Creating scenarios up front not as useful as before?

[Vy5]

When working the training type games, I saved a lot of by time creating scenarios but on recent exams, I am noticing far more local questions on this section that I always draw out question-level diagrams for. Is it bad practice to make the majority of the inferences (not laws, either/or in a particular space, etc.) upon the main diagram and move directly to the questions? The past work would then provide a guided brute force approach and help eliminate answers. Of course, drawing chains in pure sequencing and so on would be an exception.

It is my experience that recent games are a lot less smooth where making more than surface level inferences just wastes time. Am I missing something? I’ve tried to break down my approach after the September 2019 PT and so far I’ve only come up with this,

BTW, this question was inspired by a friend of mine who recommended making scenarios up front based upon the total number of questions.

[Dave Killoran]

When working the training type games, I saved a lot of by time creating scenarios but on recent exams, I am noticing far more local questions on this section that I always draw out question-level diagrams for. Is it bad practice to make the majority of the inferences (not laws, either/or in a particular space, etc.) upon the main diagram and move directly to the questions? The past work would then provide a guided brute force approach and help eliminate answers. Of course, drawing chains in pure sequencing and so on would be an exception.

It is my experience that recent games are a lot less smooth where making more than surface level inferences just wastes time. Am I missing something? I’ve tried to break down my approach after the September 2019 PT and so far I’ve only come up with this,

I still see plenty of games where templates/possibilities are the preferred approach, you just have to know when to pull that trigger. And on that S19 LSAT, a game like the Flowers game is so limited that just making the basic diagram limits possibilities so much you don't need to show each outcome—the outcomes are built right into the game. That's a good example of where the basic idea that underlies this approach is still in play, even on the hardest LG sections

BTW, this question was inspired by a friend of mine who recommended making scenarios up front based upon the total number of questions.

I don't love this as a deciding factor for making this call, fyi. I've crushed a bunch of 5 question games with templates before. The number of questions is just not determinative.

Thanks!