The Practical Guide to Using the Traditional Counting Rhyme
What is And Then He Kissed Her?
"And Then He Kissed Her" is a traditional cumulative rhyme used as a counting-out game. You say the words one syllable at a time, pointing at each participant in sequence, until the final syllable lands on a specific person. That person is then "chosen" or "out," depending on the context of the game. It's essentially a random-selection tool that predates dice, coins, and digital generators by centuries. The standard version goes: "And then he kissed her, and then she kissed him, and then they all went home again" or similar variants depending on your region. The exact wording doesn't matter much, but consistency does. If everyone agrees on the version before you start, the selection is as random as any other method. If people switch versions mid-count, you get arguments, and nobody wants that.
How It Actually Works in Practice
The mechanics are dead simple. Get a group of people in a circle or line. Pick one starter. Go around pointing at each person on one syllable per person, saying the full phrase repeatedly until the syllable count lands on whoever you want eliminated or selected. That's it. Here's where people get it wrong though. They rush the cadence. The whole point of a counting rhyme is that the rhythm is steady and even. If you speed up when you're close to ending on someone you want out, you've ruined the randomness. You're no longer selecting by chance, you're selecting by manipulation. Kids notice this immediately and call it out. Adults pretend not to notice but they absolutely know. Another common mistake is using a rhyme that doesn't divide evenly among the group size. If you have 7 people and your rhyme has 8 syllables per cycle, the first person in line gets statistically more exposure to being "the one" than anyone else, because the count resets at position 1 each cycle rather than scattering evenly. Try doing a quick modulo check in your head before you start: group size into syllable length. If it divides cleanly, great. If not, just expect a slight bias and move on.
Step-by-Step Walkthrough
Setting It Up
Gather your group. Stand them in a circle if there are more than five people. For smaller groups, a line works fine. Make sure everyone can see each other so the pointing isn't ambiguous. Ambiguity is where disputes come from. Pick the rhyme version beforehand. The most common one I see used is: And then he kissed her, and then she kissed him, and then they all went home again.
Get the Full Details
Count the syllables: that's roughly 14 depending on how you pronounce "kissed" and whether you slur "went home" into one beat. Decide as a group whether "went home again" counts as two or three syllables. This argument happens more often than you'd think. Write it down if you need to. Seriously, just agree once and stick with it.
Running the Count
The starter begins at any position. Point at the first person on "And," the next on "then," and so on. Continue around the circle, restarting the rhyme from the beginning once you finish a cycle. The person landed on when the syllable count naturally ends is the selected one. One count, one selection. Game over or round over depending on your setup. For a larger group where you need multiple eliminations, restart the count from wherever the previous selection left off. This preserves the statistical distribution better than going back to the top each time.
Handling Edge Cases
I ran into a specific problem once with a group of 13 at a family gathering. The rhyme we were using had 11 syllables per cycle. The bias was noticeable by the third round because certain positions were consistently getting selected. I switched to counting with a finger point on each beat instead of just verbalizing it, and had the group call out "stop" when they thought the count should end. That introduced enough human variance to make it feel fair even though mathematically it was still slightly biased. Not ideal, but it kept the peace. Another edge case: what if the count lands exactly between two people because someone was slow pointing? Just restart. Don't try to split the difference. Half the people will think you favored the other half, and that's how friendships end at holiday dinners.
When This Method Fails Completely
Counting rhymes like And Then He Kissed Her do not work when you need true randomness. They work when you need an acceptable approximation of chance among people who trust the process. If someone in the group suspects rigging, the whole thing falls apart regardless of whether rigging actually occurred. The social contract matters more than the math. They also fail when the group is too large. Beyond about 30 people, the rhythm drags, attention wanders, and people start counting along silently to game the result. Once two people are counting independently and disagree on where the syllable lands, you need a different method entirely. A coin flip or random number generator would be faster and carry zero social risk. There is also the problem of cultural exclusion. If not everyone in the group knows the rhyme, you're implicitly privileging people who grew up with it. I've seen this cause actual resentment at school playgrounds and team events. In those situations, the kindest move is to pick a different counting method or teach the rhyme to everyone first.
Alternatives Worth Considering
Draw straws. Classic, fast, universally understood. Each person draws a piece, one is marked or shorter, and the drawer is out. Takes about 30 seconds total. Rock paper scissors tournament. Eliminates people pairwise until one remains. More interactive, but takes longer with bigger groups. Digital randomizers. Phones have this built in now. Most people don't realize how many random-selection apps exist that handle group sizing, weighting, and fairness checks automatically. Use one if the stakes are high or the group is diverse.
Final Notes
The charm of And Then He Kissed Her isn't in its precision. It's in the shared rhythm, the physical participation, and the nostalgia factor. It's a social tool first and a randomization method second. If you need fair randomness, use a tool built for it. If you need something that feels like a game while deciding who goes first or who's on dish duty, this rhyme does the job fine. Just pick the version, agree on the count, and don't rush it.
