Those that did the decaying are called parent nuclei.
There are many different kinds of radiometric dating and not all conclusions we will reach can be extrapolated to all methods used.
Also, different radiometric dating techniques independently converges with each other and with other dating techniques such as dendrochronology, layers in sediment, growth rings on corals, rhythmic layering of ice in glaciers, magnetostratigraphy, fission tracks and many other methods. There exists different versions, or isotopes of many elements.
If the samples have been undisturbed closed systems since formation, the data will fall on the same line (the isochron from which the diagram is named).
The slope of this line is a function of the age of the rock. The reason scientists normalize with another stable isotope of the same element as the daughter is because most chemical or physical processes that occurs normally in nature does not differentiate between different isotopes of the same element when the difference in mass is as small as it is between isotopes of the same element that is used in radiometric dating.
Most young earth creationists reject all of these points.
As a scientific skeptics, we ask ourselves: is this really the case?
Furthermore, by studying supernovas far away, scientist have determined that decay rates have been constant in the ancient past as well.
Not only that, different radioactive isotopes decay differently and it is enormously improbable that a postulated difference in decay rates would affect all of them in the same way, yet as we have seen, different radiometric dating methods converge on the same date (within margins of error).
Both of these are divided or normalized by a stable isotope of the same elements as the daughter element.
So on the x-axis, we have parent/(another stable isotope of the same element as the daughter) and on the y-axis we have daughter/(another stable isotope of the same element as the daughter).
If so, the data will fall on an isochron line, but will be all over the place.