Alan_Wolfe wrote:

Just to be clear and accurate about this:

4 data points covers 1500 rpm range

6 data points covers 2500 rpm range

While that is probably the intention, that is not how the math was set up.

The Interval of the evaluation = the partition size * number of points. So 4 points x 500rpm evaluates a 2000rpm interval, and 6 a 3000rpm interval. You can't arbitrarily make the evaluation exclusive of the lower partition.

The ST rules describe a Riemann Sum(

https://en.wikipedia.org/wiki/Riemann_sum) when they add up the points as a means of evaluating different curves against one another(dividing by the constant 4 does nothing). One problem is that a LEFT sum (as opposed to RIGHT, or MID) is codified in the rules by 8.2 "If any of the above data points at higher RPM than Max HP RPM do not exist due to redline, then those potential data points will not be used in the calculation of Avg HP". LEFT sum is an OVERESTIMATION of area under the curve PROPORTIONAL TO THE SLOPE OF THE CURVE. A flatter curve means less of an overestimation.

This will not go away by adding more points.

Without even getting into the "race-ability" advantages of a flat curve(which we should not, because it too difficult to quantify), you can now see why so many people have chosen to tune their cars as flat as possible: because the ST rules math will give engines with a steep power curve a weight ratio disadvantage right off the bat.

As I suggested, the best way we can deal with this is to decrease the partition size, which increases accuracy. Simply increasing the interval does nothing.

Poor 4 cylinder NA. I feel your pain... And wait until the Rotary shows up. Poor poor rotary.

If the calculation is in fact using a Riemann Sum Approximation, then we should go to using 100 rpm intervals instead of 500 rpm intervals. This would increase the accuracy of the Approximation, especially for motors that do not have a flat HP curve. It would also reduce the difference between left rule, right rule, or midpoint rule approximations.