| Message |
The error bar is from the standard Poisson distribution for independent trials, and within 1 sigma, you have very little certainty (which is a different thing from error bars), that is why results falling within 1 sigma of each other is considered indeterminate. The dynos do not have to be controlled environments, the statistical analysis takes care of that... to get the error bar down enough for the result to be significant requires only 2-3 more trials that support the 15-17hp figure. But lacking that data, claiming the hipers make 15-17hp more is not right... PS: I like the minimal bends of the hipers, but in my experience some mods/designs sound good but make little difference. For example, I have found port matching on the intake and exhaust gives very little improvement at low-moderate power levels on the Z32. The design of the hipers may produce signicantly more hp, or it may not... pending more data, I am reserving judgement...
|
 |
