FTP test - confusing results

I'm going to assume that the KICKR was the power source for all 3 systems (ie. the 945 wasn't connected to a different power meter).

The watch detects the inflection point in heart rate variability that is often associated with threshold heart rate. However, this can potentially be different to 95% of your 20 minute power which is the method Zwift and TP use. Garmin detects my threshold at about 20W lower than 95% of my 20 minute power, although I suspect that if I did more Lactate Threshold specific training, the two calculations may get closer.

20 minutes is a pretty rigorous test, and that is probably what you should use in gauging your FTP. You can edit your Power Zones in Garmin Connect to use this figure, rather than the LTHR method.

Indeed, as mcalista said - your ftp was detected by the firstbeat component of the cycling activity:

https://www.firstbeat.com/en/consumer-feature/functional-threshold-power-ftp/

if you want to follow the garmin 'regular' ftp protocol that does the math go with: https://www8.garmin.com/manuals/webhelp/forerunner945/EN-US/GUID-C8DAC4B1-3159-4E8C-ADD9-6FA012EBA697.html - most probably you would have got the same that zwift reported.

the value you've received is 95% of the 233 - and that is 221 - that's the way zwift, trainingpeaks and other platforms are estimating your ftp based on the 20min hard effort ;-). now good luck with raising that up ;-) - it's a good value to start from.

Yes, both: Zwift and FR945 are attached to the same trainer. 
Thanks mcalista for all the explanation it makes sense.  

Sure I can change the value manually but it does not affect the FTP report in Connect so this becomes a kind of pointless to me as I need to use external tracking if I'm going to follow the 20 min FTP test protocol.

I'm a novice in this area so I'll see how it'll all work.

Would you recommend me to disable auto FTP then ?

Thanks, tsofron I hope to raise it a little :D We will see how much. 

I'm a novice in this area but I don't expect very different results from Garmin FTP guided test to what I already got, as the same component will "judge" the FTP. 

Second I'm not sure I want to conduct 2nd FTP test protocol as I plan to conduct this tests regularly in 6 weeks intervals and 20 min effort sounds far superior to me.It is closer to what I'll experience during the race.  

I also need to decide on which estimate I'll base my training on and I think I'll stick with a test I have done. 

well - you can leave auto ftp on and correct it, if the firstbeat algorithm detects a lower value (happened to me as well - i left it on for my entertainment :D ) - and use only the values that trainingpeaks or zwift are delivering (or the guided protocol from the 945, if you start such an activity). also, it would't make sense to retest now - just follow your workouts having the ftp set to 221 and have a nice surprise in 6 weeks from now.