Yoshi's Cookie(NES) FAQ version 0.5.0
by Andrew Schultz schultza@earthlink.net
copyright 2002

Please do not reproduce for profit without my consent. If you wish to do 
so then please write to me asking for specific games and addressing my 
by my first name. Thanks!

**** AD SPACE ****

My web page:
http://www.geocities.com/SoHo/Exhibit/2762

================================


           OUTLINE


  1. INTRODUCTION

    1.1. WHAT THIS FAQ COVERS

    1.2. WHAT IT DOESN'T

    1.3. GENERAL FAIRNESS/UNFAIRNESS THOUGHTS

    1.4. IN-FAQ TERMINOLOGY

  2. THE BASICS

    2.1. RULES

    2.2. CONTROLS

    2.3. SCORING

  3. THE FIRST HUNDRED BOARDS

    3.1. HARD CODED FIRST FEW LEVELS

    3.2. DIMENSIONS AND DIFFERENT SQUARES

      3.2.1. THE FACTS

      3.2.2. MATHEMATICAL INTERPRETATIONS

    3.3. MISCELLANEOUS MOVES TO LEARN

  4. THE NEXT EIGHTY-NINE LEVELS

  5. GENERAL STRATEGIES FOR THE FIRST HUNDRED BOARDS

    5.1. A QUICK(OR EVEN WINNING) START

    5.2. WHAT TO EXPECT FROM WHICH RECTANGLES

    5.3. USING THE CONTROLS EFFECTIVELY

      5.3.1. THE PIVOT

      5.3.2. WALKING OFF THE EDGE

      5.3.2. THE HALF-LINE-UP

      5.3.3. 'OTHER'

  6. GENERAL STRATEGIES FOR THE GREAT BEYOND

    6.1. A QUICK(BUT NEVER WINNING) START

    6.2. KNOCKING OUT THE KOOPA

    6.3. WHAT TO EXPECT FROM WHICH RECTANGLES

  7. TRICKS

  8. AFTER-LEVEL SKITS

  9. MY SUGGESTION FOR A NEAT SIDE GAME

  10. VERSIONS/CREDITS

================================

  1. INTRODUCTION

    1.1. WHAT THIS FAQ COVERS

  This FAQ covers strategies for the first ten boards as well as the 
next few but may tail off in detail. It aims to discuss specific cases 
that may uncover finesses for various strategies or develop intuition 
further. While Yoshi's Cookie is an easy game to hack through and manage 
to conquer the first ten levels with a few falls there are some nice 
theoretical questions and I will attempt to answer the simplest ones 
here.

    1.2. WHAT IT DOESN'T

  I don't really intend to cover two-player mode here, as I am more 
concerned with the theory of the tiles, etc. Plus the two-player version 
is theoretically trivial after this document(I hope.)

    1.3. GENERAL FAIRNESS/UNFAIRNESS THOUGHTS

  The original levels(1-10) are fair, with each different cookie showing 
up clearly. I can't fully say the same for the Yoshi-characters that 
appear in levels 11-99. You get a lot of waiting around for the right 
wild card, in which time tiles also have a decent chance of piling up 
without your having a say in anything.

    1.4. IN-FAQ TERMINOLOGY

--board and stage are interchangeable terms
--'the rectangle' means the rectangle of blocks you need to coordinate 
in rows.
--occasionally I will refer to moves in very short-hand notation. The 
move '5U' means that you go to the fifth column from the left and move 
everything there a square up. The move '2L' means you go to the second 
row from the bottom and move everything there a square up. I may also 
use (5,2)U to tell you which square is best to go to, and if you need to 
move up several times I will say 5Ux2 for instance, or (5,2)Ux2. So the 
lower left can be considered an origin of sorts.

  2. THE BASICS

    2.1. RULES

  Yoshi's Cookie is pretty simple. You start out with a rectangle of 
cookies/blocks. At any point in the rectangle you can pivot its row up 
or down one or its column left or right one. The object is to get all 
objects in a row/column the same type. Then the row/column vanishes. 
Eventually you hope to eliminate everything. Extra rows and columns 
constantly pop up from above and to the right. This may help you if you 
cannot make any new rows, but as the level wears on the rows/columns 
appear more frequently. If you wind up with a rectangle 7 on one side 
the game is over.

