment, independent of the statistical fine-tuning

I've added a list of the trump card elemental correspondences

Most of the readings with this spread should stay within the 25% - 75% range with a legitimate number of definite "Yes" or "No" answers occurring beyond those limits

As the title says, it's "not so simple," but it does take the "yes-or-no" decision out of the guesswork arena and at least get it within striking distance of SWAG territory.


While waiting for Etene to perform statistical wizardry on the changes I just made, I decided to post this major overhaul of the layout. The structure of the spread hasn't changed, just the level of detail in the written guidance along with a few tweaks to the scoring. See the update attached to the original post.


Complete Ranking Revamp

I completely changed the mechanics of this spread by eliminating the numerical scoring approach and replacing it with a sliding-scale ranking system that allows some latitude in applying judgment. (In a couple of cases it demands such judgment because I wound up with seven ranking parameters and only five sets of variables).


I've completed my work on this spread and am moving into more testing with the new ranking method. I added a worksheet to help with keeping track of the ranking. I also added a caution against trying to use this spread to go after a "% Confidence" prediction for getting a "Yes" answer. As it now stands, its much more appropriate for a fluid, inductive approach that weighs the different ranks according to preponderance of positive, neutral and negative qualities and arrives at a composite verdict. I realized that this is not so much a "yes-or-no" answer spread as it is a "positioning" spread with a "yes/no" emphasis.

I updated the spread template attached to my earlier post.