Most of the stability questions we get are really warranty questions in disguise. A venue buys a tall stool with a slim base because it photographed well, a guest leans sideways to grab a coat, the stool goes over, and the email that follows says the product is "unsafe." Usually nothing broke and nothing was out of spec — the geometry was just never discussed. So here is the geometry, with numbers, the way we walk buyers through it before we cut steel.
The triangle that decides everything
A stool tips when the combined centre of gravity — stool plus sitter — moves past the point where the base touches the floor. Two figures set how hard that is to do: how high the mass sits, and how far the base edge is from the centre line. The seated person dominates the mass, and a seated adult's centre of gravity sits roughly 250-300 mm above the seat pan. On a bar-height stool with a 760 mm seat, that puts the combined CoG in the region of a metre off the floor.
Now the base. A typical trumpet or four-leg bar-stool footprint runs 400-480 mm across, so the support edge is 200-240 mm from centre. Tip angle is the arctangent of that half-width over the CoG height: with a 220 mm half-base and a CoG around 1,050 mm, you get a tip angle of roughly 12 degrees. The same base under a 660 mm counter stool works out closer to 14-15 degrees, and a low dining chair sits past 20. That is the whole story of why bar stools go over and dining chairs do not: the bar stool starts life with a third less margin, and every millimetre you shave off the base diameter for the sake of a slim look takes more of it away.
Four legs versus a round base: the pivot line moves
A round trumpet base gives the same tip radius in every direction — what you measure is what you get, all the way around. A four-leg stool is sneakier. It tips over the line between two adjacent feet, not over a foot itself, so the effective half-width across a side is shorter than the corner-to-corner figure on the drawing. A four-leg stool with feet on a 460 mm square has a 230 mm reach to each corner but only about 230 × 0.71 ≈ 163 mm perpendicular to each side — and sideways is exactly the direction people lean. When we quote a four-leg frame against a round-base bar stool, we compare the side-on figure, not the diagonal, because the diagonal flatters the four-leg design by around 40 percent.
Splayed legs claw most of that back. Raking the legs outward 8-10 degrees adds 60-80 mm to the footprint at floor level without widening the stool at seat height, which is why almost every tall stool we build has visible splay. A dead-vertical leg on a bar-height stool is a styling decision with a stability bill attached, and we say so on the quote.
What the sitter does to the math
The static numbers above assume a person sitting still on the centre of the seat. Real venues are worse. A guest perching on one edge moves the CoG sideways before any lean starts; someone hooking both heels on one side of the footrest ring adds a steady sideways moment; and the classic failure — twisting round and leaning to reach a bag on the floor — moves mass sideways and down-and-out at the same time. This is why the stability standards do not test an empty stool: EN 1022 loads the seat and pushes to overturn in the worst direction, and BIFMA runs equivalent rearward and sideways procedures. We build and test to those reference methods, and where a contract order needs a formal report, testing can be arranged per order. The standards conversation itself is one we covered separately in our note on EN 16139 versus BIFMA for stools.
Glides and floor pads: grip is not always your friend
Here is the counter-intuitive part. When a stool gets shoved sideways, one of two things happens: it slides, or it grips and rotates. A hard nylon glide on polished tile slides — the push dissipates as a skid and the stool stays upright. A soft rubber pad on the same tile grips, and the same push becomes a tipping moment about the far edge. So "upgrading" to rubber feet can make a marginal stool easier to tip over, not harder. The flip side is floor protection and noise: hard glides scratch timber and clatter on stone, which is why hospitality specs usually want felt or rubber anyway. Our default is to match the glide to the declared floor — nylon or felt-insert glides on hard floors where a controlled slide is acceptable, rubber only where the floor is already high-grip or the venue insists — and to flag the trade-off in writing when a buyer asks for rubber on slick tile.
Uneven floors are the quieter problem. A three-point stance never rocks; four points and a trumpet ring do. On the old stone floors a lot of bars have, a stool that rocks gets a folded coaster jammed under one foot by week two. Adjustable threaded glides cost a few cents per foot and remove the complaint entirely; we fit them on request to any of the frames in our product range.
The numbers to put on your spec
If you take one thing from this, make your stool spec say three things it probably does not say today: a minimum base footprint relative to seat height (as a working rule we like the floor-level footprint to be at least 55-60 percent of seat height — about 450 mm under a 760 mm seat), the glide material matched to a named floor type, and a stability test to EN 1022 or the BIFMA procedure for the destination market. None of these costs real money at production time. All of them cost money after the stool is on the floor.
Send us your seat height, your floor type and a photo of the room, and we will come back with the base geometry and glide spec we would actually build — through the contact page or at [email protected].