"R"(a stats package) and got it to process it for me. The distance is in km. It's worth saying that virtually none of these contacts are SSB or FM - I am almost exclusively digital or CW
20m appears to have provided me with my best DX(a ZL), so it retains its label as a DX workhorse of the bands. followed by 10m, 15m and then 17m. I'd say 12m was underperforming - probably my antenna. I don't spend much time on 30m. I find 40m crowded. I don't stay on 80 or 160m long enough in the evening to pick up real DX.
What is interesting is that the median distance - half the contacts are above, half below - (indicated by the bold line in the middle) on 20m is way down on the median distance on 17m - I think this is caused by heavy local (i.e. European) QRM drowning out the DX, hence my impression that 17m at M0DEV is more DX-y. If you have a very high presence of european stations, then you won't hear the DX - even though it is there. Indeed, 17m at M0DEV has the highest Median. Of course, I am probably skewing this by going after DX contacts calling CQ, and ignoring the locals. The 20m distribution (and 15m and 40m) is classic positive skew, bunched up at the bottom end. The 17m distribution looks more "interesting".
It would be interesting to take this further, and try to tease out factors which are caused by my equipment, and factors which are really "there".
So what does it all add up to? Very litte!