I am not certain what the data are, how how the angles and deviations relate. Perhaps either polarplot or polarhistogram will do what you want.
ang = linspace(0, 2*pi, 50);
dev = rand(size(ang));
r = 1;
th = linspace(0, 2*pi, 180);
xc = r*cos(th);
yc = r*sin(th);
rx = xc(1:15:end);
ry = yc(1:15:end);
[xs,ys] = stairs(ang,dev);
xs = [xs; xs(1:2)];
ys = [ys; ys(1:2)];
[xp,yp] = pol2cart(xs,ys);
figure
plot(xp, yp)
hold on
plot(xc, yc, '-k')
zv = zeros(size(rx));
plot([zv;rx], [zv;ry], '--k')
hold off
axis('equal')
Ax = gca;
Ax.XColor = 'none';
Ax.YColor = 'none';
.
