North was inclined to raise her partner’s one-over-one major-suit responses irrespective of whether she held three or four-card support: maybe a great compliment to South’s declaring skills but not always an effective way to bid.
After getting the good news of heart support opposite, South rebid two spades, ostensibly a try for game, but when North showed help for potential spade losers and a good raise, South launched Blackwood in search of a slam bonus.
Now in this instance, South realized the heart Queen was missing from the North-South assets as his partnership employed Keycard Blackwood in which five hearts showed two aces but no trump-suit Queen.
“No problem” thought South: “Partner probably has four hearts so the Queen is very likely to fall and even if she only has three, it might still fall”.
Of course, there’s a wide gap in probability between “very likely” and “might” and good slam bidders will usually avoid contracting for twelve tricks when missing four trumps including the Queen when the total combined point count makes another non-trump loser very likely.
Proof of the pudding: one intractable trump Queen and an unavoidable diamond loser.
Back to the drawing board: despite North’s reluctance to rebidding one notrump with her balanced minimum and only three-card hearts, resisting making immediate raises of one-over-one response with only three-card support will make subsequent bidding far more accurate.
Yes, some pairs have elaborate schemes for asking partner if she’s made such a raise with three or four-card support but it seems far more effective to, as the French prescribe, require four cards in the first place.