How to interpret column 'N' in tblastx output?

My understanding from this page https://www.biostars.org/p/272364/ is that it’s related to ‘High Scoring Pairs’, which are in turn related to the way BLAST works. I’m still not sure how to interpret that for my data.

I’ve blasted a set of contigs that contain at least some of my genes of interest (and often nothing but) against a database of my genes of interest in known organisms. Top hits are all E=0, no issue here. Where things get complicated is that I know the top-scoring hit (let’s call it ‘A’) is from a slightly more distantly-related organism. In fact, among the top hits (bit score 988-991), all that distinguishes the ‘correct’* match (let’s call it ‘B’)is the fact that it also has the highest N.

*I call it a ‘correct’ match because there is additional information (phylogenetics of the genes involved, genome-level phylogeny) which lend support to the fact that the organism associated with the contig in question is more related to B than to A.

Now, before I go and use N as a guideline in the rest of my blast searches, what does it actually mean? Instinctively I’d assume that a higher N is worse, because ‘better’ match would just have one very large High Scoring Pair, while a ‘worse’ match might have more HSPs for the same area, with mismatches inbetween them. But surely this isn’t the case…

Help would be appreciated!