Nubby's Number Factory

Nubby's Number Factory

94 ratings
A Mathematician's Guide to Nubbyology
By weh
Using rigorous mathematical analysis to help you stop losing at Round 40.
7
8
2
4
2
   
Award
Favorite
Favorited
Unfavorite
Basic Nubbyology
This document outlines the fundamental mathematical systems that underlie gameplay in Nubby's Number Factory, independent of specific item or perk interactions.




1. Board Structure

- The board consists of 21 peg slots.
- Pegs have values that are powers of two: 1, 2, 4, 8, 16, 32, ...
- Pegs are popped when Nubby hits them or when a summon or item causes a pop.
- When a peg is popped, it is reduced to half its value (rounded down), and the value popped is added to the score.
- A peg is removed from the board entirely once its value becomes 0. This is called a full pop.




2. Scoring From Pegs

Popping a peg of value `v` gives the following total score (if popped fully down to 0):
Score(v) = v + v/2 + v/4 + ... + 1 = 2v - 1 (since this is a geometric series with ratio 1/2)
This is the maximum score that can be obtained from fully popping a peg.




3. Restocks

A restock occurs when the player's score in a round equals or exceeds the current quota.

Each restock performs the following operations in order:

1. For each empty slot, insert a new peg of value:
value = 2^(floor(round / 5))

2. Perform a single merge pass:
- For every value `v`, if there are at least two pegs of that value, one pair will merge into a single peg of value `2v`.
- Only one merge pass is performed. Merges do not recurse.
- Merge order is not adjacency-based, and pegs can merge even if they are in opposite positions.

3. Award the player 1 coin per restock.

Each restock increases the total peg value on the board, leading to higher quotas.




4. Target Score (Quota) Calculation

The quota is calculated once at the beginning of each round and does not change mid-round.

Let `V_total` be the sum of all peg values on the board, computed using:
FullValue(v) = 2v - 1 (as per section 2)

So,
V_total = sum of FullValue(v_i) for each peg i on the board

Let `multiplier` be the value from the following table based on number of shops visited:
shops = number of Nubby-Mart or Grab-a-Tron visits multiplier = lookup from table below

Shop Visit Count | Multiplier -------------------|------------------ 0 | 1 / 4.25 1 | 1 / 3 2 | 1 / 2.25 3 | 1 / 1.8 4 | 1 / 1.75 5 | 1 / 1.6 6 | 1 / 1.45 7 | 1 / 1.25 8 | 1 / 1.1 9 | 1 10 | 1.1 11 | 1.15 12 | 1.25 13 | 1.5 14 | 1.6 15 | 1.75 16 | 1.95 17+ | 2 * (1.3)^(shops - 17)

Then the target score (quota) is:
Quota = V_total * multiplier

VERY IMPORTANT NOTE - THIS MEANS (as of patch 1.3), YES, GOING TO THE BLACK MARKET DOES NOT TICK UP YOUR QUOTA MULTIPLIER. YOU SHOULD BASICALLY ALWAYS BE GOING TO THE BLACK MARKET.




5. Board Clear Bonus

If all pegs on the board are fully popped (board is empty), a Board Clear occurs, which doubles your total score.

This is a powerful end-of-round reward, and an important consideration for score efficiency.




6. Peg Behavior Summary

Each peg has the following properties:

- Value `v`, where `v` is a power of two.
- Full score potential = `2v - 1`
- When popped, adds `v` to score and changes to `v / 2`
- When `v == 1`, popping removes the peg (full pop)
- Pegs with same value may merge during a restock (once per restock)




7. Restock Value Increase

Each restock increases the total value on the board through:

- Adding new pegs of fixed value per round tier
- Merging existing pegs into higher values

Let `r = round number`, then new peg value on restock is:
new_peg_value = 2^(floor(r / 5))

Over multiple restocks, the board's average peg value increases geometrically, especially if merges are efficient.




8. Game Phases

The game is divided into 4 acts:

- Act 1: Rounds 0–20 (Early Game)
- Act 2: Rounds 21–40 (Midgame)
- Act 3: Rounds 41–60 (Lategame)
- Act 4: Rounds 61–80 (Endgame)

Score expectations, peg values, and quota sizes increase significantly with each phase.




9. Randomness and Determinism

- Peg merge behavior is deterministic unless the board has been shuffled.
- Board shuffles introduce positional randomness, but not value randomness.
- Peg pops are determined by Nubby's physics and bounce angle, which includes slight RNG.