  There's a gauge in the lower right that shows about how many of each 
shape you've put into a row. When the gauge is full, a Yoshi pops up. 
Yoshi counts as a wild card and is useful on level 10+(a twist on the 
original game, shown when level 10 is completed) as well as near the 
end.

  What's nice about Yoshi's Cookie is that you can always continue from 
the level where you left off. So if you just want to breeze through 
it(generally takes me about two hours) you can do so.

    2.2. CONTROLS

  B button: causes pieces to move in faster
  arrow key: moves the cursor. If you are at an edge of the rectangle 
and move off it you appear on the other side. Seems to cause blocks to 
move more quickly too; if you push up or down, the descending blocks 
come in a bit quicker.

  A button + direction: moves the row/column you're in in that 
direction.

    2.3. SCORING

  You get 10*2^(x+y) points for clearing off a row of non-Yoshis, where 
x = # of blocks in the row - 2, and y = the number of chain reactions 
before your current row-clearance.

  Per level I tend to get 5000 for the first level up to 10000 per 
level(if I survive all the way through) for the later ones.

  The Yoshi tends to make scores go weird. It seems to add ten points to 
whatever row it touches instead of doubling--so 2 in a row + yoshi is 
less than 3 in a row.

  3. THE FIRST HUNDRED BOARDS

    3.1. HARD CODED FIRST FEW LEVELS

    The first few boards seem to be set up for quick wins. I'll indicate 
different pieces with 1/2/3 as the colors seem random although the 
formations aren't. Learning how to deal with these boards expediently 
will help you to perhaps complete others without adding extra blocks. 
It's not too hard to make a row, but making a row and leaving yourself 
open for even more later will help tremendously in the long term. Given 
how monochrome they are, they probably won't come up long term, but the 
strategies involved may help you later.

  Board 1-1:

  1 2 1 2

  2 1 2 1

  1 2 1 2

  2 1 2 1

  This one is pretty hard to botch. 2U, 4U is one way to do it although 
there are so many solutions to start off it is laughable.

  Board 1-2:

  1 2 1

  2 1 2

  1 2 1

  Interestingly, smaller is tougher here. The odd number of squares 
means a move of 2U sticks you where you have to wait for new tiles to 
come in. So instead what you should do is (3,3)UL. That will get rid of 
a row and a column, and then any move wins.  

  Board 1-3:

  1 2 1 2 1

  2 1 2 1 2

  1 2 1 2 1

  2 1 2 1 2

  This is your first non-square rectangle, but it is as easy as the 
first one. And it gets you double the points, too.

  2U, 4U will clear the board nicely. It'll take longer to watch the 
rows removed and points tallied than for you to make the right moves.

  Board 1-4:

  1 2 1 2 1

  2 1 2 1 2

  1 2 1 2 1

  2 1 2 1 2

  1 2 1 2 1

  Similar to board 1-2 only this time a little more fiddling may be 
necessary. This may not be the optimum solution, but it works. My first 
try(move every other row down one) leads to a general problem--if you 
have a 1-by-any rectangle, you cannot make further progress. So what I 
do is to go 5U 3U 1D. This reduces the structure to a 3x2. There's still 
a chance to mess up unless you go 1D and win.

  Board 1-5:

  1 2 3

  2 3 1

  3 1 2

  This level has a nice symmetry about it. No matter where the cursor 
is, just move your row left, go up one, and move the row right. You'll 
win.

  Later levels seem to be drawn from a template or possibly even a large 
set of possible designs(note--I count boards where you can swap colors 
and they'll look the same to be identical.) With three colors the boards 
are somewhat checkerboardy and later on I see familiar patterns cropping 
up where you'll have the following in several rows:

  1

  X

  1

  If you can move these vertically to line them up you have a good 
start. Board 1-6 has several different possibilities, so I can't give an 
explicit solution.

  But in general on level one your big hazard may be leaving a single 
row, where you'll have to wait for reinforcements. Take a pause before a 
chain reaction; you can probably afford it.

    3.2. DIMENSIONS AND DIFFERENT SQUARES

      3.2.1. THE FACTS

  This document describes the starting rectangle dimensions and number 
of different colors at the start of each board.

