Mk5 R32 stock,  on 98Ron, ecu relies on knock sensors to retard the stock ignition under WOT by 3-6 degrees. Apparently this is normal. So, in my wisdom, just for learning purposes, I changed Lamfa from 0.9 to 0.84. I expected power to probably decrease, but to also see less or no retard I was astounded when the logs now show even more retard of 8-10 . This seems to go against all I have read. Any ideas?

If the ECU needs to pull more than 3 degrees timing, the octane is too low for the calibration or the static compression ratio is too high or the intake charge temp is higher than expected and lacking sufficient compensation.

That being said the R32 actually is two 3 cyl. engines connected to one crank and the two engines are not the same. The induction runners/chambers/operating temps etc. are different from one bank to the other. Most engines have a point where additional fuel beyond Leanest for Best Torque (LBT), will tend to reduce spark knock - up to a point. Most engines also can have a point where the AFR is so rich that you have secondary combustion that produces detonation of the remaining intake charge.

Obviously different people's experience is going to vary. I have personally observed Ford I-4 Turbo 2300cc engines that run fine at WOT with the OE 11:1 true AFR measured by actual fuel and air flow measurement in real time on an engine dyno yet the same engine will go into spark knock at 10.9 or richer AFRs, in my experience using in-cylinder pressure transducers and real time knock sensing equipment.

At .84 lambda you should not experience excessively rich AFR spark knock if the lambda reading is correct on the R32 engine. It would be common to run closer to .80 @ WOT with forced induction. OE C-O-P ignitions are somewhat output power limited due to reduced coil windings but due to the fact they do not use ignition cables a lower voltage is required to fire the sparkplug. In a NA application the OE C-O-P coils should not cause a misfire unless they are bad.

Detonation doesn't show as a misfire but it can be seen by the knock sensor if the frequency of the detonation is in the filtered "window" for the knock sensor and it is functioning properly. Unfortunately other engine components can sometimes produce a noise in the same knock sensor frequency range and cause ignition timing to be pulled when there is no actual detonation occurring. If the knock system is pulling 8-10 degrees of timing at WOT then you can typically hear the engine "rattling" under WOT as this would be severe and often terminal engine detonation. In a situation like this the sparkplug porcelain will typically have what appear to be aluminum speckles from the detonation abrading the pistons. The center porcelain may also be blistered or broken from excessive heat or shock. This may be your "misfire" if the plugs are cracked or otherwise damaged.

BTW, if you live where it gets cold, Winter fuel mixtures have cold starting additives that will cause unexpected detonation in some engines under WOT. The reason for this is the two types of lab test engines used to determine fuel octane ratings are not identical to many modern engine designs. Motor octane is a more severe test and tends to be more representative of WOT use where as Research octane is a very old test of much milder engine operating conditions.

None of the aftermarket tunes use SAE MBT/LBT calibration IME which is how you produce the maximum power at all points in the operating range of an engine. Basically the tuners try to find what they believe is an acceptable AFR/timing range that increases power as best they can. I'd take all chassis dyno data with a large grain of salt as the only proper way to actually calibrate an engine is on an engine dyno with proper instrumentation, true fuel/air flow in real time, knock sensing measurement, etc. in steady state mode with fixed operating temps for fluids, induction charge, etc. Flash readings on a dyno can be made to say pretty much whatever the dyno operator wants them to say.

It's worth noting that VW and other car makers use internal EGR to actually lower cylinder pressure at specific points where the peak cyl. combustion pressure exceeds what the fuel will tolerate to prevent detonation and allow a couple degrees more timing. They achieve this by advancing the intake cam so that some of the burnt exhaust blows up the intake port and dilutes the incoming charge. Retarded timing causes excess heat and poor drivability so with high static compression ratio and 98 RON fuel, proper calibration is important.

Just so that you are clear - a richer AFR does not make more power than the proper LBT (Leanest for Best Torque), AFR calibration. For an NA engine MBT and LBT will produce the maximum power possible for a given octane fuel. With FI ignition timing is generally knock limited and a richer AFR is typically used to be able to add a couple degrees more timing or to cool the pistons/intake charge for reliability.

Think ULEV ex. emissions and mpg for the different MK V ex. cam vs. the MK IV cam.

Engines run at factory LBT/MBT calibration should not have knock if the engine is mechanically sound and the proper octane fuel is used. Ignition retard is used to save the engine when the engine or operating conditions are different than how the engine was originally calibrated, i.e. lower octane fuel, carbon build-up in engine, timing chain or other engine wear, etc. What you are doing is attempting to recalibrate the engine for it's current fuel and/or mechanical condition which is fine as long as you understand the current engine condition is slightly different than when the engine was factory calibrated. LBT/MBT by definition will deliver the maximum power possible for the engine and fuel used.

For those unfamiliar with how factory engineers perform LBT/MBT calibration on an engine, it's quite simple. They run the engine under controlled conditions (steady oil/air/water temps, etc.), on an engine dyno at WOT in steady state mode at each RPM point from just above idle to maximum RPM. With the engine at WOT, they adjust the AFR to find the maximum torque. Then they adjust the ignition timing for maximum torque without spark knock. This becomes the base calibration for that rpm point in the base map. The process is repeated for every cell point over the entire RPM operating range.

Thank you GolfSportWagen & nyet for your input. Its got me a bit worried. The car runs well, and I can get a quarter mile done in 14.2s which is slightly better than factory specs and much better than the 15s I've seen on youtube but one possible explanation I just found is that the fuel where I live (AU) may not be real 98 octane. On the gov Federal  Register of Legislation, I just found this sentence. So I could well be putting 95 in the tank, I have no idea.

Absolutely. Not could. Will.