All articles ←

What do you need to match on? The formula that changed the calculation

10 September 2026·7 min read·By The Zugot Project team

Two people can give exactly the same answer to a question and still not be a match. Two others can give different answers to that same question and be an excellent one. The difference isn't in the position — it's in how much that position matters to them in the first place.

The problem with comparing positions

Suppose a matching system asks you to rate your daily halachic observance from 1 to 5. You answer 4. A candidate answers 4 as well. The system sees a distance of zero and declares a perfect match on that item. Reasonable enough.

Now take a second case: you answer 4 and the candidate answers 3. A distance of one. The system deducts points. The question nobody asked is whether that distance matters to either of you. It's entirely possible that you're both flexible here, that you both know excellent couples who aren't precisely aligned on it, and that one step of difference is, for you, noise. It's equally possible that you both see this as the backbone of the home you intend to build, and that one step of difference is a chasm.

The same number — a distance of 1 — describes a negligible mismatch in the first case and a decisive one in the second. A system that looks only at positions cannot tell them apart, because the information that separates them was never collected.

Nomothetic versus idiographic

Psychology has a name for this distinction. A nomothetic approach measures everyone with one yardstick: the same questions, the same weights, the same formula for every person. An idiographic approach measures each person with a yardstick they defined themselves: the same questions, but the weight of each is set by the respondent.

Most compatibility questionnaires are nomothetic. They assume that if a question is on the form, it matters — and matters equally — to everyone filling it in. That assumption is computationally convenient and simply untrue. People differ not only in their answers, but in which answers they care about.

The formula itself

The calculation per item is a product of two factors. The first is closeness: take the distance between the two positions and subtract it from the scale maximum. Two people who answered identically get full marks; two at opposite ends get the minimum.

The second factor is weight: the average of the importance ratings both sides gave that item. If both marked it as highly important, the item enters the calculation at full force. If one is indifferent, the weight drops. If both are indifferent, the weight approaches zero.

The product of the two is the whole story. High closeness on an item neither person cares about contributes very little — correctly, because there is no reason to reward a couple for being aligned on something neither is invested in. Conversely, a gap on an item both rated critical cuts at full weight, because that is where a gap will actually decide things.

The small detail that decides it: indifference must be worth zero

There's a technical decision here that looks minor and isn't. When converting an importance rating from a 1-to-5 scale into a weight, the natural approach is to divide by the maximum: a rating of 1 becomes 0.2, a rating of 5 becomes 1. The problem is that a rating of 1 means "this doesn't matter to me at all" — and yet the item continues to enter the calculation at a fifth of its weight.

That isn't negligible. Someone who explicitly marked a subject as irrelevant still receives matches shaped by it. They asked to opt out, and the system kept them partly in. So the conversion here works differently: the scale is stretched so a rating of 1 maps to a weight of exactly zero. Someone who said "doesn't matter" genuinely has that item removed from the equation.

Where it actually runs

The personal card has three core items that each get this treatment separately, with their own position and their own importance rating: daily halachic observance, attitude towards Torah study and spirituality, and openness to engagement with the modern world.

These three were deliberately separated from the rest of the questionnaire. They get individual importance ratings rather than one blanket rating for the category, because in the religious community they are the most decisive items — and because people genuinely differ across them. Some are entirely flexible about engagement with the modern world and uncompromising about Torah study; for others the picture is reversed. A single blanket question would erase the difference.

The same formula runs on prayer rite, where the distance is categorical rather than numeric — Ashkenaz, Sefard, Edot HaMizrach, Yemenite and others. There too the second question is asked: how much do you need a partner who davens the same nusach? For someone to whom it is foundational, full weight. For someone to whom it is a technicality, zero.

What this means when you fill in your card

The practical conclusion is simple: the importance rating is not a polite follow-up question. It is half the calculation. Someone who marks 5 on every importance question is effectively demanding a perfect match on every dimension simultaneously — and since such a person barely exists, the result is a very short match list, or an empty one.

Someone who marks 1 everywhere gives up the system's ability to distinguish between candidates on their behalf. If nothing matters, everyone matches equally, which is precisely the situation we set out to escape.

The real value comes when the ratings discriminate. Two or three subjects where you genuinely have no flexibility, and honest flexibility marked on the rest — that is what lets the system find someone aligned with you on what is critical, instead of disqualifying them over a clause that would never have bothered you.

And that, in the end, is what the Lutz formula buys: not greater mathematical precision, but the recognition that two people filling in the very same questionnaire were not necessarily asked the same question.