Given the same board state and sequence of actions (with no shuffling), outcomes are deterministic.
Case Study: One-Restock Board Model pt 1.
The following table will show you the board state assuming you only ever manage a single restock, with no items or anything else. You can consider this toy model a "lower bound" and analyze it for behaviors.

Round
Peg Summary
Empty Slots
Shops Visited
Multiplier
Board Value
Quota
2
6x2, 5x4
10
0
0.23529
53
12
3
5x2, 4x4, 2x8
10
0
0.23529
73
17
4
6x2, 2x4, 2x8, 1x16
10
0
0.23529
93
22
5
5x2, 3x4, 1x8, 2x16
10
0
0.23529
113
27
6
1x2, 8x4, 2x8, 1x32
9
1
0.33333
152
51
7
2x2, 3x4, 5x8, 1x16
10
1
0.33333
186
62
8
2x2, 4x4, 2x8, 3x16
10
1
0.33333
220
73
9
1x2, 3x4, 2x8, 3x16, 1x32
11
1
0.33333
263
88
10
2x2, 1x4, 4x8, 1x16, 2x32
11
1
0.33333
298
99
11
3x2, 2x4, 1x8, 4x16, 1x32
10
2
0.44444
324
144
12
1x2, 2x4, 2x8, 2x16, 2x32
10
2
0.44444
375
167
13
3x2, 3x4, 3x8, 1x16, 1x64
10
2
0.44444
435
193
14
3x2, 4x4, 3x8, 1x32, 1x64
9
2
0.44444
504
224
15
4x2, 3x4, 2x8, 1x16, 2x64
9
2
0.44444
573
255
16
2x2, 2x4, 2x8, 2x16, 2x32, 1x64
10
3
0.55556
606
337
17
2x2, 1x4, 2x8, 2x16, 1x32, 2x64
11
3
0.55556
659
366
18
3x2, 2x4, 2x8, 1x16, 2x32, 1x64
10
3
0.55556
712
396
19
2x2, 2x4, 2x8, 1x16, 1x32, 2x64
11
3
0.55556
779
433
20
2x2, 2x4, 3x8, 1x16, 2x32, 1x64
10
4
0.57143
839
479
21
1x8, 6x32, 2x64, 2x128, 1x256
9
4
0.57143
1668
953
22
1x8, 1x16, 4x32, 3x64, 1x128, 2x256
9
4
0.57143
1956
1118
23
1x8, 5x32, 3x64, 2x128, 1x512
9
4
0.57143
2244
1282
24
1x8, 1x16, 5x32, 3x64, 1x128, 1x256, 1x512
8
4
0.57143
2531
1446
25
1x8, 1x16, 5x32, 3x64, 2x128, 1x256, 1x512
7
4
0.57143
2786
1592
26
1x8, 1x16, 7x64, 1x128, 2x256, 1x512
8
5
0.62500
3235
2022
27
1x8, 1x16, 5x64, 4x128, 2x512
8
5
0.62500
3747
2342
28
1x8, 1x16, 5x64, 2x128, 2x256, 1x1024
9
5
0.62500
4260
2662
29
1x8, 1x16, 1x32, 5x64, 2x128, 1x256, 1x512, 1x1024
8
5
0.62500
4835
3022
30
1x8, 1x16, 1x32, 5x64, 2x128, 2x256, 1x512, 1x1024
7
5
0.62500
5346
3341
31
1x8, 1x16, 1x32, 6x128, 1x256, 2x512, 1x1024
8
6
0.68966
6243
4306
32
1x8, 1x16, 1x32, 4x128, 4x256, 2x1024
8
6
0.68966
7267
5012
33
1x8, 1x16, 1x32, 4x128, 2x256, 2x512, 1x2048
9
6
0.68966
8292
5719
34
1x8, 1x16, 1x32, 1x64, 4x128, 2x256, 1x512, 1x1024, 1x2048
8
6
0.68966
9443
6512
35
1x8, 1x16, 1x32, 1x64, 4x128, 2x256, 2x512, 1x1024, 1x2048
7
6
0.68966
10466
7218
36
1x8, 1x16, 1x32, 1x64, 1x128, 5x256, 1x512, 2x1024
7
7
0.80000
12258
9806
37
1x8, 1x16, 1x32, 1x64, 5x256, 3x512, 2x2048
7
7
0.80000
14050
11240
38
1x8, 1x16, 1x32, 1x64, 1x128, 4x256, 3x512, 1x1024, 1x2048
7
7
0.80000
15842
12674
39
1x8, 1x16, 1x32, 1x64, 4x256, 3x512, 2x1024, 1x2048
7
7
0.80000
17634
14107
40
1x8, 1x16, 1x32, 1x64, 1x128, 3x256, 3x512, 1x1024, 2x2048
7
8
0.90909
19426
17660
Case Study: One-Restock Board Model pt 2
41
1x8, 1x16, 1x32, 1x64, 1x128, 6x512, 2x1024, 1x2048
6
8
0.90909
29665
26968
42
1x8, 1x16, 1x32, 1x64, 1x128, 3x512, 3x1024, 2x2048
7
8
0.90909
32738
29762
43
1x8, 1x16, 1x32, 1x64, 1x128, 1x256, 4x512, 2x1024, 2x2048
6
8
0.90909
36322
33020
44
1x8, 1x16, 1x32, 1x64, 1x128, 1x256, 3x512, 2x1024, 2x2048
7
8
0.90909
40520
36836
45
1x8, 1x16, 1x32, 1x64, 1x128, 5x512, 1x1024, 1x2048, 1x4096
7
9
1.00000
43489
43489
46
1x8, 1x16, 1x32, 1x64, 1x128, 7x1024, 1x2048, 1x4096
6
9
1.00000
49633
49633
47
1x8, 1x16, 1x32, 1x64, 1x128, 4x1024, 4x2048, 1x4096
6
9
1.00000
55778
55778
48
1x8, 1x16, 1x32, 1x64, 1x128, 3x1024, 2x2048, 4x4096
7
9
1.00000
62945
62945
49
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 4x1024, 1x2048, 3x4096
6
9
1.00000
69090
69090
50
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x1024, 3x2048, 3x4096
7
10