Board|Dim's|Colors
-----+-----+------
1-1  | 4x4 | 2
1-2  | 3x3 | 2
1-3  | 5x4 | 2
1-4  | 5x5 | 2
1-5  | 3x3 | 3
1-6  | 4x4 | 3
1-7  | 5x4 | 3
1-8to| 5x5 | 3
1-10 |     |
2-1  | 4x4 | 3
2-2  | 5x4 | 3
2-3  | 5x5 | 3
2-4  | 6x5 | 3
2-5  | 4x3 | 4
2-6  | 4x4 | 4
2-7  | 5x4 | 4
2-8  | 5x5 | 4
2-9  | 5x5 | 4
2-10 | 6x5 | 4
3-1  | 6x5 | 3
3-2  | 6x6 | 3
3-3  | 5x5 | 4
3-4  | 5x5 | 4
3-5to| 6x5 | 4
3-7  |     |
3-8to| 6x6 | 4
3-10 |     |
4-1  | 4x4 | 4
4-2  | 5x4 | 4
4-3  | 5x5 | 4
4-4  | 6x5 | 4
4-5  | 6x6 | 4
4-6  | 4x4 | 5
4-7  | 5x4 | 5
4-8  | 5x4 | 5
4-9  | 5x5 | 5
4-10 | 5x5 | 5
5-1  | 5x4 | 5
5-2  | 5x5 | 5
5-3to| 6x5 | 5
5-5  |     |
5-6to| 6x6 | 5
5-10 |     |
6-1  | 6x5 | 5
6-2to| 6x6 | 5
10-10|     |

      3.2.2. MATHEMATICAL INTERPRETATIONS

  It appears that with two colors, you can always find a way to reduce 
the entire board before play begins. I have no proof and there doesn't 
seem to be one without strong induction. I suspect you can show that you 
can always leave a corner in place where you have:

 2 2
 1 1 2

  ...i.e. a pair of each color, so that any chain reactions will not 
leave you with a dead end.

  The first four boards are pretty trivial since they are pre-planned as 
you see above, but when you get to a few more colors some observations 
can come in handy. The 'Pigeonhole Principle' is useful here. In a 
nutshell it states, for our purposes:

  If you have X squares and Y different colors, then you must have at 
least (X/Y) rounded up of one color. So 12 squares of 5 colors mean that 
you must have at least 3 of 1 color. But with 10 squares you need only 
have 2 of any one. If you have 2 of every color you won't have 3 of any.

  So let's start with applying this on board 1-6. Four by four, sixteen 
squares, three colors. You will have six of one color. Unless there's a 
color with four squares(reducing to a two-color board would make this 
level very easy,) eliminate a row of four squares with that color, and 
you'll have at least two of that square's color. If you eliminate a 
color with five squares, then there will be one left, and that is bad.

    3.3. MISCELLANEOUS MOVES TO LEARN

  4. THE NEXT EIGHTY-NINE LEVELS

  5. GENERAL STRATEGIES FOR THE FIRST HUNDRED BOARDS

    5.1. A QUICK(OR EVEN WINNING) START

    5.2. WHAT TO EXPECT FROM WHICH RECTANGLES

There are 28 possible rectangles with sides from 1 to 7, which is what 
you will see in the game; I don't count 2x3 and 3x2, for instance, as 
separate.

In general it is better in the long run to try to make a rectangle 
slimmer than closer to a square. However if you are pressed for time and 
just want to reduce your workload and can't see any way to do this, of 
course you need to cut down what you can. So for instance if you have a 
5(horiz)x6(vertical) rectangle it would be ideal to find six vertically 
in a row but copping out with five horizontally in a row is okay. With 
experience you will be able to target which cookie you have the most of 
and how to shuffle it around quickly. The rule of thumb to remember is 
that narrower rectangles provide a chance for easier rows to knock out, 
but on the other hand you don't want to knock out a row just for fun.

1x1: automatically solved
1x2-6: will automatically become 2x3-7 as there is nothing you can 
really do. But this is not so bad as it can be immediately reduced 
drastically.
1x7: you'd better hope the 1 side is filled next and not the 7, or it 
will be game over.
2x2: assuming all four colors are different, we should try to push for a 
2x3 rectangle instead of a 3x3 rectangle. Hold up or right on the 
controller--or try for a diagonal move.
2x3: the 'pigeonhole principle' says that if there are y items and x 
slots to put them in, one slot must have at least the integral part of 
((y/x)+1) items. Replace slots with colors and y with cookies and we can 
apply the problem to the game. In this rectangle we will have 1+(6/5)=2 
of one color for sure.
2x4-7: these are very advantageous. You can always reduce to 2x2 and 
rather quickly I might add. Pick any type of cookie of which you have 
two(there always will be one, and preferably the total number should be 
even, but this is a finesse.) Assume it's color A and look below.

