The result is a point; the move is a path. The summary figure "price rose" says only where price ended up by the close of the horizon, and stays silent about everything along the way.
Why that matters
An episode that finished two percent higher may have dropped fifteen percent on the way and come back. In a summary it looks identical to one that rose calmly throughout.
For anyone planning to act on the information, those are two completely different cases. They would not have survived the first.
What to look at besides the result
Extremes along the way. The highest and lowest points price reached inside the horizon, measured from the match. They show how smooth the path was.
Spread between episodes. If across ten cases the rises were half a percent and the falls were ten, the share that rose misleads more than it helps.
Agreement. Did the similar cases resolve in different directions. If they did, that is not a "weak signal" but a direct answer: nothing follows from this shape.
How many cases are enough
Fewer than you would like, and more than you usually get.
At a handful of matches, talking about shares is pointless: the difference between "2 of 4" and "1 of 4" is one case. At dozens the picture is steadier, but it is still not a sample you compute probabilities from.
The practical conclusion: the fewer the matches, the more sense it makes to look at episodes one by one rather than at the summary.
Keep the horizon in mind
The outcome is measured over a stretch of fixed length after the match. That length is chosen in advance and does not adapt to the individual episode.
So "price rose" means "rose by the end of that stretch", not "rose and kept going". What happened beyond the horizon, the summary does not know.
A common mistake
Comparing outcome shares across different settings as if they were the same quantity. Changing the similarity threshold, the sample length or the timeframe changes both the sample and the horizon — such numbers are not comparable with each other.