1.10000
83425
91768
51
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 6x2048, 1x4096, 1x8192
6
10
1.10000
95713
105284
52
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x2048, 4x4096, 1x8192
6
10
1.10000
108001
118801
53
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 4x2048, 1x4096, 3x8192
6
10
1.10000
120289
132318
54
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x2048, 3x4096, 3x8192
6
10
1.10000
132577
145835
55
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 4x2048, 2x4096, 4x8192
6
11
1.15000
157154
180727
56
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 5x4096, 1x8192, 1x16384
7
11
1.15000
185824
213698
57
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 1x2048, 4x8192, 3x16384
5
11
1.15000
206305
237251
58
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x4096, 3x8192, 2x16384
6
11
1.15000
230881
265513
59
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 4x4096, 2x8192, 3x16384
6
11
1.15000
255458
293776
60
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x4096, 2x8192, 3x16384
7
12
1.25
255458
319322
61
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 5x8192, 1x16384, 2x32768
7
12
1.25
312802
391002
62
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 1x4096, 4x16384, 2x32768
5
12
1.25
370144
462680
63
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x8192, 3x16384, 2x32768
6
12
1.25
411105
513881
64
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 4x8192, 2x16384, 3x32768
6
12
1.25
460257
575321
65
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x8192, 2x16384, 4x32768
7
12
1.25
509410
636762
66
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 5x16384, 2x32768, 2x65536
7
13
1.50
624098
936147
67
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 1x8192, 4x32768, 3x65536
5
13
1.50
738784
1108176
68
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x16384, 3x32768, 3x65536
6
13
1.50
820705
1231058
69
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 4x16384, 2x32768, 4x65536
6
13
1.50
919009
1378514
70
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x16384, 2x32768, 5x65536
7
13
1.50
1017314
1525971
71
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 5x32768, 2x65536, 1x131072
7
14
1.60
1246690
1994704
72
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 1x16384, 3x65536, 2x131072
5
14
1.60
1476064
2361702
73
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x32768, 3x65536, 2x131072
6
14
1.60
1639905
2623848
74
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 4x32768, 2x65536, 3x131072
6
14
1.60
1836513
2938421
75
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x32768, 3x65536, 3x131072
7
14
1.60
2033122
3252995
76
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 5x65536, 2x131072, 1x262144
7
15
1.75
2491874
4360780
77
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 1x32768, 4x131072, 2x262144
5
15
1.75
2950624
5163592
78
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x65536, 3x131072, 2x262144
6
15
1.75
3278305
5737034
79
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 4x65536, 2x131072, 3x262144
6
15
1.75
3671521
6425162
80
1x8, 1x16, 1x32, 1x64, 1x128, 1x512, 3x65536, 2x131072, 3x262144
7
16
1.95
4064738
7926239
Basic Peg Popping
Nubbyology II: Peg Popping and Score Math


Introduction
This guide provides a rigorous mathematical breakdown of peg popping mechanics and their role in score generation, item/proc activation, and viability in building toward quota-clearing runs. This section builds on the foundational mechanics described in Basic Nubbyology, focusing specifically on pop-based systems.



