I've done little dyno work in the past. You're doing good to get 1% accuracy from run to run if you just take the car or engine off the dyno and put it back on in the real world.

1% of 340 is 3.4 so figure their numbers are not much if any better than plus/minus 3.4 hp for each test. If one tests give results 1% high and the other 1% low, then any small differances you are looking at are buried in the error.

 

I don't doubt that it can be that much. The only issue in using such a statistic is that I have never heard of someone adding a minor mod and reporting back that on three pulls they lost power. For example for those that add an underdrive pulley the nay-sayers state that the 9rwhp they got on the average of three pulls was dyno error when the fact is that I have never seen anyone loose 9rwhp on 3 consecutive runs. Or even better where is the -xx hp run in a tripple? A guy does a baseline of three pulls I always see 3 runs that go in one direction usually due to the car getting warmer (or hot). In a set of three pulls its 341, 344, 345 not 341, 344, 336 by you error rule. Such an exception is not the norm as I have never seen it and I asked RWTD about it and in his experience it never happens.

 

This is all doubtless true. It pops in most threads that discuss mods that only offer small gains. The usual argument is that, if the gain is within the + and - or + or - accuracy tolerance then it can't be measured. If it can't be measured then we really don't know if it is useful. Once again, that is all quite true. However, it may be possible to add a mod that is slightly useful even though it can't be measured on a dyno. Ported throttle bodies, throttle body spacers, plug wires, certain spark plugs, MAF sensors might all provide gains that are too small to measure but, collectively, they might add add 1 or more units of significant gain.

Back to the dyno. I assume that the readings are correct unless there is some obvious flaw in a given pull. I acknowledge that I will sometimes waste money because I just really can't tell that I didn't. Common sense, SOTP and other factors have their place in the analytical gaps left by superior tools like dynos.

This is one time time when I am glad that I am not an expert. It makes it easier to sleep at night!

 

Today we have dynos and etc..... to help us evaluate performance products. When I started racing it was trial and error. You changed one part at a time and if the MPH increased...... that was good.

In any case the problem with performance parts is advertising period. If you consider how large the performance market has become in just the last few years...... it is astounding.

Take any performance product Cold air intakes, throttle bodies and the like. The performance gains between brands is not material (assuming there are any gains at all). We all get caught up in the advertising hype, trying to get the most bang for our performance buck.

Do not get me wrong, there are combinations of parts that work very well together and make more horsepower in an particular application. Our problem is finding that mix of parts and installing it on our car.
I think that those of us who have been playing around with cars for quite some time are always looking for information, helpful hints of what works and what is marginal or just BS.

In any case this forum is helpful, as none of us know it all and even though we may have quite a bit of experience........ I have found that in most cases one can always use the benefit of another opinion. It is my firm belief that there is never only one right way to accomplish anything. And yes there is a place for aluminum flywheels however, I have an automatic