Originally Posted by
davidad
Back in the mid 90's when Bicycling was a decent magazine they had an article on this subject. At the time Mavic built 200 wheels for the European pro teams using all 4 possible patterns. They found that mirror image spokes with the pulling spokes elbow out stayed the truest and had the fewest broken spokes.
I'm aware of that, and was gratified to learn that I somehow chose to do it "right" back in 1967 or so. My theory at the time was that the elbows out spokes tended to touch and be supported by the flange rim, while the elbow in spoke were out in space angling away from the flange. It was then and remains only a loosely held theory, but for me enough to tip the scales slightly from a 50/50 proposition.
Until JB wrote his book, Americans tended to followed the European model, and many who were building before the book stayed that way. These days I can almost tell how long someone has been building wheels by which way they lace, ie. pre JB, post JB.
But, while I go the same way as Mavic, I don't think it makes a critical difference. IME well built wheels will fail from rim wear or fatigue before breaking spokes anyway, and that's if they aren't crashed, stolen or obsoleted first.