Peg Popping: Core Mechanics

Each peg begins with a value `v`, where `v` is a power of two (1, 2, 4, 8, 16, ...).

Each time a peg is popped:
  • You gain score equal to its current value.
  • The peg’s value is halved (rounded down).
  • If its value reaches 0, the peg is removed from the board ("full pop").

The total score from a full pop of a peg is the sum of its value over successive halvings:

FullPopScore(v) = v + v/2 + v/4 + ... + 1 = 2v - 1

This formula gives the upper bound on the score that can be gained from a single peg, assuming it is popped all the way to 0.



Score Contribution From Pegs

To estimate the total board score potential:
  • Let `v_i` be the value of each peg on the board (up to 21 pegs).
  • For each peg, compute `2v_i - 1`.
  • Sum these to get `TotalFullBoardScore`.

TotalFullBoardScore = sum(2v_i - 1 for all pegs i)

This gives the theoretical maximum score for a round, before any multipliers or bonuses are applied.



Board Clear Viability and Quota Matching

From the quota multiplier table, we know that:

Quota = TotalFullBoardScore * multiplier

If a player fully clears the board, they receive a 2x board clear bonus:

BoardClearScore = TotalFullBoardScore * 2

Therefore:

BoardClearScore >= Quota (pre-endless mode)

This implies:
  • Any build that consistently fully clears the board will, by definition, meet or exceed the quota (outside of endless mode).
  • Even as the multiplier increases with more shop visits, the board clear bonus ensures survival is possible with perfect clearing.
  • Builds that do not consistently clear the board must rely on supplemental score sources (items, perks, etc.) to remain viable.



Pops as a Resource

Each pop is also a resource used to:
  • Trigger items (e.g., every 5 or 10 pops)
  • Trigger perks (e.g., every 10, 15, or 20 pops)
  • Chain summoner effects
  • Contribute to restock chains

Thus, optimizing for total pops per round has benefits beyond score alone. For example:
More pops → more item triggers → more pops → restock → multiplier → new pegs → more score

The game strongly rewards recursive pop loops, even when individual pop values are low.



Pop-Triggered Items and Perks

Below is a non-exhaustive list of systems that rely on peg pops directly:

Pop-Triggered Items:
  • Flutty: every 2 pops
  • Cactus: every 5 pops
  • Squirmy: every 5 pops
  • Pants: every 10 pops
  • Crazy Straw: every 10 pops
  • Ton of Feathers: on full pop

Pop-Triggered Perks:
  • Drainer: every 10 pops
  • House'O'Cards: every 15 pops
  • Eggy: every 20 pops

Each of these may increase score directly or indirectly by generating more pops, spawning summons, or force-triggering other items.



Summary

  • Popping is both the core scoring mechanism and the primary trigger system for a wide range of effects.
  • Each full pop provides a score of 2v - 1, meaning that even low-value pegs contribute significantly over time.
  • Consistent board clears ensure quota matching through the 2x multiplier, making them a mathematically viable long-term strategy.
  • Non-clear builds must compensate with external score generation to remain sustainable.
  • Pop triggers and recursion are central to score acceleration, restock chaining, and late-game viability.

This concludes the core math on peg popping. Let's get into some case studies.
Case Study: House of Cards + Pants Loop


This section models the well-known "House of Cards Pants Loop", where Pants is used to recursively pop the board via repeated triggers, enabled by the House'O'Cards perk.

Mechanics Recap
Pants:
  • Triggers every 10 pops.
  • Each Pants item pops 1 random peg.
  • That peg, when popped, gives 2x score (unupgraded) or 4x score (upgraded).
  • Each Pants trigger adds exactly one pop per Pants item.

House'O'Cards Perk:
  • Triggers every 15 pops.
  • Force-triggers all "X pegs popped →" items, including Pants.
  • Multiple copies of the perk stack: if you have `h` copies, it will retrigger Pants `h` times every 15 pops.

Important Notes:
  • Pants do not trigger individually; all Pants activate together at every 10-pop milestone.
  • Each Pants activation results in `p` pops, where `p` is the number of Pants equipped.
  • All of those pops count toward the next pop thresholds (10, 15, etc.).

