I completely agree with BPA, but I just want to point out that Edmonton did that, and it led to problems when they kept winning the lottery where the BPA were high end forwards and they desperately needed high end D.
Not saying we shouldn't pick Kakko if he's there. Just pointing out a similar situation led to issues elsewhere in the past.
To be somewhat controversial, my personal philosophy is not "BPA", it's more of a "Contextual BPA".
What I mean by this is essentially tier-based. Let's just say I'm the Devils picking in the second round (33 OA). On our scouting board, the BPA is an undersized LW who our scouts have rated 21st OA at a rating of 7.3 and the second player on our board is a big, fast RD who we have 23rd OA at a 7.1 rating. In this particular case, if I were GM I would draft the defenseman. It's a more important position and a greater area of organizational need, as if the Devils pipeline is overflowing in one area, it's undersized, skilled LWs (Bratt, Boqvist, Zetterlund, Gignac).
This is also why I would have drafted Ty Smith or K'Andre Miller last year, even if Joel Farabee and Grigori Denisenko were both still on the board.
Conversely, let's say I'm picking 33 OA and the best player on my draft board is an undersized, skilled LW who we have ranked 14th at a 7.9 rating, and the second best player is a defenseman ranked 23rd at a 7.1 rating. In this case, I would take the LW, due to the fact that we have him rated a full tier (0.5) over the next best player. The only other option I would consider in this particular hypothetical would be a trade-down scenario.
Personally, I always weigh Cs and D over Ws. A center can always be converted to W, not so much the other way around.