Was I too anxious, or were they avoidant?

Probably a bit of both, and neither one is a verdict. What you were likely looking at was a loop between two tendencies, one person reaching and one pulling back, and loops are built by two people in a particular situation, not by one broken person.

What were you actually looking at?

Say it’s Leo, 24, and Sam, who he saw for about three months last spring. Now Leo is scrolling back through the thread at midnight. There’s his long message after the rooftop night, three paragraphs about how much fun he had. Sam’s reply: “haha yeah for sure.” Two weeks later, Leo at 1 a.m.: “are we good?” Sam, the next afternoon: “yeah just need some space this week.” Then it thinned out until there was nothing left to thin.

Leo has two theories, and he swaps them every few days. Either he was too anxious and scared Sam off, or Sam was avoidant and nothing he did would have mattered.

Split it the way you’d split anyone else’s story. What happened: different texting volumes, a late check-in, a request for space, a fade. The reading: one of them is the problem. The fear: that it’ll happen again, and that it’s him. The next step: still open, and that’s the part worth spending time on.

That pattern, one person pressing for more and the other pulling away, has a name in couples research: demand-withdraw. A meta-analysis of 74 studies with more than 14,000 participants linked it to lower satisfaction, less intimacy and poorer communication.1 Notice where the problem sits in that sentence. It’s in the pattern, not in either person.

Were they avoidant, or avoidant with you?

That’s a real distinction. Attachment researchers measure two sliding scales rather than types: anxiety (worry about rejection or being left) and avoidance (discomfort with closeness and depending on someone). And the two don’t behave the same way across relationships.

In a study of college students who rated their attachment across five different relationships, João Moreira found that avoidance varied more from relationship to relationship than from person to person. Anxiety was the opposite: it tended to come along with a person across their relationships.2

Two things follow, and both are kinder than either of Leo’s theories:

• “They were avoidant” may describe how Sam was in that relationship, at that time, with that much uncertainty in the air. Not who Sam is.
• Your worry may be something you carry, which isn’t a flaw. It’s useful to know, because it means the next relationship will probably poke at the same spot, and you can get ahead of it.

Can you tell from the thread?

Less than you’d hope. In a 2025 study of the actual message histories of 40 early-stage college daters, Rachel Reymann Vanderbilt and colleagues found that people higher in avoidance texted less often. Anxiety didn’t predict how often people texted, but more anxious daters used more future-focused language, the “next weekend” and “we should” kind of messages.3

A much smaller 2012 study points another way. Robert Weisskirch had 31 college women check their phone records, and those higher in anxiety had sent and received more texts with their partners.4 Small samples, different methods, partly different answers. That’s what research on this looks like right now.

And the thread leaves a lot out. Maybe Sam was in finals or working doubles. Maybe Sam hates texting and is great in person. Maybe Sam grew up somewhere texting is for logistics. Some people, including some neurodivergent people, just lose threads without meaning anything by it. A short reply isn’t a diagnosis, and neither is a long one.

Were you “too much”?

Wanting regular contact with someone you’re seeing isn’t too much. Wanting less isn’t cold. The trouble starts when two different needs go unnamed, and each person’s way of coping (reaching more, or stepping back) makes the other’s need louder. That’s the loop.

Mismatch itself may be the bigger story than either person’s wiring. In a survey of 205 adults aged 18 to 29 in relationships, Jonathan Ohadi, Leora Trub and colleagues found that feeling similar in texting habits, especially in who starts conversations and who checks in just to say hi, went along with higher satisfaction, even after accounting for attachment anxiety and avoidance.5 Leo and Sam may simply have been texting in two different languages, and nobody said so.

How anxious you felt wasn’t only up to you, either. Jeffry Simpson and Nickola Overall reviewed evidence that partners can buffer each other’s insecurity, helping the more insecure person feel calmer and act more constructively.6 The same worry can stay quiet with someone whose responses are steady and get loud with someone whose aren’t. That doesn’t make Sam responsible for your feelings. It does explain why you might not recognize yourself in that thread.

What’s yours to look at isn’t whether you were too much. It’s what you did when the fear spiked. Did you send another message, go quiet to protect yourself, or ask a clear question? That move is the part you get to change.

What to do this week

• Write the four lines. What happened, how you read it, what you were afraid of, and what you’d want to do differently. Keep each to one sentence.
• Stop rereading the thread. You’ve pulled out everything it can tell you. More scrolling mostly rehearses the fear.
• Name your early warning sign. The 1 a.m. draft, the third text, the urge to go quiet first. Knowing it is half the work next time.
• Practice one clear sentence for next time. Something like “I like hearing from you most days. Does that work for you?” Said early, in daylight, it saves a lot of decoding.
• Leave their label alone. You don’t need to decide what Sam is to move on. You only need to decide what you’ll do.

If the worry about relationships is constant, keeps you up at night, or is getting in the way of work or friendships, a licensed mental-health professional is the right person to help you with it.

What doesn’t the research show?

• The demand-withdraw research mostly studied couples, often married ones, not three-month situationships.1
• The relationship-specific finding used college students rating parents, friends and partners, with questionnaires over a few weeks. It’s suggestive, not a rule about any one person.2
• Both texting studies were small. Forty daters and 31 women can show a direction, not settle a question.34
• The texting-similarity study surveyed people already in relationships, at one point in time, so it can’t say mismatch caused anything.5
• Partner buffering is a review of couples research, mostly established relationships, not a few months of dating.6
• None of this lets anyone diagnose an ex, or themselves, from a phone. Attachment scales describe tendencies, and tendencies move.