What nine dates produce.

You ran nine dates over twelve weeks. Three calibration, six selection, with one no-show that didn’t count. So the data we have is eight dates of usable signal. You filled four of nine slots. The five you skipped tell us something specific, more on that below.

The calibration phase set the bar at 6.0.

Your first three dates averaged 6.0 on Compatibility. That number is your line. Anyone who clears it on the markers that actually predict the long arc, commitment, appreciation, curiosity, healthy conflict, is worth taking seriously. Anyone who doesn’t, isn’t, no matter how high the spark.

The selection phase showed the pattern, twice over.

Of your six selection dates, three were high-Spark, low-Compatibility, and they were the ones who matched your type the closest (8.4/10 type-match on average). Two were lower-Spark but cleared the calibration bar easily. One was meh on both axes and we just don’t learn much from it. The pattern your audit named, the pull toward people who feel electric in the first hour and disappointing by week three, showed up clean in the data.

The pattern is doing a job.

Gottman’s research on relationship stability is consistent on this: the things that predict whether a relationship lasts have almost nothing to do with chemistry, and a lot to do with appreciation, curiosity, and how disagreements get handled. Attachment theory adds the frame: what feels electric is often anxiety wearing a tuxedo, the activation a familiar shape brings when someone is just inconsistent enough that you have to chase a little. The pattern probably worked once, when the alternative was something worse. It’s outlived its usefulness.

Two people cleared the bar. Both were the same person.

Date 5 and Date 9 were the same human, eight weeks apart. The first time you noticed the markers but flinched at the absence of spark. The second time, after watching the pattern repeat with two more high-spark candidates, the spark had caught up. Date 9 was the highest combined Spark and Compatibility score in the experiment. That isn’t a coincidence. The audit did its work. You learned to read your own data.

What to change going forward.

Three things. First, take Date 5/9 seriously. Verification phase exists for this. Stedi keeps tracking the four markers across the next stretch of dates with this person, watching whether commitment, appreciation, curiosity, and conflict trend up or down. Second, the slots you skipped, “different decade energy,” “different pace of love language,” “from outside the apps,” were the ones that would have stretched you most. If verification doesn’t hold, those are where you go next. Third, the pull toward Date 1 / Date 4 / Date 8 isn’t your real preference. The data says so. Trust it more than the spark next time.

The bottom line.

You ran the experiment. The data answered. Date 9 is the lead. You have a benchmark, calibration 6.0, that anyone you date next has to clear — whether that’s Jordan, or someone Stedi hasn’t met yet. You have a pattern named in your own words and shown in your own scores. The next decision — verify with Jordan, keep dating new people against the bar, or run a second experiment — is yours. The framework just gave you the floor to stand on, and a benchmark you keep using.

Generated from your Type Profile, your nine date cards, and decades of research from Gottman, Bowlby, Mikulincer & Shaver, and the optimal-stopping math behind the 37% rule. Informed inference applied to your specific data. Working hypotheses, not certainties.