The Loop Setup

Let:
p = number of Pants h = number of House'O'Cards perks

Trigger chain:

1. Reach 10 pops → all `p` Pants activate → +`p` more pops
2. Reach 15 pops → all House'O'Cards perks activate → force-trigger Pants `h` times
3. Each forced Pants trigger → `p` additional pops
4. These forced pops count toward next 10/15 thresholds

This creates a pop loop.

Loop Condition

To go infinite (clear the entire board without further player input), the loop must produce more pops per cycle than it consumes.

Let’s compute pops produced per cycle:

At pop 10:
+ p pops (Pants activates naturally)

At pop 15:
+ h * p pops (h copies of House'O'Cards → each triggers Pants once → p pops each)

Total pops gained from one 15-pop cycle:

delta_pops = p + (h * p) = p * (1 + h)

Pops consumed to reach the next cycle: 15

To go infinite:
p * (1 + h) >= 15

Minimum Required Values

To find the minimum number of Pants and House'O'Cards to go infinite, solve:

p * (1 + h) >= 15

Try integer combinations:

Case 1: 5 Pants
5 * (1 + h) >= 15 → (1 + h) >= 3 → h >= 2
Infinite at 5 Pants and 2 House'O'Cards

Case 2: 3 Pants
3 * (1 + h) >= 15 → (1 + h) >= 5 → h >= 4
Infinite at 3 Pants and 4 House'O'Cards

Case 3: 7 Pants
7 * (1 + h) >= 15 → (1 + h) >= ~2.14 → h >= 2
Infinite at 7 Pants and 2 House'O'Cards

Conclusion

The loop becomes infinite when:

p * (1 + h) >= 15

This models how many total pops are generated per loop cycle versus how many are required to trigger the loop again.

Once this condition is met:
  • The board will be cleared through recursive Pants activations.
  • The pop count will increase geometrically.
  • The loop terminates only when the board is empty.
  • This guarantees a Board Clear and 2x multiplier.

If you are having trouble beating the game, this is a decent strategy to fish for. Pants is a good item on its own.
Case Study: Stacked Disguise Glasses


This section analyzes the behavior of Disguise Glasses, particularly when stacked or upgraded, and how they can be used to facilitate recursive board-clearing loops.

Mechanics Recap

Disguise Glasses:
  • Trigger condition: A peg of the current highest value is popped.
  • Effect (base): Pops 2 random pegs.
  • Effect (upgraded): Pops 2 random pegs, then repeats this action once more (total of 4 random pegs popped).
  • Stacking multiple copies results in multiple triggers on a single highest-value peg pop.

Key Mechanics

  • Only the highest-value peg(s) on the board will trigger Disguise Glasses.
  • If there are multiple pegs tied for the highest value, popping any of them will also trigger the item.
  • Random peg pops caused by the glasses may pop other pegs tied for highest, recursively triggering the effect again.
  • Pops occur in sequence, and if a peg is reduced to tie for highest (e.g., 1024 → 512), it may then contribute to additional triggers.

Frame-by-Frame Example

Suppose the board contains:
1x 1024 (initial highest) 3x 512 others < 512

1. You pop the 1024:
- It becomes 512
- Glasses trigger: pop 2 random pegs

2. If one of those pops is a 512:
- That 512 is now a "highest" peg
- It triggers Glasses again

3. This chain can continue as long as:
- Random pops hit pegs of highest value
- Pegs remain on the board to be targeted

Statistical Modeling

Let:
g = number of Disguise Glasses u = number of upgraded Disguise Glasses r = number of pegs currently on the board h = number of highest-value pegs at that moment

Each Disguise Glasses trigger pops:
2 pops per unupgraded item 4 pops per upgraded item Total pops per trigger = 2g + 2u (since upgraded triggers the effect twice)

Assuming random targeting and uniform peg distribution (this is not the structure of a board, but modeling an arbitrary board at any given round is difficult, so for now this will have to do):

Probability that a given pop hits a highest-value peg:
P(hit highest) = h / r

Expected number of new highest-value peg pops per trigger:
E(highest hits) = (2g + 2u) * (h / r)

This value determines whether the chain will sustain itself.

Recursive Chain Condition

To sustain a chain or move toward a full clear:

E >= 1 → (2g + 2u) * (h / r) >= 1

If the expected number of highest-value peg pops per Disguise Glasses trigger is ≥ 1, the chain is self-sustaining.

