I lost for six straight weeks. Not narrowly — comprehensively. I would sit down, build the wall, do the Charleston, and then spend the entire hand looking up at the card like it was going to tell me something new.
What I was doing was trying to find a hand on the card and then collect the tiles for it. Every week. It never worked, because by the time I had found a hand I liked, the tiles for it had been discarded by three other people who were faster than me.
Then one evening a woman at the table watched me do it for about twenty minutes and said, mildly, "Why don't you look at what you've got first?"
It sounds like nothing. It changed everything. The next week I looked at my rack before I looked at the card, picked the thing I was closest to, and committed. I still lost that week. The week after, I won twice.
The lesson was not about reading the card. It was about deciding. I had been keeping every option open for six weeks and calling it strategy.
What actually changed that week
The advice sounds like it is about the order of two glances. It is really about the order of three judgments, and the middle one is the one nobody tells you about.
First, read your own rack. Not "which hand do I want" but "what am I already close to". A pair, two pairs, three tiles in sequence in the same suit. That is a five-second read and it needs no card.
Second, read the discards. Not one tile at a time: the shape of them. What has gone that you needed, what is still coming around. This is the step I had been skipping entirely, and it is the one that decides whether the hand you picked is still reachable.
Third, pick from the card, and pick something the discards still allow. A hand you are close to that three other players have already been feeding is worth less than a hand you are two tiles further from but nobody has touched.
Read the rack, read the discards, then read the card. Six weeks of losses came from doing those three in reverse.
The part that was harder than it sounds
Deciding early feels reckless when you can still see three possible hands. It is not. Holding all three is what kept me from building any of them: I was discarding tiles that were useful to all three and essential to none.
The version of me from week one would have argued that keeping options open is free. It is not free. It costs you the discards you should have been making three turns earlier.
What I do now, at the first discard after the Charleston: name the one thing I am building, out loud if the table is friendly. Saying it makes it harder to quietly abandon, and it makes the abandon a decision rather than a drift.
If you are in the middle of a run like that, the more detailed versions of steps one and two are in how to read the NMJL card and reading the table. The short version is the one above.
For the other half of it (the part where you have to decide whether to keep pushing a hand or fold it while you still can) see push or fold. That article took me longer to write than any other piece on this site, because it took me six weeks to live through.