PM5 data v ‘other’ data

[Tony Cook]

Why is there a difference between what is shown on the PM5 and (for me) ErgData, and then the downloads from ErgData?
Specifically a WP L4 60’ piece. I set the PM for 10’ intervals with zero rest so I can see how many strokes I take for each 10’ sequence. On the ErgData display I hit the exact number of strokes for 5 of the 10’ sequences and went over by 1 on the the other one.
Target strokes was 1032 (I thought I’d done 1033 with the extra on each I counted) and the ErgData summary showed 1030. The 2 ‘rogue’ strokes were +1 and -3 on two pieces that I counted, and ‘live’ ErgData showed as on target.
The metres are even more confusing.
PM5 and ErgData summary shows 14,679m. Total of each piece equals 14,684 - I assume the 5 is down to rounding. My sequences were: +3, -6, +2, -4, -2,+1,-6 for an overall 6m under target - pretty good I thought. But then the biggest difference/ confusion is the .csv download from my C2 log that shows I was: -2.7,-17.8,-3.6,-20,-8.3,-16.2 for a total of 68.6m under target, that seems a big difference from what the PM5 shows.
Can anyone explain why the .csv file is overall different from what the PM5 shows?

[strupotter]

If you take a stroke that overlaps the end/start of an interval then that stroke gets voided, so my guess that's what happened with your stroke count. It's happened to me before. Better to do 1 hour single time with 10 minute intervals, and either write down your stroke count target somewhere you can see while you're on the ergo. Or you can practice the mental arithmetic times table of whatever rate you're doing.