Example 1: 2 Unupgraded Glasses, h = 2, r = 16
E = (2 * 2) * (2 / 16) = 4 * 0.125 = 0.5 → Chain will likely die out

Example 2: 3 Upgraded Glasses, h = 3, r = 12
E = (2 * 3 + 2 * 3) * (3 / 12) = 12 * 0.25 = 3.0 → Chain is self-sustaining, possibly board-clearing

Practical Observations

  • Disguise Glasses are most effective on boards with multiple high-value pegs of the same value.
  • Their chain potential weakens rapidly as high-value pegs are removed or fall below the top tier.
  • Chains tend to "fizzle out" as the board becomes sparse or unevenly distributed.
  • Upgraded copies are significantly stronger, as each provides twice the chance of recursion.
  • Even in non-infinite scenarios, stacked Glasses can rapidly thin out the board and contribute toward full pops and restock cycles.

Summary

To loop with Disguise Glasses: (2g + 2u) * (h / r) >= 1

This equation defines the condition for expected recursion. While it is not a deterministic loop like Pants + House'O'Cards, a well-balanced board and a strong stack of Disguise Glasses can reliably clear large portions of the board and maintain value density in a score-efficient way.

Ideal use cases include:
  • Midgame boards with merged values and few or no orphans
  • Item slot builds focused on reactive chains and pop synergies
  • Boards aiming for restock acceleration or combo multiplier setups
Basic Flat Points
This section examines flat point items - those that add fixed score values independent of peg values - and analyzes their effectiveness throughout game progression.

Flat Point Items: Definition and Examples

Flat point items provide score through fixed additions rather than through peg value manipulation. Primary examples include:

- Cheese House: +300 points when Nubby dies (+3000 after round 30)
- Poop Butt: +15 points × number of restocks when Nubby hits their first peg
- Finger Puppet: Points equal to bounce count when Nubby bounces
- Flutty: +3 points for each trigger (every 2 pegs popped)
- Squirmy: +5 points for each trigger (every 5 pegs popped)

Mathematical Model: Early Game Economics

To understand the effectiveness of flat point items, we need to compare their contribution to the round's quota. Let:

- FP(r) = Flat point contribution in round r
- Q(r) = Quota for round r
- E(r) = Effectiveness ratio =
FP(r) / Q(r)

Early Game Analysis (Rounds 0-20)

From the One-Restock Model, we can see quota progression in early rounds:

Round
Quota
Cheese House Contribution
Effectiveness Ratio
2
12
300
25.00
5
27
300
11.11
10
99
300
3.03
15
255
300
1.18
20
479
300
0.63

As shown, a single Cheese House can single-handedly clear quotas in the very early game, with E(r) > 1 indicating the flat points exceed the quota. Even at round 15, one flat point item contributes significantly to quota completion.

Mid-Game Transition (Rounds 21-40)

The effectiveness ratio declines sharply as quotas scale:

Round
Quota
Cheese House Contribution
Effectiveness Ratio
25
1,592
300
0.19
30
3,341
3,000
0.90
35
7,218
3,000
0.42
40
17,660
3,000
0.17

Note that at round 30, Cheese House's contribution jumps to 3,000, temporarily boosting its effectiveness, but the exponential quota growth quickly outpaces this increase.

Late Game Collapse (Rounds 41+)

By late game, flat points become increasingly irrelevant:

Round
Quota
Cheese House Contribution
Effectiveness Ratio
50
91,768
3,000
0.03
60
319,322
3,000
0.01
70
1,525,971
3,000
0.002
80
7,926,239
3,000
0.0004

Even with multiple flat point items, their contribution becomes negligible in the endgame. The E(r) approaches zero as board values and quotas increase exponentially.

The Restock Acceleration Problem

Beyond their diminishing returns, flat point items create a secondary issue: they accelerate restock cycles too quickly without corresponding board-clearing capability.

Consider what happens when flat points trigger early restocks:

1. Premature restock occurs before efficiently popping existing pegs
2. New pegs spawn with base values of 2^(floor(round/5))
3. These merge into higher-value pegs
4. Board value increases, leading to higher quotas next round
5. Flat points become even less effective against these elevated quotas

This creates a "debt spiral" where quota requirements outpace your ability to clear the board.

Mathematical Proof: The Exponential Gap

The fundamental mathematical issue is the mismatch in growth rates:

- Quota growth: Exponential (powers of 2)
- Flat point contribution: Linear or at best quadratic (for cumulative items like Flutty)

Specifically, board value in round r under the One-Restock Model approaches:
V(r) ≈ C₁ × 2^(floor(r/5))

Where C₁ is a constant based on board composition.