** **
** **
** **
A1 **
** **
** **
** A2

Obviously you will get things done if you cycle the columns through 
whether you push A1 or A2 down or up. It's a bit tougher with the 
following:

** **
A1 **
** **
** **
A2 **
** **
** **

In this case put the cursor on one of the cookies, move right once and 
then up/down to get them to match.

There's a finesse if you have exactly three of a type of cookie. Let's 
say you have the following:

** **
** **
A1 **
** A2
A3 **
** **
** **

It looks like the easy choice would be to move up/down. But actually it 
is not. Moving A2 off to the left may be the better option. Because if 
you erase two, only one will be left and you may not be able to solve 
this. If you put all three to the side, you will still be able to clear 
something(even with 2x4 there will be 5 squares remaining and 4 cookie 
types--all except A--to choose from.) You can always reduce to a 2x3 and 
then it will be reduced to a 1x3 right away. So for instance below we 
could do the following:

A1 B1
C1 A2
A3 C2
B2 D1
D2 E1
E2 B3

Here you can do the following: 5L 3R 2D to give:

A1 B3
A2 B1    A1 B3
A3 C1 => A2 B1
D1 C2    A3 C1
D2 B2    D1 C2
E2 E1    D2 B2

Now to eliminate the D's and C's since the E's are matched. Let's focus 
on tackling the D's--easy enough. 2L 1U. Note that knocks out the C's 
too.

A2 B3
A3 B1    A2 B3
C2 C1 => A3 B1
D2 D1    A1 B2
A1 B2

Oh dear, the A's and B's go too, with a little extra bonus.

More conventional play may have isolated an A or a C.

3x3: the pigeonhole principle only assures us that there are at most two 
of one type. So we may be out of luck here.
3x4: We may have 12/5+1=3 of one type--in fact, of two types of cookie. 
First see if for any type of cookie you have one in each column. That is 
the easiest way to settle business. However if you have four of one type 
of cookie, not always the case, you might want to try to go for a 2x4.

This illustrates another rule of thumb--try never to leave one of one 
type of cookie--and another--when one type of cookie gets rare, try to 
leave an even number, or if there are three of it put them in a row so 
two are not accidentally gone(it also helps in the future if more attach 
to the rectangle.)

3x5-7: unlikely you'll have 5-7 of any of a type of cookie but if it's 
easy to make a big column of them try to do so. Fortunately it'll be 
easy to reduce to 3x3. Note that if you have four of one type of cookie 
you probably don't want to get rid of it unless you have to. It's also 
not possible to have four of each type of cookie remaining or you'd have 
4, 8, 12, 16, 20--not 3x5, 6, or 7=12, 15, or 18. Look around a bit for 
the cookie you have the most of.

4x4: it's still easy to peg if you have five of a certain type of cookie 
but it's not always possible to reduce a 4x4 to nothing at all. For 
instance if you have 2 types of cookie represented 5 times and the rest 
2, your first reduction will leave 1 of 1 type of cookie. I believe for 
just about any rectangle above this you can cut it down entirely without 
leaving one of one type of cookie.

Caveat for the rectangles below--I included a lot of analysis for 
getting rid of stuff right away. Take what you feel is necessary.

4x5: If you do not have 4 of each type of cookie then you have at least 
5 of 1(otherwise your total would be <= 4+4+4+4+3 = 19.) Try to line it 
up so that you reduce to a 3x5 unless you have 6 of the greatest number. 
In which case reduce that to 2--in the 4x4 remaining you have 14 squares 
among 4 colors and so you must have 4 of one. Then unless you have a 5-
3-3-3 combination you can reduce even further(the 4 can go.)

4x6: Unless you have a 5-5-5-5-4(which means you've solved it anyway; 
match up the color with 4 in one row and then the 5-colors can all match 
up and you're done if your fingers are quick enough.) You'll have six or 
more of one color, maybe more. Get rid of that(unless there are seven) 
and the resulting 3x6 is easy.

