Weight does not change the glide.
Start with the fact that surprises every student: adding weight does not reduce your glide ratio. Lift and drag both come from the same dynamic pressure, so when weight goes up, the whole flight simply happens at a higher speed — the geometry of the glide is untouched. A glider at 42:1 dry is still a glider at 42:1 with full water.
On the polar this has a beautifully simple form. Multiply the weight by some factor, and every point of the curve slides outward along a ray from the origin: both the speed axis and the sink axis scale by the square root of the weight ratio. Fifty percent more weight is √1.5 ≈ 1.22 — every speed on the polar, including best glide, moves 22% to the right, and every sink rate grows 22% with it.
So the heavy glider glides just as far — it merely does so faster, coming down the same slope at a higher speed and a higher sink rate. That single scaling law is the entire aerodynamic content of water ballast. Everything else is tactics.
The same tangent, further out.
Remember from the MacCready article that best glide is a tangent from the origin to the polar. Scale the polar outward along rays from the origin and the tangent line does not move — only the touching point does. Same glide ratio, higher speed. That is why ballast is pure profit on a fast final glide and in the cruise between strong climbs: you fly the same slope at more km/h.
The tool below puts both polars on one chart. The grey curve is the dry glider, the blue one carries water. Drag the wing loading and watch the curve stretch outward; then set the thermal strength and see whether the stretch is worth it.
Load the wings. Watch the trade.
Both gliders climb worse than the core value — circling costs sink, and the heavy glider circles wider and faster, further from the strong middle of the thermal. Try a 1.5 m/s core: the water hurts. Then try 4 m/s: the same water is suddenly worth real km/h, because the climb penalty barely dents a strong thermal while the cruise gain is permanent.
The price is paid in the climb.
If ballast were free, everyone would fly full every day. The bill arrives in the thermal, three ways at once. First, minimum sink scales up by the same √ factor as everything else — the heavy glider is simply sinking faster while it circles. Second, the circling speed is higher, and at a given bank angle the turn radius grows with the square of speed — the heavy glider is forced onto a wider circle, further from the core. Third, thermals are not solid columns of equal lift: they are strongest in the middle and fade outward, so a wider circle flies through weaker air.
All three losses land on the same side of the ledger, and they explain the whole character of a ballasted glider: magnificent between thermals, mediocre in them. In a strong, wide thermal the losses are a rounding error. In a weak, narrow one they can eat the climb entirely — the tool above will happily show you a heavy glider that cannot climb at all in air the dry glider uses to get home.
When the water pays.
The break-even is not a fixed number — it moves with thermal width, with how much of the flight is cruise, and with wind — but the shape of the answer is robust. As a rule of thumb, on a day of 2 m/s achieved climbs or better, a full-ballast glider beats a dry one; below about 1.5 m/s the water is a net loss; in between it is a judgement about how the day will develop, not about aerodynamics.
Two refinements matter in practice. Streets and ridge energy shift the answer toward water, because they raise the fraction of the flight spent cruising and reward the higher cruise speed twice. And the forecast matters more than the present: the classic error is dumping at the first weak climb of a day that was about to switch on. The classic opposite error is dragging full wings into the last, dying hour.
Dump it before you need it.
Water leaves the wings at a rate the designer chose, not the rate you suddenly require — typically three to five minutes for a full load. That number belongs in your planning: if the day is dying, the decision point is one thermal before the weak climb you will have to take, not inside it. Trying to save a bad low point by opening the dumps is usually two minutes too late.
Three habits close the subject. Dump before the last glide if the arrival is marginal — the heavy glider needs more height for the same distance only in the sense that it arrives faster; but it lands faster too, and an outlanding with full wings is a measurably worse event. Dump fully before any off-field landing. And respect temperature: water in the wings below freezing altitude is a structural question, and most flight manuals forbid it outright.
Set the actual wing loading in your computer and the rest follows: the polar, the speed ring and the final glide all move together, and the instrument stops advising a glider you are no longer flying.
Set the ballast once and every display recomputes the polar, speed-to-fly and final glide for the weight you actually carry.
Standalone vario with the ballast-corrected polar behind its MacCready audio and final glide.
The racing computer — wing loading is a cockpit setting, not a guess, and the optimization follows it.