← All stories

What we learned from three hamsters’ nightly routines

Three hamsters across ten years, 4,549 km of running between them — two records complete, one still being written. They offer a glimpse of life after lights-out — and show why knowing your own hamster’s routine is more useful than judging a single quiet night.

You can spend time with your pet every evening and still miss much of his wheel activity. A sensor helped me follow Hamster 1’s nights in 2016, Hamster 2’s nine years later, and Hamster 3’s now. Across 864 recorded nights, the numbers became a way to remember their routines and notice how they changed.

Hamster 1, a pale Roborovski hamster, photographed in his cage in 2016
Hamster 1
Roborovski hamster
2016 – 2017
371 nights recorded
2,630km
total distance
20.1km
best single night
7.1km
average night
Hamster 2, a dark-furred Djungarian hamster, sitting up in his cage in 2025
Hamster 2
Djungarian hamster
2025 – 2026
461 nights recorded
1,735km
total distance
14.3km
best single night
3.8km
average night
Hamster 3, a sandy Roborovski hamster, looking out of his house in 2026
Hamster 3
Roborovski hamster
2026 Aug – present
32 nights recorded, as of 20 September 2026
184.2km
total distance
13.4km
best single night
5.8km
average night

Here is each hamster’s monthly activity. These are individual pets, of different species, living in different rooms and measured with different equipment. Their records are observations, not a controlled study or a guide to how far another hamster should run.

Nov–Jan 0 5 10 15 km per night, monthly average Hamster 1 Mar May Jul Sep Nov Jan Mar

Swipe to explore the monthly record.

Nov–Jan 0 5 10 15 km per night, monthly average Hamster 2 Mar May Jul Sep Nov Jan Mar May Jul

Swipe to explore the monthly record.

Most of their running happened while we slept

Before looking at changes over months, it helps to see when each hamster used his wheel. Despite the different rooms and equipment, both records show most running at night. That gives us a useful starting point, though it does not make the two setups directly comparable.

Every spin, sorted by hour of the day
Share of each hamster's lifetime running, by local hour
Hamster 1Hamster 2
night 0% 5% 10% 15% 20% % of all spins 12 15 18 21 00 03 06 09 hour of day — centred on midnight

Swipe the chart to explore all dates or hours.

Hamster 1 did 92.2% of his recorded running between 21:00 and 07:00; Hamster 2, 98.7%. Both records show very little wheel activity from 08:00 to 19:00. A quiet wheel does not by itself confirm sleep, and the first counter did not distinguish zero activity from missing readings.

In these records, running generally picks up around 21:00, is busiest between midnight and 03:00, and has largely tapered off by 06:00. The wheel tells us when they ran; it does not tell us exactly when they slept.

Their distances are not a fair contest. Species, size, surroundings and equipment all differ. The useful comparison is each hamster with his own earlier activity: what was usual for him, and how did that change?

Two quieter winters, seen in context

Both records begin in mid-March because both hamsters arrived as birthday presents. That means the calendar months line up, even though the records were made nine years apart. Below, each hamster’s monthly average is compared with his own first April–October average. This reference point, or baseline, helps show change without treating their distances as equivalent.

Two records, compared with each hamster’s earlier activity
Monthly average distance per night, as a % of that hamster's first Apr–Oct average
Hamster 1Hamster 2
Nov – Jan 0% 50% 100% 150% 200% % of own summer baseline Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul

Swipe the chart to explore all dates or hours.

100% marks each hamster’s first April–October average: 10.1 km a night for Hamster 1 and 5.4 km for Hamster 2. This is a comparison point, not a health target. Hamster 2’s January average was 2% of that reference and later rose again.

Hamster 2’s average fell from 7.7 km a night in September to 0.1 km in January — a 98% drop based on the underlying values. For six or seven weeks, his wheel activity was far below his earlier routine.

The timing raises a question about the season, but the wheel record cannot answer it. We do not have enough evidence here to separate the effects of light, temperature, age or health. Low activity alone does not establish torpor, rule out illness or show that a change is harmless.

A long record helps you notice what has changed. It cannot, on its own, tell you why.

Over the following months, Hamster 2’s recorded activity rose again, reaching 78% of baseline in May. His food, cage and wheel had not changed. Keeping the record meant this return was visible month by month, rather than something I had to piece together from memory. The monthly values are in the table below.

