Chassis dyno graph

My guess is, data was being recorded prior to the throttle being wide open.

I’m basically disregarding everything below 4500 rpm or so.

On an engine dyno, you get the motor stabilized at WOT prior to making a pull, so all the data is at WOT.

On the wheel style of chassis dyno, traction on the rollers can be an issue, so often times the bottom of the curve isn’t representative of the actual power output, since that part of the curve isn’t always at WOT.

With an eddy current absorber, provided traction wasn’t an issue, it should have been possible to have the engine stabilized at WOT, at the stall speed to the pull being started.
However........ if the roller/wheel speed was well below what would be correct for whatever that stall speed was....... when the pull was started, you’d see a huge spike at the beginning of the curve from the large amount of TQ multiplication that’s happening from the crank/trans input shaft speed differential.

It’s always important to understand that the dyno graphs between a chassis dyno vs an engine dyno can have vastly different shaped curves...... especially if there is a fairly high stall converter involved.