My tek 3'd MY2003 STI made 290 at Powerstation

Was a bit dissapointed because I wanted 300+ but most people I speak to say powerstation's rollers are more accurate (but also read lower than others)

 

and drops the 5.9 to 4.8 to 60 as well (:

 

You could debate all day about which rollers are accurate, but it is true that powerstation's rollers are more pessimistic with their numbers than several others. A standard decat/ecutek exhaust will make perhaps 260-265 at powerstation, more like 280-285 at power engineering. I don't recall ever hearing of a PPP making book figures at powerstation, for what it's worth.

With a standard WRX turbo (TD04), you would need some further mods and some luck to reach an indicated 300bhp on optimistic rollers. I don't believe it has ever been done on a standard (non hybrid) TD04 at powerstation, & I doubt it ever will.

End of the day, roller figures are good to compare like with like (same car on same rollers in same conditions before and after mods, or 2 different cars on the same day). They are never going to be super-accurate representations, I'd always take the figures as ±20bhp. What the car drives like on the road is what matters,

 

Does this variation have something to do with the way the various RR owners calculate the flywheel power based on the "at the wheels" reading - I heard it was as simple as using a slightly different multiplier ?

If so then maybe the "at the wheels" figure is the accurate one?

onder2:

 

There's more to it than that. Most rollers try to measure in some form or another the transmission losses during "coast down" (i.e. when the car is in neutral at the end of the run), although they do include other factors in there. I'm not sure whether there is a straight multiplier in the calculations, but they are generally more sophisticated than "power at flywheel = power at wheels * 1.2" or similar. Perhaps someone else can comment

It is true that the "relationship" between the car and the rollers does have a significant effect. Tyre pressures can make a few bhp difference. At the most extreme case, I have seen one car "gain" over 20bhp as a rolling road day between two runs about an hour apart. The only change was to get the suspension geometry adjusted (mainly front tracking IIRC). This to me suggests that rolling road figures are a little fickle! To be fair, there are numerous examples of fairly consistent performance as well.