Here's the deal with the TPS. Most, if not all, electronically controlled ignition and fuel systems use one to gauge engine load. The more advanced ones add a manifold pressure sensor to get a more exact picture of the situation. They also use a tach signal to read the engine speed. With this info, the control module decides what the ignition timing should be. The TPS is set up so that it sends a voltage back to the ECM that varies with the throttle opening, and all of them that I can think of right now are set up so that high voltage is a closed throttle, and low voltage is an open throttle. What the normal response to a more closed throttle at higher speeds is is to advance the timing more than it would be at full throttle to improve part throttle power and fuel economy, exactly like the vacuum advance units on older car distributors did. While the max ignition advance on an engine might be something like 34 degrees at full throttle with the revs up, the max advance at cruise might be as high as 50-60 degrees. The reasons are a little complicated and relate to dynamic compression, but that's what happens. If you unplug a TPS, the voltage returned to the ECM will be zero. The ECM will respond to that either by seeing it as a fault and entering a default mode of operation with fixed timing, or by simply seeing it as full throttle all the time, and going about its business as usual. Most modern systems react with a combination of the two; they complain about the fault, but continue to control what they can based on the information they have. The peak power increase that Tryce has shown indicates that YZF CDI clearly changes the full throttle timing when the TPS is unplugged, but you might get similar results by adjusting it. You could also make it ping at part throttle that way, though. The choice to have an open (unplugged) TPS be full throttle is both deliberate and important. If the assumption is that the throttle is open, and full throttle requires less advance, the engine will still operate safely at full throttle, and only part throttle operation will be negatively affected because the timing will be retarded from the optimum levels. As far as the '03 YZ450 specifically, I find it pretty interesting. I can't and won't argue with T-Dub's results; they are what they are. But my '03 does not have the dip in its power curve that it did when it was stock, and the TPS is still plugged in. I made it go away by putting the FMF Power Bomb on it. It comes back if I use the stock header with either the FMF or the stock muffler, although it's worse with the stocker. Based on my understanding of how the PB works, my guess is that there is a complex situation involving the exhaust system pressure levels during the overlap event at that speed that is helped by changing the advance curve slightly, and that is changed by the PB header, but I have no tools at hand to verify that. I leave mine connected because I like the way it runs as is. If it did work so well, I'd fool with it.