Ball flight. m dv/dt = m g + Fdrag + FMagnus, with Fdrag = -½ ρ Cd A |vrel| vrel (Cd = 0.47 held constant, ρ = 1.2 kg/m³, ball 0.142 kg, radius 32 mm) and vrel = v − wind. The Magnus force is ½ ρ A CL |vrel|² along the direction of ω × vrel, using the smooth-sphere fit CL = 1/(2 + 1/S) with spin parameter S = rω/v. Backspin lifts, sidespin curves the ball toward the side that the front of the ball turns to.
Stick sling. The stick is a single rigid lever that turns about the lower hand. The pocket moves on a circle, so the release speed is v = ω Leff, with Leff = stick length − 0.25 m (hand 0.13 m above the butt, pocket 0.12 m below the tip). The velocity is tangent to that circle, so the launch angle is the stick angle at release; a run-up speed is simply added to the horizontal velocity.
Integration. RK4 with a fixed step of 1/600 s. Ground contact uses a restitution of 0.55 and horizontal speed factor 0.8; the net is a heavy damper. The goal is 1.83 m x 1.83 m (6 ft x 6 ft); a shot counts when the ball centre crosses the plane inside the posts and bar, minus one ball radius.
Simplifications, stated honestly. Cd is constant although it depends on Reynolds number and surface roughness; spin does not decay in flight (real balls lose a few percent per second); the CL fit comes from smooth-sphere data and a rubber lacrosse ball is rougher; the stick is one pivot, not the two-hand push-pull of real shooters, so real elite shots (about 45 m/s) need a longer effective lever than a single-pivot model implies; no ball deformation, no stick-ball friction. Treat the numbers as an educational model, not measured data.
Checks below compare the integrator against closed forms: the vacuum range, the terminal-velocity law v(t) = vt tanh(g t / vt) with vt = √(2 m g / ρ Cd A), the finite-difference speed of the stick pocket, and a 8x smaller time step.