Structuring the game and drawing inferences

[RoxanaA644]

Hello,

I missed almost every question on this game and I had a hard time drawing things out since it had 3 different levels. Can anyone help efficiently draw out the game so it's easy to use while answering questions? I also missed questions 2-6 and could really use some explanations on those.

Thanks!

[charlotte.f.blatt.18]

I too had a hard time with this game, could anyone (in particular a Manhattan Geek perhaps ) recommend other similar games like this one or like PT 74 Game 4 which is also a conditional grouping game to practice with?

Thank you!

[ohthatpatrick]

It is definitely an annoying game, just because it's so loose. I don't use the Open Board that Manhattan Prep uses for games like these. I just write out scenarios horizontally, because that allows me to more quickly test out some plug-n-chug hypotheticals.

My setup / diagram would look something like this (it's hard to make stuff look right in this typing environment):

1 ≤ G R W ≤ 3

__ + .... __ + .... __+
M .........J ...........S

If we were doing the Open Board setup, we'd represent the maximum possible per column, which seems to be 3 by default, although rule 2 shows us that column S could never have all 3.

__ __ /
__ __ __
__ __ __
M..J...S

RULES:
Gj --> Rj
~Rj --> ~Gj

Rs --> ~Ws
Ws --> ~Rs
(as we said before, this guarantees that R and W will never BOTH be in column S)

Xm --> Xs
~Xs --> ~Xm
(we tend to use "X" when we're looking to represent the idea of "any character")

INFERENCES:
We got conditional rules, so we should see if any of them chain together. Rules 2 and 3 could chain together to give us
Rs --> ~Ws --> ~Wm
and
Ws --> ~Rs --> ~Rm

Since being in M automatically sends you to S, people like R and W who are mutually exclusive in column S must also be mutually exclusive in column M.

If we're using the Open Board, we could update column M to reflect that it would never have all three people.
/... __ ../
__ __ __
__ __ __
M..J...S


FRAMES?
Are there any chunks or 2/3 way splits that would be worth investigating?
There are no chunks. It's hard to think of a good 2 or 3 way split.

My first instinct was to think about the various possibilities for column M, since that would also tell me about some of the people in S.

But it seems like the possibilities for M are too numerous, like it could be
GR
GW
G
R
W

I don't really want to do 5 different scenarios.

The Gj -> Rj rule is almost a chunk. If you frame a conditional, you make one frame where the trigger happens and one where it doesn't.

When G is in column j, we would know that R is there, but we wouldn't know anything else.
When G isn't in column j, we would know that G is in column S (it has to be at least somewhere, and it couldn't ONLY be in column M).

But both of those seem too anemic to be worth sketching out.

So I would just get cranking on the questions. After doing the orientation questions, I would prioritize all the "If" questions first so that I can get a bunch of possible scenarios out on the page.

My scenarios would look like this on my page:
G | G R W | G R/W

G W | R | G W

etc.

I have the master diagram on my page to reference in terms of reminding my brain that the leftmost cluster is M, the middle one is J, and the rightmost one is S.

Otherwise, all I'm doing it playing around with these three letters, making sure I use each one at least once and making sure I obey the three conditional rules.

OTHER SIMILAR GAMES
The most immediate comp that came to mind was June 2007, game 4.
PT62 game 2, about stained glass windows also comes to mind.

The game from PT74 (while definitely hard) doesn't seem as similar to me because you only use everyone once (the 6 people will only appear once, either in R, S, or OUT)

Part of what makes this game and the others like it annoying is that we can re-use characters.

[andrewgong01]

I used the Open Board diagram and since the test is 1.5 weeks away I don't think I will switch my strategy for these types of games but I am curious for the open board why did you make the base of the board JMS (the plays) instead of the student.

I did it by student because I thought that was easier to track because Q21 and Q22 gives a condition diretly on how many plays each student had and making the students the base makes it easier to track Q23 since the conditional rule in the question requires seeing the visual of who had what parts. I know typically the orientation question would hint at the set up as to what to use as the base but I think this time the orientation base was "misleading" and I ignored it this tie. I would agree though that your way makes it easier to do Q20 by making the plays the base.

[ohthatpatrick]

Honestly, I just default into always making the people the movable parts and the other things the groups.

When I first size up a game, I'm mainly thinking, "Is this Ordering or Grouping?", and if I see we're sending people into various categories, I think "grouping" and make the non-people the groups.

And because the rules are conditional statements, it seems even more like a Grouping game.

The sort of game you're thinking of, where we make the exotic decision to make the PEOPLE base is what many of us would call "an OPTIONS" game.

It's certainly possible to construe this as an OPTIONS game, in terms of the setup paragraph.

Each of the three people will present on at least one of three possible subjects.

But, none of the rules end up acting like they would in an Options game.

Options game rules would sound like:
- George presents on more subjects than Rita does
- Rita and Wendy don't present on any of the same subjects

Without "comparative quantities" rules or "matching / mismatching options" rules, I wouldn't think it's an OPTIONS game.

And OPTIONS games are also so rare (five or six ever? all of them between test 40-55) that it would take a lot to push me in that direction.

That said, it's obviously perfectly doable to do this game either way (and may even be better the way you diagrammed it).

[JenniferK632]

I used a sectioned off In/Out board for this game, and it worked pretty well. The three sections were J, S, and M. I think I've seen other MP games taught this way.

Overall, I find that though it's hard to put all conditional inferences on the board, it's pretty necessary for me. Does anyone have time-saving tricks for diagramming or answering questions?

[Laura Damone]

This is a pretty standard open grouping game, so I wouldn't advise diverging from that standard setup.

M J and S should be the base, and we know that most from looking at the answers to q.19. If you're ever unsure of which element set should be the base, check the answers to the first question! The rules themselves don't give you enough solid clues in this game, and checking those answers prevents you from having to redo the diagram to switch it up after realizing your mistake!

As far as tips for managing conditional rules, I always think about guarantees and prohibitions. Rule 1 guarantees that if G is in J, there's a GR chunk there.

Rule 2 prohibits a WR chunk in S
The combo of rules 2 and 3 prohibit a WR chunk in M (since it would force that chunk into S as well).

Rule 3 guarantees that if someone doesn't present on M, they do present on J.

Thinking about the rules like this, instead of just as abstract conditionals and chains, can make them easier to use as you play the game. It's a great step for your Big Pause. It's also helpful to think about what will drive the game during your Pause. I predicted the combo of rules 2 and 3 because the double anticunk is so limiting. That helped me get 20 and 21 without drawing anything out. 22 and 23 were a bit of a slog, but sometimes plugging and chugging is just what you gotta do! You only needed to test 5 answers total between those two questions, so it's not too bad, all in all.

Hope this helps!