4x7: Since 4x6 can be reduced to 3x6 and you may be in trouble of losing 
anyway just fill in any old row you see.

5x5-7x7: the bigger a rectangle gets, the easier it is to pick a color 
to eliminate a row. However, it's more urgent as well. First you may 
want to note if there is a cookie type you can eliminate exactly--in 
that case you really have control of the situation. Let's look below:

5x5 will have 6 of one cookie type or be very easy to reduce(5-5-5-5-5.)

5x6 will have 7 of one cookie type or be very easy to reduce(6-6-6-6-6.)

5x7 will have 8 of one cookie type or be very easy to reduce(7-7-7-7-7.)

In 5x6 or 5x7 of course it is easier to eliminate a row of 5 but if you 
can get the 6 or 7 by all means do so. In general you want to leave 2 or 
4 of a certain cookie type as even numbers of cookie types are easier to 
get rid of if the structure collapses to a 2-by-x formation.

6x6 will have 8 of one cookie type.

6x7 will have 9 of one cookie type.

7x7 will have 10 of one cookie type.

For these, just go with your gut feeling and try to wipe out the most 
prevalent cookie type UNLESS you are able to count a rarer cookie type 
and know you can get rid of it.

    5.3. USING THE CONTROLS EFFECTIVELY

      5.3.1. THE PIVOT

    The most effective direction I know of is what I call 'the pivot.' 
If you need to move just one block to a certain square to complete a 
row, then you should be able to do so. For instance take the example 
below.

X X X X X X

X 1 X X P X

X X X X X X

X X X X X X

1 1 1 1 X 1

X X X X X X

    The pivot point is marked with a 'P.' Move the cursor there, click 
three times right, then three down. This is the most expeditious way to 
do things; if you can lump the moving and shifting together, it will 
save you time and thought. The rule is, in general:

  1. move the candidate square parallel to the almost-complete row until 
it is on the pivot square, which should be horizontally/vertically 
aligned with the candidate square's original position and destination.
  2. move perpendicularly until the row is complete.

      5.3.2. WALKING OFF THE EDGE

    It's useful. If the next square you need to access is well on the 
other side, go off the edge, especially if it's in the opposite corner. 
But don't pause to see if the distance would be farther that way as that 
will waste the time you generally save by fewer moves. It takes a bit of 
getting used to in order to get your intuition up to scratch.

      5.3.3. THE HALF-LINE-UP

    I often find that, at the start of a level, you have a good deal of 
one certain color. Usually there will be one of it in each column. For 
whatever reason there often turns out to be two of it. Don't worry about 
trying to line up two rows at once, but once one is lined up you should 
see something vaguely resembling a row of the color that's just 
disappeared.  Move horizontally to fix the columns as you need to.

      5.3.4. 'OTHER'

    This section is reserved for later strategies.

  6. GENERAL STRATEGIES FOR THE GREAT BEYOND

    Levels 11-99 echo 1-10 except there is always a random Koopa shell 
placed somewhere. This is very annoying--if you apply the pigeonhole 
principle you now see serious problems with getting stuff done. In 
particular:

2x3 rectangle may have all 5 regular shapes + Koopa shell. In other 
words there is no way to reduce it. The probability here is 
5*4*3*2*1/5^5=120/3125=3.84%. However a 3x4 rectangle can be reduced, 
and having a Yoshi helps.
4x4 rectangle similarly may have 3 each of all 5 regular shapes + Koopa 
shell. That leaves no way to reduce it. Generally speaking if the 
rectangle is big enough you can always find ways to reduce, but if it is 
pretty small you may just be out of luck. Then even if you get rid of 
the Koopa shell you still have the conventional slog to wipe everything 
out and the pieces will probably be going a bit faster.

    6.1. A QUICK(BUT NEVER WINNING) START

    The tactics above(5.1) should apply. 

