1 / 5
Pasa — Indian long dice, the d4 you can actually roll — image 1
1 / 5

Pasa — Indian long dice, the d4 you can actually roll

Order a print of this model

We calculate the final price from your selected file. You review and accept it before payment.

Saved on this device.

A long die: a square rod marked on its four long sides and capped at both ends so it can never stand up.

Rating
0 / 5
Likes
1
Downloads
4
Makes
0
Views
21
Reviews
0

Updated at source:

Model information

Category
Board Games
License
Creative Commons — Attribution

A long die: a square rod marked on its four long sides, capped at both ends so it can never come to rest on one. You roll it like a pencil and read the number on top.

Why this shape exists. The d4 is the one polyhedral die that does not work. A tetrahedron has no face parallel to the table, so there is no "up" face to read — you read a corner, or the bottom edge — and it does not roll, it stabs. The long die predates the cube and solves both problems by abandoning the idea that every readable face has to be a face of a solid.

Two numberings, same body:

  • pasa — faces 1, 2, 5 and 6, the traditional Indian form used in Chaupar / Pachisi, the dice game of the Mahabharata. Three of them throw 3–18 with gaps and the classic "special" results.

  • d4 — the same rod numbered 1–4, for ordinary RPG use.

The balance rule — the part worth stealing for other dice

Cube dice pair opposite faces to sum to 7. That is a balancing choice: the 6-pip face has the most material engraved out of it, so you put the 1-pip face opposite it, leaving 5 pips of missing material on that axis.

On a long die the same pairing is available, but the right answer is the opposite, because nothing forces opposite faces to sum to anything — you only ever read the up-face, so the pairing is free and its only effect is bias. So pair each face with its nearest count:

layout

opposite pairs

pip difference per axis

centre of mass off axis

naive (sum-to-7 habit)

1↔6, 2↔5

5 and 3

0.0536 mm

balanced (shipped)

1↔2, 5↔6

1 and 1

0.0130 mm

1-2-5-6 is unusually lucky here — its four values fall into two adjacent pairs, so every axis is off by exactly one ring of engraving. 4.1× less bias for zero cost in material, print time or legibility. The 1-2-3-4 stave gets the same treatment (1↔2, 3↔4) and lands on the same absolute 0.0129 mm.

The ends

Both ends taper at exactly 45° — self-supporting with the rod lying down — onto a small sphere. There is no flat of any size at either end, so a 62 mm rod cannot balance there. That is measured, not eyeballed: the total area of every facet pointing along the rod axis in the end zone is 0.52 mm² (the gate allows 1.5; a square-cut control reads 198).

Marking

Ring-and-dot — a cut annulus that leaves the centre island flush. It is the authentic mark on pasa, astragali and bone dice generally, and it suits FDM: it is a groove rather than an island, so nothing stands proud of a face and the bed-side face still prints cleanly. Paint-fill it, leave it bare, or print the rings in a second filament.

Two colours — and why it makes the die perfectly balanced

pasa_stave_2c.stl and d4_stave_2c.stl are the same dice split into two solids: the body, and exactly the material the rings take out of it. In Bambu Studio: Import → right-click → "Split to Parts", then give each part a filament. Same 62 × 10 × 10 mm, same geometry.

The interesting consequence is on the balance work above. Filling the rings flush with a second colour of the same polymer puts the missing material back — body + ink equals the unmarked blank to 0.0000 mm³ — so the two-colour die's centre of mass sits on the axis. The 0.0130 mm of engraving bias is not reduced, it is gone. It is exactly what casino dice do with flush paint, done in filament instead.

The 14 ring annuli are welded into one shell by 1 mm traces buried 0.7 mm under each face, so Split to Parts hands you two parts to colour rather than fifteen. The traces are narrower than a ring's 2.0 mm centre island, so every island keeps body on both sides of it and cannot ship loose.

Cost, so you can decide: the weld puts ink on nearly every layer of the 10 mm height instead of only the ring bands, which takes a plate from 31 to 50 bi-colour layers, about 18 g of purge. That figure is per plate, not per die — so print all three staves of a Chaupar set together and it costs the same as printing one.

Files

file

qty

what

pasa_stave.stl

×3

the Chaupar stave, faces 1-2-5-6 — one colour

d4_stave.stl

×1+

same body, faces 1-2-3-4 — one colour

pasa_stave_2c.stl

×3

two-colour Chaupar stave: body + ink, Split to Parts

d4_stave_2c.stl

×1+

two-colour 1-2-3-4 stave

pasa_dice.scad

 

the full parametric source

Print the _2c file instead of the plain one if you have an AMS — not as well as. A Chaupar set is three pasa staves; a d4 for RPG use is one.

Printing

Lies flat on a face. No supports, no brim. The only overhangs are the two 45° end tapers. Any material; something with a bit of density rolls better than a lightweight PLA at 10 % infill — 25–40 % infill is worth it, and infill is symmetric so it does not disturb the balance work above.

The bed-side face's rings come out very slightly softer than the other three (first-layer squish). If that bothers you, put the face you care about least downwards.

Printed and rolled — the photo is the actual print in green PLA with white ink.

Source

Hand-authored parametric OpenSCAD, self-contained, attached. stave_len, stave_w, edge_r, tip_r, ring_od / ring_id / ring_depth / ring_pitch, plus web_w / web_t / web_over / riser_w for the two-colour weld. The end geometry is derived rather than free: core_len = stave_len - stave_w is what makes the taper land at 45° with the tip sphere reaching exactly stave_len. die="pasa" or die="d4" picks the numbering; balance=false reproduces the naive pairing if you want to measure it yourself.

Tags