Four bags each, 27 feet, cancellation scoring to 21 — and a bag with a fast side and a slow side.
You0
Them0
Frame1
Bags4 / 4
Your pitch
Pick your aim, your power, your loft — and which side of the bag lands down.
Face down
Power is an offset from the throw that lands on the hole at the
loft you have chosen, so zero on every slider is a shot at the hole and the sliders stay
usable at any loft. Negative lands short — which is what the fast side wants.
Suede down. Lower friction, so the bag keeps sliding after it lands.
Predicted landing appears here.
No frame completed yet.
The board is empty.
Games 0
Game seed 0 — the same seed replays the same game exactly.
Rules, and where every number came from
Cornhole has two rule-making bodies and they do not agree everywhere. This
game cites both: the ACO (American Cornhole Organization, Official Rules
for the Sport of Cornhole, which numbers its articles) and the ACL
(American Cornhole League, Rules & Regulations and Equipment, Bags &
Boards, which names its rules). Where they differ, the choice is stated.
The board and the court
Playing surface 48 × 24 in (ACO 1.1 allows
47.5–48 × 23.5–24; ACL allows 47.75–48.25 ×
23.75–24.25 — 48 × 24 is inside both).
Hole 6 in diameter, centre 9 in from the top
edge, centred side to side. Both bodies agree, and the ACL says it twice in two
different forms that also agree: its equipment page puts the hole’s edge
6 in from the top, and 6 + 3 = 9.
Back of the deck 12 in above the ground (both bodies, exactly).
Front of the deck 2.8 in — that is the ACL’s figure;
ACO 1.1 says 3–4 in and gives no deck angle. The ACL also publishes the
angle, 11° ± 0.5, which makes one of its four
numbers redundant and therefore checkable: asin((12 − 2.8)/48) =
11.05°, inside its own tolerance. The ACO’s range implies
9.59–10.81°, so the two bodies overlap only in 10.50–10.81° and this
game follows the ACL, whose numbers are self-consistent.
Boards 27 ft apart, front edge to front edge, ±½ in
(ACO 1.5 and ACL, identically). The widely repeated “33 feet” is in
neither rulebook, and neither body publishes a shorter or junior distance at all.
Pitcher’s box 4 ft × 3 ft beside each board, and
the front edge of the board is the foul line (ACO 1.5; ACL says the
same). In singles both players stay in one lane for the whole game (ACO 4.1).
The bag — and the two sides
6 × 6 in, 16 oz, 1.3 in thick, filled
with plastic resin (ACO 1.3 mandates “plastic resin/beads which will not
breakdown”; ACL gives the same nominal figures). Whole corn is not legal in either
body’s competition.
The two-sided bag is in the rules — but only in the ACL’s.
Its Set Uniformity rule requires “the same design and colour for each
corresponding side on all four bags” and its own worked example is “a gray
suede side and a red duck cloth side”. So a bag with two different materials on its
two faces is explicitly legal and explicitly regulated. The ACO rulebook does not
mention bag faces anywhere (all 786 lines searched).
Which side is faster, and by how much, is in no rulebook. Neither body
prints a coefficient of friction, or the words “fast” or “slow”.
So the mechanic is rules and the ordering is practice, and every friction number in this
game is reconstructed — listed below.
One bound is implied: ACO 1.1 requires a surface “not so slippery
that the bags slide back down the platform”, and on an 11.5° deck that needs
μ ≥ tan 11.5° = 0.2035. Both faces
here clear it.
Why the bag has to deform at all
A rigid 6-inch square cannot pass through a 6-inch round hole: the largest
circle it fits through is its own inscribed circle, exactly 6.000 in, and its
8.485 in diagonal stands outside. That holds for every legal bag against every legal
hole, at every tolerance corner. So every three-point score in cornhole is a
deformation event, and a rigid-body simulation of this sport cannot score at all.
The bag here is a limp resin-filled envelope whose contact fraction grows with impact
speed; a hard landing spreads it, which makes it harder to drop through the hole, and
a soft arrival keeps it compact and drops in. That is what the loft control is really for.
Scoring
Through the hole — a cornhole — 3 points
(ACO 4.1). ACO 3.15 defines it as any bag that “passes through the
cornhole board hole at any time within the frame”, which is a statement
about time and not about cause: a bag knocked in by another bag scores 3.
Resting on the surface — a woody — 1 point
(ACO 4.1), judged “at the conclusion of the frame” (ACO 3.15).
A bag hanging in the hole is 1 point, stated outright by the ACL
(Bag Values). The ACO reaches the same answer only by implication.
Anything else is a dead bag, 0 — including a bag that touched the
ground before the board, and a bag resting on the board while touching the ground
(ACO 3.15; ACL prices it at 0 explicitly).
