range of quantities in Game

[pixelzoom]

The design doc indicates that quantities for the Game range from 0-10, but the mockups show a single digit for the spinners. 2 digits will be 2x wider, so the spinners will need to shrink. For example, see screenshot below (from the design doc). Reducing the spinner size to accommodate 2 digits may make them unusably small. And we should try to avoid having different-sized spinners for different equations in the game.

[pixelzoom]

@ycarpenterphet wrote:

Ah, good point. Let me dig into why I wrote that we needed 10 molecules in the game, and if it really is needed, I’ll do a quick adjustment and new mockup.

@ycarpenterphet wrote:

Do you have a complete list of the game equations? I am not sure if the list in the Java-sim design doc is up-to-date. Also, can you tell me how the problems are generated in the game (i.e. how the number of molecules in the challenge is determined)?

[pixelzoom]

Answers to both questions are in the Java source file (NumberOfVariablesChallengeFactory.java) here in the Unfuddle SVN repository:

https://phet.unfuddle.com/a#/projects/9404/repositories/23262/file?path=%2Ftrunk%2Fsimulations-java%2Fsimulations%2Freactants-products-and-leftovers%2Fsrc%2Fedu%2Fcolorado%2Fphet%2Freactantsproductsandleftovers%2Fmodule%2Fgame%2FNumberOfVariablesChallengeFactory.java&commit=74168

The comment towards the top of the file documents the strategy for generating challenges. The constants LEVE1_LIST, LEVEL2_LIST and LEVEL3_LIST define the pools of equations for each game level.

To parse the equation names… Single underscore separates terms, double underscore separates sides of the equations. For example: Reaction_H2_F2__2HF.class is H2 + F2 -> 2HF

[pixelzoom]

@ycarpenterphet wrote:

Ok, I did a bit of testing and let’s go with the following plan:

(1) Keep all spinners (across all 3 screens) to a maximum value of 8. (2) Keep layout as is (since the spinners will remain a single-digit width.) (3) Use stacked molecules in Sandwiches + Molecules screens, and randomized molecule layout in the boxes in the game.

As a note, the reaction LEVEL2_LIST.add( Reaction_P4_10Cl2__4PCl5.class ); will now only ever be used in the case where no products are generated (in either L1 or L2), but I don’t think that’s a problem, pedagogically. If that’s a problem for the coding, or makes the selection process somehow inefficient, I’d be ok with removing this reaction from the pool.

How does that sound?

[pixelzoom]

Sounds good. I’ll evaluate Reaction_P4_10Cl2__4PCl5 when I get around to implementing the game.

[pixelzoom]

Reaction_P4_10Cl24PCl5 is causing problems in the generation of game challenges. One of the requirements is that every game will contain exactly one zero-products challenge. To make a long story short, the algorithm for generating challenges assumes that it can choose any reaction, and that it has control over determining whether the challenge will have zero products. Handling Reaction_P4_10Cl24PCl5 will require modifying the algorithm to adapt to situations where the reaction cannot possibly have non-zero products because of the quantity range restrictions. I’d rather not modify something that is working well. So unless you’re attached to Reaction_P4_10Cl2__4PCl5, I recommend dropping it.

[ycarpenterphet]

After double-checking the full set of game reactions, I feel totally comfortable removing the reaction Reaction_P4_10Cl2__4PCl5 from the pool.

It serves no particular pedagogical niche, and removing it allows us to avoid all sorts of exception cases for the algorithm to handle in game, so it's not worth keeping.