Meanwhile, flat point contribution scales at best as:
FP(r) ≈ C₂ × r²

For any positive constants C₁ and C₂, there exists a round number r* such that:
∀r > r*: V(r) > FP(r)

This guarantees that flat points eventually become irrelevant, regardless of how many you stack.

Strategic Implications

Despite their late-game falloff, flat point items serve important strategic roles:

1. Early Economy Building: They can help you survive the first 20 rounds with minimal board-clearing ability
2. Transition Resource: The extra coins from early restocks let you purchase more powerful items
3. Supplemental Scoring: They can provide the "last push" to reach quota when your board-clearing falls just short


The optimal approach is a transition strategy:

- Rounds 0-20: Utilize flat point items to build economy while developing your long-term plan.
- Rounds 30-40: Finish transitioning to your main strategy and perk loadout.
- Rounds 40+: Phase out and sell flat point items in favor of multipliers and loop enablers

Conclusion

Flat point items follow the same mathematical pattern as many game resources: powerful early, negligible late. Even in the pessimistic One-Restock Model, they serve a critical function in jump-starting your economy before recursive strategies become viable.

The key insight: flat points are tools for acceleration, not sustenance. Recognize their declining effectiveness curve and plan your transition to sustainable pop loops before reaching the "flat point cliff" around rounds 30-40.
Case Study: Cheese House + Trophy Perk
This section examines a powerful but temporary economic exploit I have termed the "YIMBY Combo" that leverages specific perk-item interactions to generate unprecedented restock chains early in the game.

Core Components

1. Cheese House (item): Provides +300 points when Nubby dies (+3000 after round 30). Poop Butt also works in this role, but less efficiently, in exchange for being actually useful after this round.
2. Trophy (perk): 50% chance to force-trigger the item in slot 3 when passing round goal

Trigger Mechanics

The YIMBY Combo works because of a critical recursive interaction:

- Each "pass round goal" event gives Trophy a 50% chance to trigger
- When Trophy triggers, it force-activates Cheese House
- Force-activated Cheese House gives +300/+3000 points
- These points can immediately trigger another quota clear
- Each new quota clear gives another 50% chance for Trophy to trigger

This creates a probabilistic chain reaction that can generate multiple restocks in rapid succession.

Mathematical Model: Chain Explosion

Let's define:
P = 0.5: probability of Trophy triggering on a single quota clear
Q(r): Quota in round r
CH = 300: Points from the Cheese House.

The critical insight: each Cheese House activation can trigger multiple quota clears, and each quota clear gives Trophy an independent chance to trigger.

For example, at Round 9 with a quota of 88:
- Cheese House gives +300 points
- This clears the quota 3 times (⌊300 ÷ 88⌋ = 3)
- Each quota clear gives Trophy a 50% chance to trigger
- Probability of at least one trigger in 3 attempts =
1 - (1-P)^3 = 1 - 0.5^3 = 0.875
(87.5%)

With an 87.5% chance of continuation, the chain becomes explosive:

Stage
Cheese House Activations
Quota Clears
Probability of Continuation
Initial
1
3
87.5%
After 1 repeat
2
6
98.4%
After 2 repeats
4
12
99.97%
After 3 repeats
8
24
~100%

The expected number of Cheese House activations follows:
E(activations) = 1/(1-p)

With p = 0.875, E(activations) = 8, but this is a conservative estimate because each activation can trigger multiple new activations, causing the chain to have an increasing probability of continuation as it accelerates.

This creates an exponential cascade that can easily reach hundreds or thousands of restocks, effectively maxing out at the game's internal restock limit of 9,999, though practical observations show 300-400 restocks being common.

Optimal Timing Windows

The YIMBY Combo has two primary viable windows:

Window 1: Round 9
From the One-Restock Model, we know the quota at Round 9 is approximately 88 points.
Cheese House provides 300 points, which is 3.4× the quota.

For the combo to chain:
- First restock occurs naturally
- With 300 points and quota of 88, you clear the quota 3 times
- Each quota clear gives Trophy a 50% chance to trigger
- If any of these trigger, Cheese House activates again
- Each new +300 creates 3 more quota clears
- Each of these gives more chances for Trophy to trigger
- This creates an exponential cascade that can reach hundreds if not thousands of restocks

Window 2: Round 30
At Round 30, the quota is approximately 3,341 points.
Cheese House now provides 3,000 points (.9x of quota), while an upgraded Cheese House provides 7,500 points (2.24x of quota).
At this point, you may also have multiple Trophy Perks.