    6.2. KNOCKING OUT THE KOOPA

    Since you do not have any Koopas dropping from the sides, you need 
to find a Yoshi to wipe it out. In particular you will almost certainly 
need to have a 1-by-something row where the incoming Yoshi can make a 
row with the Koopa and eliminate it. This is especially difficult to do 
because chance will not always be on your side--it may take a few Yoshis 
before you manage to eliminate it. Then there's the matter of waiting 
out the end as in previous levels, which is complicated a bit by the 
less descript colors. As in the general strategies if you have a 1-by-x 
rectangle then the probability you can eliminate a lot of the squares 
and in particular the Yoshi will line up with the shell is quite good--
the next objects form a 2 by x+1 rectangle(i.e. add x+2) and there are x 
slots it can go to where you can place the Yoshi next to it. It's hard 
to get a 1-by-x just when Yoshi appears but if you do you're in great 
shape. See below for example of a 1x6.

N N

1 Y
2 Y
3 Y
4 Y
5 Y
S Y

    If Yoshi appears at any of the 'Y' squares you can move the shell to 
match with it. If he's on an 'N' square there is no way you can do so. 6 
Y's, 2 N's.

  7. TRICKS

    Thanks to gamewinners.com for this one: to get past level 10, go to 
the start menu and then turn the speed to high, turn the music off, go 
to the level selector, and hold up while you select the level.

    (Note: the game's totally impossible without switching the speeed 
back to low/medium after you choose the level. And the three songs you 
can re-choose are different from the 'original' Yoshi.)

    Actually the game tells it to you, but I was too lazy to wait for 
the finishing credits the first few times. Then I gave gamewinners 
credit, played through level 10 one more time, went to fix a snack 
and...voila!

  8. AFTER-LEVEL SKITS

    After completing board 1-10 you see Mario running from the left(he 
always does) to catch a cookie. He gets it and waves to the player.

    After completing board 2-10 you see Mario chase the cookie, catch 
it, and spin around with it as it rolls to the right.

    After completing board 3-10 you see Mario chase the cookie off the 
right side of the screen. It comes back with fangs and double the 
circumference, bouncing towards him. Very Pac-Man.

    After completing board 4-10 you see Mario chase the cookie, which 
bounces off a building to the right and topples Mario on the way back.

    After completing board 5-10 you see Mario run after the cookie, 
which spins like a coin. Yoshi comes from the right and eats it.

    After completing board 6-10 you see Mario run down the hill past the 
cookie.

    After completing board 7-10 you see the cookie roll against the 
building. Mario jumps on it this time and deflates it.

    After completing board 8-10 you see Mario chase the cookie until it 
drops. Mario walks over it, tries to balance, and falls over.

    After completing board 9-10 you see Mario chase the cookie off a 
cliff. It rises back up frowning with a halo on its head, and Mario 
looks to you and shrugs.

    After completing board 10-10 you see Mario and Yoshi run from 
opposite sides to trap the cookie, which smiles when they catch it. Then 
you get the credits scene and the hint to get to level 11 and beyond.

    After completing boards 11-10 through 98-10 you get nothing of note. 
Although the border around the level changes to reflect the new 
characters.

    After completing board 99-10 you get a final list of credits 
including the monster names and so forth.

End of FAQ proper

  9. MY SUGGESTION FOR A NEAT SIDE GAME

    Being the perfectionist I am, I would like to point out this side 
game. First, if you've bought the original game, get an emulator. Save 
the state before starting a board in the first ten levels. Keep trying 
it until you've solved it WITHOUT getting reinforcements. This will not 
always be possible(see 4x4 with 4 colors,) but it's worth a try.

    I suppose you could also write a computer algorithm for this sort of 
thing as well, or at least a cheap C++ game to simulate this sort of 
thing.

    Later I may use some actual level starts as examples of how to play 
this game. Perhaps it may make the later levels completely trivial, as 
the level starts seem to be chosen from a relatively small group.

================================

  10. VERSIONS/CREDITS

0.5.0: sent to GameFAQs.com 8/14/2002 mostly complete. The regular 
errant strategy tweaks I bet and some parts are just plain MIA but oh 
well. I am trying to push myself out of a layoff--this thing was in the 
hamper for two months.

vimm.net for the ROM image which helped me futz with a lot of stuff more 
quickly(replaying a level a few times)
www.gamewinners.com for the hint which allowed me to explore the later 
levels. It's not this excellent site's fault said levels are tedious. 
Before in the spirit of not reading the instruction book 'til I have to 
I just killed the game after 'winning' the first ten levels a few times 
before I actually saw it. But that site is always there for an awesme 
database of codes.