The first form is easy to identify. Somebody holding nothing watches a price climb, and every additional day of climbing adds to the sense of what non-participation is costing, until at some point they establish a position. This version gets discussed at length, and those involved will usually concede afterwards that their judgement was affected by the preceding rise. It is recognisable because the act itself is conspicuous. Nothing existed the day before and something exists the day after, which makes the decision impossible to overlook in a record.
The second form occurs among existing holders. The position is already there, yet after a rise it appears too small relative to what it could have been, and so it gets increased. Seen from outside, this is a holder adding to something they were positive about, which is ordinary behaviour requiring no particular explanation and attracting no comment. Seen from inside, what drove the decision may be identical to the first case, and the person adding may have no way of telling the difference.
The second form is well disguised because it comes with legitimate descriptions ready to hand. Adding can be presented as following through on the original view, as concentrating into a high-conviction position, or as responding to confirmation of a developing trend. Each of those descriptions is sometimes accurate and describes a defensible action. Their accuracy is exactly what prevents them from distinguishing one kind of decision from the other, since a true description of the surface tells you nothing about what produced it.
Both forms share a structure in where the decision originates. The change in size follows a change in price, while understanding of the asset has stayed roughly where it was. Rising prices do carry information, which cannot be denied and should not be dismissed, since a market repricing something is telling you that other participants have concluded something. The quantity of information carried is generally far smaller than the weight it seems to have at the moment of deciding, and that gap is where the trouble sits.
One consequence of that structure is entirely predictable. Total exposure reaches its maximum when prices are at their highest. No individual decision was mistaken, each addition looked reasonable at the time, and the reasoning behind them may have been perfectly sound in isolation. The problem lies in the distribution across time rather than in the quality of any single judgement. Where rising prices are the trigger for adding, additions will by definition cluster after rises, and the aggregate position will be shaped accordingly.
The test takes the form of a question about timing. Was the reason for this addition equally valid before the price rose. If it was, then the thing requiring explanation is why nothing was done at that earlier point, and the answer is usually about conviction rather than about information. If it was not, then the price supplied the decision and the reasoning was assembled afterwards to accompany it. Neither outcome is disgraceful, and only the second is worth changing.
Answering honestly tends to be uncomfortable, because the answer is usually the first one. The reasoning did hold earlier, and conviction was simply not sufficient then to act on it. That points towards something else worth noting. Writing out a sizing plan before establishing a position is considerably easier than assessing the motive behind each addition while it is being made. The earlier version of the question arrives without any money attached to the answer. Motives are much easier to read honestly when nothing currently depends on the reading.