This creates two scenarios:
1. If your score was already near quota, Cheese House can push you over
2. If you've accumulated even a small score buffer, each activation can trigger another restock

Economic Impact Model

Given the potential for hundreds of restocks, let's model a more realistic YIMBY chain at Round 9:

Restock Range
Board Evolution
Coins Gained
1-10
Board quickly fills with level 2-8 pegs
+10
11-50
Extensive merging creates level 16-64 pegs
+40
51-100
High-value pegs (128-512) dominate
+50
101-200
Extremely high-value pegs form (1024+)
+100
201-300
Board stabilizes with maximum value pegs
+100

Total: +300 coins (effectively maxing out at 99), board value equivalent to round 30 or so.

This represents a massive economic discontinuity, essentially skipping 20 rounds of progression in a single turn.

Quotas After YIMBY

The immediate consequence is a catastrophic jump in quota for the following round:

Normal Round 10 quota = 99 Post-YIMBY Round 10 quota ≈ 5,000-10,000

This represents a 50-100× increase in quota requirements, which renders flat point items completely irrelevant overnight. Even Cheese House's +300 becomes less than 6% of a single quota clear, effectively forcing an immediate transition to mid-game strategies.

Board Composition Effects

The rapid merges create a significantly different board composition:
- A few extremely high-value pegs, surrounded by
- Many relatively low-value pegs
- Higher value variance
- More challenging board to clear efficiently

Strategic Implications and Limitations

The YIMBY Combo creates a "one-time economic boost" with specific trade-offs:

Advantages:
- Massive early-game coin acceleration
- Potential to hit the 99 coin cap extremely early
- Access to the cafe and items much earlier, allowing you to go all in on Strawberry and Kebab strategies ASAP

Disadvantages:
- Immediate obsolescence of flat point strategies
- Forces transition to value-based strategies early
- Creates extreme quota pressure in subsequent rounds

Accessibility and Reliability

The minimum requirements for attempting the YIMBY Combo are:
- Reaching Round 9 with Cheese House in slot 3
- Obtaining Trophy as your first perk
- Having a low enough quota that Cheese House will clear it at least twice, but ideally more than that. The lower, the better.

Additionally, the round 30 variation is an excellent way to rescue a flagging economy should you have the tools to access it and a low enough quota.

Conclusion

This combo demonstrates the ultimate form of flat point exploitation - using a single flat point item to create such an overwhelming economic advantage that flat points themselves become obsolete. It's the mathematical equivalent of burning the ladder you climbed up on, forcing an immediate transition to entirely different strategies.

The YIMBY is less a sustainable strategy and more an all-in economic reset button that catapults the player from early to mid-game in a single round. Players should decide whether they want to experience the game's natural progression curve or accelerate directly to mid-game conditions with maxed economy but exponentially harder quotas.
WIP
...
14 Comments
DJUNGELSKOG 2 Jun @ 4:08am 
you helped me get to round 41:steamhappy:
Mitrimind 1 Jun @ 9:10am 
Seal + Tardigrade + Alien might be a good case study. Was able to get googol points involving these items.
Mitrimind 1 Jun @ 9:05am 
Disguise glasses actually becomes more effective if you fully pop pegs and is guarantee to clear if you get down to one or two pegs.
RandomZeugma 1 Jun @ 2:29am 
Unfortunately the YIMBY combo is impossible to trigger today, as the Trophy perk has been updated to grant three non-probabilistic activations at the first restock and the first restock only, which makes the setup described deterministic, with only 13 restocks.
Still, a very interesting guide, you should probably look at updated asbestos next, the multiplier properties of insulation can make for more easily scalable endless builds.
midJoaKz 17 May @ 10:15pm 
Astounding, amazing work :gnomechild:
Caius Cosades 16 May @ 1:04pm 
all this arithmetic and you cant account for Nubbys incessant need to halt all horizontal momentum
Cooper 5 May @ 5:02pm 
Are you mentally stable? These are too many numbers for one mentally stable person to use.
sand_020 2 May @ 8:42am 
incredible work. gonna study this both for nubby knowledge and guidance on how to write technical reports
Sommy 27 Apr @ 9:38am 
i did read this . what the fuck . very well done
Tactical Tac0 20 Apr @ 9:06am 
Excellent analysis! I am curious what the math is on halving strategies, I haven't been able to make any builds with halving work thus far