The issue is whether bent shaft paddles are more efficient for forward stroking than straights. More "efficient" means, to me, that the very same person when FORWARD STROKING : (1) can go slightly faster with a bent at high stroke rates; and (2) can paddle with less energy expenditure at any stroke rate, including touring, recreational and lily-dipping stroke rates. Based on 40+ years of always having a straight and bent in my various canoes when paddling flatwater, I consider this conclusion to be inarguable.
I like the wording of Caplan's conclusions since they don't need any qualification or interpretation. Both Caplan and Burke conclude that for a seated paddler, if 2 paddles are identical in every way but bend angle then the bent paddle has a slight edge in efficiency.
Conclusions
The aim of this investigation was to develop a mathematical model of the canoe paddling stroke in order
to investigate the influence of paddle orientation through the stroke on mean boat velocity. The model,
which was based on Newton’s second law, was shown to be valid against on-water data for an elite female
dragon boat paddler in an O1 outrigger canoe. Due to the availability of appropriate data, it was only
possible to validate the model for a single female paddler in an O1 outrigger canoe. Further research should
aim to validate the model against data for a range of boat classes and crew sizes.
Paddle offset angle relative to the shaft was adjusted incrementally. The simulations showed that by
reducing the angular offset of the paddle blade by -20 degrees, peak mean boat velocity was achieved. The
angle of attack when peak paddle velocity occurred was close to 0 degrees for an offset of -20 degrees. This
finding suggested, at least for the paddler investigated here, that the fastest boat velocity was achieved by
ensuring that the face of the paddle was vertical at the point in the stroke when the paddler had accelerated
the paddle to the greatest extent through the water.
These findings suggest that, at least for the paddler investigated here, the orientation of the paddle blade
through the stroke should be adjusted to suit the change in paddle velocity generated by the movement of the
paddler. In order to more accurately model the stroke, future developments should increase the degrees of
freedom of the model to allow for the influences of vertical forces and their effect on the drag experienced by
the boat. The effect of yaw, pitch and heaving motions should also be investigated.