The walk you'll remember from this week is the bad one. It won't be the four quiet loops around the block, or the morning your dog saw a neighbor's terrier, looked at it, and then looked back up at you. It will be Thursday: the off-leash dog that came out of nowhere, the lunging, the stranger's face, the long walk home with your heart pounding. By Sunday, Thursday will feel like the whole story. You'll lie awake wondering if any of this is working, and whether the dog you love is just like this. That 2 a.m. doubt is real, but it isn't a fair measurement. It's what memory does to a record nobody is keeping.
This article is about one idea: with a reactive dog, how you feel about progress and how much progress there actually is are two different things, and your brain is very bad at telling them apart. Once you see why, you can start measuring the thing that matters, and most people find they've come further than they thought.
Your memory is a terrible training log
Psychologists have described for decades how lopsided our memories are. Negativity bias is the well-documented tendency for bad experiences to grab more attention, feel heavier, and stick in memory longer than good ones of similar size. From an evolutionary point of view this makes sense, since the rustle in the grass that turned out to be a predator mattered more than the hundred that didn't. On a dog walk it means one blowup outweighs a week of calm.
Then there's the peak-end rule, from Daniel Kahneman and his colleagues' research on how people remember experiences. We don't judge a past experience by adding up every moment. We judge it mostly by its most intense moment and by how it ended. A forty-minute walk that was lovely for thirty-eight minutes and ended in a barking fit gets filed away as "a bad walk."
Put those together and a reactive dog's guardian is in a tough spot. The moments that show progress are quiet by nature: a dog who notices and lets it go, a bark that stays one bark, a faster settle. Quiet moments don't make strong memories. The moments that feel like failure are loud. So without a written record, almost everyone underestimates how much has changed.
That matters for more than mood. People who are sure nothing is working tend to quit the plan, switch methods every two weeks, or push harder than their dog is ready for. Most of those choices come from bad data.
What progress actually looks like
Most people measure progress with one yes-or-no question: did my dog react? That's like judging recovery from a broken leg by whether you can run a marathon yet. It's too blunt to pick up the changes that come first.
Applied behavior analysts, the people who measure behavior professionally, don't track a behavior as simply present or absent. They look at things like frequency (how often it happens), duration (how long it lasts), latency (how long before it starts, or before it stops), and magnitude (how intense it is). Applied to a reactive dog, that gives you several separate dials, and they rarely all move at once:
- Distance. How close can a trigger get before your dog can't eat, sniff, or respond to you anymore? If that distance shrinks from across the park to across the street, that's real change, even if they still bark when a dog passes right beside them.
- Intensity. A dog who used to scream, spin, and hit the end of the leash and now gives a stiff stare and two barks has improved a lot. The bark is still there, but there's far less of it.
- Recovery. How long until your dog sniffs the ground again, takes a treat gently, or answers to their name? Shorter recovery is often one of the earliest and most reliable signs that the nervous system is changing.
- Frequency. How many reactions per walk, or per week, compared with a month ago?
And there's a dial that hardly ever gets counted: the reactions that didn't happen. The trigger your dog saw and chose to walk away from. The dog at the end of the block that would have set them off in spring and got a glance in autumn. Those non-events are what success looks like, and nobody remembers them unless they write them down.
The quiet wins are the real ones
Picture it. A dog appears at the far corner. Your dog's ears go forward and their body goes still. Six months ago, this is where the explosion started. Today, after a second or two that feels much longer, your dog turns their head back toward you, as if asking what happens now.
You probably won't celebrate it. You might not even notice it, because you're busy watching the other dog and getting ready to steer. But that head turn is the target behavior. It's a dog noticing the thing that scares them, feeling the fear, and still having enough of their thinking brain left to make a choice. Every protocol worth using is built to make that moment happen more often.
If you only count explosions, you'll miss it every time.
Why bad days still belong on the chart
The hardest part of measuring honestly is that the line doesn't go straight up. Behavior change is uneven. A dog's tolerance on a given day depends on sleep, pain, weather, how many stressors have already piled up that week, and what showed up around the corner. A bad Thursday doesn't erase the trend any more than one cold day cancels summer.
This is why a log beats a feeling. Once you have a few weeks written down, you stop asking was today good? and start asking are the bad days less bad, less frequent, and shorter than they were? That second question is the one that tells you whether training is working. Often the answer is yes, even in a week when it doesn't feel that way.
A log also catches real plateaus. If after several weeks your distance, intensity, and recovery numbers truly haven't moved, that tells you something useful: maybe it's time to bring in a qualified behavior professional, check for pain with your vet, or change the plan. A log isn't there to make you feel good. It's there to make you accurate, and sometimes accuracy is bad news you need to hear.
How to keep a log you'll actually keep
The best log is the one that takes less than a minute, because a detailed spreadsheet you give up on after four days tells you nothing. For each walk or outing, write down:
- What the trigger was (dog, person, bike, whatever it was).
- Roughly how far away it was when your dog first noticed.
- What your dog did, on a simple scale: noticed and moved on, stared or got stiff, barked, or lunged and lost it.
- About how long it took to recover, using a rough count: seconds, a minute, or the rest of the walk.
That's it. Over a month, those four details turn into something your memory can't argue with.
Your next moves
- Tonight, write down your baseline. Before any more training, jot down what a typical bad encounter looks like right now: the distance, the intensity, how long recovery takes. Date it. It's hard to see change when you have nothing to compare against.
- Log every walk this week, even the boring ones. Especially the boring ones. A walk with no reactions is data, and it's the kind your memory drops first.
- Count one quiet win per walk. Look for the head turn, the sniff after a sighting, the bark that stayed small. Write it down in a few words. You're training your own attention as much as your dog's.
- Time one recovery with your phone. Next time your dog reacts, start a timer when it begins and stop it when they can take a treat or sniff again. Do the same in two weeks and compare.
- Put a review on your calendar for four weeks from today. Read the whole log in one sitting. Don't judge progress by any single day before then.
Where Mellow fits
Mellow was built around this exact problem: the gap between how a reactive dog is actually doing and how it feels to the person at the other end of the leash. Along with guided behavior-modification sessions for anxious and fearful dogs, it gives you a quick way to log distance, intensity, and recovery after each outing, and then shows you the trend over weeks so a bad Thursday sits in its proper place. If you'd like a clearer view of how far you and your dog have come, you can take a look at Mellow here. The log above works just as well on paper, though. What matters is that you start keeping one.