Bingo, you nailed what I was talking about here.
To restate the smith chart example in my words...
The smith chart, love those, that is a great tool to learn about feed lines and antenna matching. The X and R variables (aka complex impedance) along a theoretical lossless feed line will create a circle on the smith chart, and on that circle will be every possible combination of X and R that creates that SWR value. The center of the circle is the center point of the smith chart and a perfect match. All points that is a given distance from said center point have the same SWR, this forms an "SWR circle". The circle that the theoretical lossless feed line creates is a perfect match to this "SWR circle".
Once you add losses to the feed line the circle created by the feed line will become a spiral towards the center of the chart, with enough feed line it will eventually get to a perfect match which, again, is the center point on the chart, this happens irregardless of the starting mismatch. The lossier the coax the faster the spiral spirals inwards. What this is telling us is that the more losses there are in the feed line the better the match will appear to be on the radio end of the feed line.
If you have a situation where a short change in the length of coax makes a large change in SWR, then you have a problem that needs to be dealt with. Generally this is caused by common mode currents, a ground loop would be the next biggest cause, I have also seen a bad piece of coax do this.
The DB