Cancellation scoring: “the points of one player cancel out the
points of their opponent… only one player/team can score in each frame”
(ACO 4.1). The rulebook’s own example: 7 − 2, so red scores 5.
First to reach or exceed 21 at the end of a frame wins, and there is
no need to win by two (ACO 4.2).
There is no bust rule in either body. The ACL states it positively:
“A player or team is not penalized for achieving a score over 21 points.”
ACO 4.2’s “reach or exceed” says the same. Busting, and
going back to 15, is a recreational variation.
The side that scored leads the next frame; on a wash the previous leader keeps
first pitch — it does not alternate (ACO 3.4; ACL Wash/Push).
A frame is scored 5 seconds after the last bag lands, and a bag that
moves after that is scored where it was (ACO 3.7). That is the settle window this
game integrates.
What is reconstructed
No rule in this game is reconstructed. Every scoring, foul, turn-order and
win rule above carries an ACO article number or a named ACL rule. What is reconstructed is the
physics, because neither rulebook contains a single coefficient of friction,
restitution, drag or compliance — they specify equipment and procedure only. Thirteen
numbers, all of them mine:
μ of the fast (suede) face, 0.26
μ of the slow (duck cloth) face, 0.42
the assignment of “fast” to suede (the ACL names both materials and ranks neither)
the spread velocity 4.20 m/s, at which the contact patch saturates
the resting contact fraction, 0.55
the deformation-friction coupling, 0.35
the bridging threshold Λ* = 1.35, which decides whether a bag straddling the rim hangs or drops
restitution against the board, 0 (the bag does not bounce; the behaviour is sourced, the number is not)
the drag coefficient, 1.1
bag-on-bag restitution 0.05 and friction 0.55
where inside the 4 × 3 ft box the player stands (the box is a rule; the spot is not)
the release height, 0.62 m
the three opponent skill levels
One figure is derived rather than chosen, and it is worth stating because it came
out right: modelling the flat-laid bag as a 6-inch square whose corners are pulled in to the
ACL’s minimum 7.5-inch diagonal gives a footprint of 34.79 in², and
16 oz in 34.79 × 1.3 in³ is a bulk density of
0.612 g/cm³ — squarely inside the 0.50–0.65 band for
bulk plastic resin pellets. The ACL’s four independent bag numbers are mutually
consistent, and nothing had to be invented to reconcile them.
A related tension, recorded rather than resolved: ACO 1.3 caps the corner-to-corner
span at 8.25 in, and a sharp-cornered 6-inch square measures 8.485 in
diagonally — so the ACO’s own rules require a bag whose corners are pulled in.
But ACO 1.3 also caps corner rounding at 0.5 in, while a 7.5-inch diagonal
implies 1.19 in. The diagonal therefore comes from the fill pillowing a sewn square, not
from cutting the corners, and the footprint used here is a flat-lay projection of a pillowed
bag rather than a cut outline.
The opponent
Measured over 1000 pitches per level by this game’s own test harness:
Social 4.8% in the hole and 58.5% on the board, Club 17.1%
and 74.8%, Pro 45.4% and 54.6%. These are measurements of this opponent and
are not calibrated against any published figure: the ACL’s statistics pages
return 404 and its home page links no statistics section, so no external anchor was
obtainable. Treat the levels as tunable, unanchored settings.
How the model is tested
The engine ships with two offline harnesses that are not part of the page: 9,744
assertions and 13 must-fire controls over the physics and the rules, plus 4,413 assertions
and 5 controls over the geometry of every triangle drawn. Cancellation scoring is verified
exhaustively — all 5,764,801 ways eight bags can land across the seven
outcomes — against an independent scorer. An independent simulator that shares no code
and no method with the engine (an exact closed-form solution of the drag equations in the
flight-path angle, adaptive Runge-Kutta-Fehlberg sliding, event-located collisions,
grid-quadrature areas) agrees on every bag outcome and every scored round. The flight
integrator’s convergence order is measured at 3.99, not assumed.
Controls
Space or Enter pitches. F flips the face.
V switches between the pitcher’s view and the overhead plan.
Escape closes this panel.
Credit
Cornhole is a folk sport with no single author; its modern codified form is governed by the
American Cornhole Organization (founded 2005) and the American Cornhole League (founded 2015),
and the rules quoted above are theirs. This is an independent reimplementation
written from those published rulebooks. It is not affiliated with, endorsed by, or connected to
the ACO or the ACL. No text, art, code or data from either organisation is included: the
dimensions are facts, and everything drawn here was written for this app.
Differences from a real match, stated plainly: singles only; no doubles, quads or skins
format; no pitch clock, no timeouts and no officials, so the 15-second clock, the three-strike
foot-foul escalation and the 12-point touching penalty are implemented in the scorer but not
reachable through play; the bag is thrown flat with no spin of the player’s choosing;
and the hole-capture criterion is this app’s own, since no rulebook describes when a
hanging bag falls.