Bias: it’s not just found in opinions, it’s also found in your paint
Red-Yellow-Blue; primary-secondary-tertiary. Most everyone who’s in any phase of design has probably had thier run-in with mixing colors, either with light or with paint, and most of us can, no doubt, remember the fine feeling of subtle panic which which we approach mixing colors when we do illustration.
Anyone whose had to mix colors knows the drill: experiment and experiment, come up with mud, throw away mixed paint, or settling for a color that’s kind of what you had in mind.
When studying color, painters are trained in the use of the RYB color model. The three primaries, red, yellow, and blue, are the bases from which all other colors depend, and which cannot be in theory mixed from any other. The secondaries–those mixed only from adjacent primaries–become red+yellow=orange, yellow+blue=green, and blue+red=violet. This forms the basis of the color wheel with which we are all so familiar, giving us a foundation toward finding the seemingly-infinite range of colors that we would use to create our worlds on paper or canvas.
For beginning students of color it’s a sound structure, but, in our world of reality, where nothing is perfect–especially primary colors–it has significant, though conceptually subtle flaws. The simple color wheel, with us in concept from the time of Goethe and his symmetrical color circle first advanced in his Theory of Color, is correct, but only in the ideal. Though it might seem obvious to state it’s easy to overlook that, in practical terms, there can be no such thing as a theoretically pure primary color in reality.
In simpler terms, your red paint will always have the merest amount of yellow or blue in it; your blue will have some amount of red or yellow, and your yellow will have some miniscule component of red or blue.
This tendency is called color bias. First written widely about (to our knowledge) by artist and color theorist Michael Wilcox in his book Blue and Yellow Don’t Make Greenin about 1987, color bias theory has been gaining currency since (the author recalls color bias being touched upon in his own painting classes in school). When implemented, color bias moves color mixing away from a thing of chance and experimentation and becomes a thing of reason and thought.
As a thumbnail example, lets consider a notional beginning painter wanting to mix a standard orange from what they think are ideal primary tubes of yellow and red that they happen to have. Basic theory tells them that if they mix equal amounts of each they will get an ideal orange, and they do this, but further assume that the either or both colors have a bias toward blue (refer to the Wilcox Wheel illustrated to picture what is meant here):
This mental excercise, remember, is predicated on the assumption that our notional student is unaware of bias theory and assumes that the primary paints they used are ideal. It should be little stretch of the imagination to see how the equations would change with awareness of color bias: in each situation knowledege of the bias provides not only how much of each one should use, but a clue as to how one might remedy the situation of a less-than-idea color mix.
Mr. Wilcox has built his implementation of color bias theory into an organization called the Michael Wilcox School of Color; its homepage for the American market is here. It has a variety offerings, from newsletters to its own line of paints, that will increase your knowledge of color bias theory. His book, Blue and Yellow Don’t Make Green, is available in its current printing from Amazon via this link.
On color theory in general, Wikipedia has a very literate and extensive discussion here.

