I was recently ripping about an inch off of a 6/4 thick, 5" wide board, and had the kerf close up on the riving knife. Thankfully the knife saved me from kickback, but I'd still like to avoid this situation in the future.
Is there a way to tell how a board will move when ripped by looking at the grain, or is it pretty much random?
Internal stresses in the board that are released when cut cannot be entirely predicted, no. A square, flat, straight board might rip or resaw into a problem piece.
That being said, a cupped, bowed, or twisted board will often yield a smaller board with similar problems, even after going through the process of squaring / truing up.
I've never noticed any predictable pattern when ripping timbers.
If you get a particularly strong or stubborn piece, you can knock some pre-cut timber wedges into the saw groove on the outfeed side in order to prevent excessive pinching, hammering/pushing them in as you go along.
As the others said, no true prediction. It has to do with internal stresses and cutting the board releases them. It might have to do with where the wood came from in the tree (a leaning tree or branch is more likely to have stress wood when cut into lumber)
Some of it can be how the wood was dried. Case hardening etc. There is a good chance that if it's the drying process, or the wood location, a lot of the wood from the same batch will likely have similar problems.
This not so much an answer, but to provoke some discussion.
When I rip clear, straight-grained wood, I do not expect any problems.
When I rip wood with "interesting" grain, I expect to have difficulty half of the time - the wood will either bind into the blade or splay apart.
The definition of "interesting" is up for grabs. Perhaps crotch wood or wood near a knot, or perhaps wood near the pith.
