How do you avoid trial and error on LG? Forum

[NomNom]

I keep seeing "avoid trial and error", but on questions like "COULD BE TRUE", how do you avoid this? I try to setup "maps" or frames to see what is possible, but sometimes I don't get them all down. Just wondering what others do.

It could be that I am not doing enough setup/inference work before diving into the questions I imagine...but I panic about time.

[minnbills]

Make sure you do the local questions first. They give you free hypos that you can use on questions like "could be true."

If you do have enough hypos you will usually get "could be true" questions without having to do any heavy lifting.

[Bobeo]

I'm interested in this as well.

[Tiago Splitter]

Very rarely will you completely avoid trial and error on a game. Just get the global diagram going right away, try to make a few inferences, and then get the pencil moving.

[hyakku]

Make sure you do the local questions first. They give you free hypos that you can use on questions like "could be true."

If you do have enough hypos you will usually get "could be true" questions without having to do any heavy lifting.

This. I wasn't doing this up until like a week and a half ago, and while my games usually were between -1-4, I wanted it more consistent. Been doing this and disregarding careless errors I'm mostly -0, -1 now. I used to think the same thing.

I also find that doing those gets the rules more situated in your head without you really having to try and often times you'll be able to beast out hypos in a second or two once the rules are fixed in your head, helping out with those cannot or except questions that are particularly challenging.

[NomNom]

Great advice, thanks!!

[notaznguy]

You can't completely avoid it. You'll usually have to do trial and error on at least 2 of the questions, but there are usually inferences where you can eliminate at least 1 of the 5 answer choices so you don't need to waste time doing all 5. Sometimes you just end up plugging and chugging for 3 answer choices instead of 5.

The best thing you can do is to note down inferences and some interchangeables-randoms, especially for linear games. Usually the interchangeables are in the answer choices and you can eliminate both of them because if 1 of them doesn't work, then obviously the other one isn't going to work either. I don't mean you automatically eliminate ALL interchangeables, but there are some examples where you can.

For instance, if the rules are:

There are 6 letters, LMNOPQ

K or L > M
M > N
-----
K or L > M > N

P and Q are interchangeable because they are random.
L and K are interchangeable as well.

And a question says: "Which one of the following CAN'T go first?"

A) L
B) M
C) P
D) Q
E) K

Obviously we know that L can indeed go first, and since L and K are interchangeable, we can eliminate them. Obviously P and Q are both randoms and therefore have no restrictions, so they can go first too. Since they're interchangeable, we knew that if P could go first, so could Q. That leaves us only with answer choice B) M.