Add all valid distances for one covering, then divide by the number of trials for that covering. Repeated values still count as separate trials. Compare the means in this test without claiming that every individual roll or future roll must follow their order.
Worked example
In a separate test, distances 3, 7, 8 cm have mean 18 / 3 = 6 cm. Distances 4, 6, 8, 10 cm have mean 28 / 4 = 7 cm. The second mean is greater, but its 4 cm reading is below the first set's 8 cm reading.
Two Wheel Coverings
These are fictional paper records; no moving-cart experiment is needed. One toy cart was released from the same marked ramp height onto the same floor, using the same distance method. Only the wheel covering changed. Covering P: 8, 14, 14, 12 cm across four trials. Covering Q: 9, 13, 11 cm across three trials. All readings are valid. Unequal trial counts do not prevent describing these means, but a balanced follow-up would provide a better planned comparison.