Their spring routines took different paths

Hamster 1’s activity also fell over winter. His December, January and February averages were all below his earlier baseline, as the chart and table show. At that point, the two records could appear to be following a similar pattern.

But Hamster 1’s March average was 12% of baseline, and April was 6%. His activity did not rise again in spring. The last recorded wheel movement was on 1 May 2017.

Looking back, the difference is clear: one hamster’s recorded activity rose again, while the other’s continued to fall. That describes what happened; it does not diagnose the cause. Two complete records cannot establish a rule for distinguishing a seasonal change from ageing or illness, and waiting for spring is not a way to decide whether a pet needs care.

Week by week, each on his own scale
Kilometres on the median night, by week of the record
0 5 10 15 20 Apr Jul Oct Jan Hamster 1 — km per night
0 5 10 15 Apr Jul Oct Jan Apr Hamster 2 — km per night

Swipe the chart to explore all dates or hours.

Separate scales let each record’s pattern remain visible. Hamster 1 (top) shows a summer plateau followed by a decline from November that does not reverse. Hamster 2 (bottom) shows a winter trough, four months of rising activity through spring 2026, and a sharp fall in June. A 12-day recording gap near the end is missing data; the readings on either side show lower activity.

Hamster 2’s later pattern was more abrupt. After reaching an average of 4.2 km a night in May, his activity dropped sharply in the first half of June. His last recorded spin was on 8 July 2026. Seen alongside his earlier nights, the size and timing of that change are much easier to appreciate.

A closer look at the hours they ran

Total distance is only one view. We can also look at hours in which the wheel moved and ask how many turns were recorded during those hours. Hamster 2’s tracker saved readings roughly every 30 seconds. The chart below compares each hamster’s typical active hour with his own busiest month; “median” means the middle value of those active hours.

Wheel turns during active hours
Median spins in an active hour, as a % of that hamster's best month
Hamster 1Hamster 2
Nov – Jan 0% 25% 50% 75% 100% median active hour, % of own peak month Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul

Swipe the chart to explore all dates or hours.

Only hours with recorded wheel movement are included. Hamster 2’s turns per active hour fell in winter and rose through spring; Hamster 1’s continued to fall. The measure combines how long and how fast they ran within each hour, and the older counter could group delayed readings together.

Fewer turns within an active hour could mean shorter bouts, slower running, or both. This measure does not separate those possibilities or measure physical capacity. It adds detail to the activity pattern, while the older counter’s timing limitations call for extra caution.

What this means for your pet

These records matter because they belong to pets whose routines were easy to miss while I slept. The longer record brings those unseen nights into everyday care: something concrete to look back on when a familiar routine begins to change.

Learn what is usual for your own pet. The winter changes here, including a fall of around 90% across parts of the records, are observations about these two animals. They are not a normal range or a safe threshold for another pet. A personal history gives you a more useful reference than someone else’s distance.

Use changes as a reason to pay attention. In February 2017, about ten weeks before Hamster 1’s record ended, I remember thinking he seemed a little slow but otherwise himself. His recorded activity was already much lower. That memory helps explain why I value tracking, but it is not evidence that the wheel predicted his health or the end of his life.

That is why long-term tracking with Collidera is useful: it keeps a picture of nights you cannot watch and details you might not remember. You can see when a change began, how large it was, and whether it continued. Those observations can support a conversation with a vet; they cannot replace an assessment of your pet.

The numbers behind the charts

Aligned by month of the record — both begin in mid-March, so the calendar months line up. Hamster 1 runs Mar 2016 to May 2017; Hamster 2, Mar 2025 to Jul 2026.

Month of recordHamster 1 km/night% of baselineBest night Hamster 2 km/night% of baselineBest night
Mar12.8127%15.110.3191%14.3
Apr12.1120%20.16.0112%12.6
May12.2122%16.26.5120%10.2
Jun11.7116%15.03.462%5.8
Jul8.080%12.95.6104%11.3
Aug7.878%11.34.991%10.0
Sep9.998%14.67.7142%11.8
Oct8.787%14.73.769%9.7
Nov8.585%11.90.611%2.1
Dec3.837%9.50.47%1.8
Jan +1y2.726%7.00.12%1.6
Feb +1y2.121%3.21.528%5.1
Mar +1y1.212%3.02.649%8.0
Apr +1y0.66%1.43.564%6.6
May +1y———4.278%8.8
Jun +1y———0.23%1.1
Jul +1y———0.00%0.0