Michael Eriksson
A Swede in Germany
Home » Politics | About me Impressum Contact Sitemap

Bias test


In today’s world, there appears to be an enormous amount of bias and hypocrisy-around-bias, especially when factors like race, sex, and sexual orientation are concerned.

Below, I suggest a simple self-test through which the reader can spot some own weaknesses.

The test

U.S. version

Imagine a movie featuring eight characters in a car race. There is no prize money—but they all want to win equally strongly. (Reasons are beside the point, but a simple “to prove that I am the best of us” is an example.) We have one person per car, no teams, no cooperation. Everyone has approximately the same screen time, and the movie is not angled to make any given character seem artificially more sympathetic than the others.


The last sentence is quite important for the test, as movie viewers very often root based solely on screen time and angling, even among otherwise equally worthy characters.

Imagine, further, that these racers differ in sex, race, and one additional characteristic that you have picked yourself, with the one restriction that it must feel highly relevant to you. (What this is, will depend very strongly on the individual at hand. Examples include sexual orientation, occupation, religion, being or not being a parent, and age group—but countless other examples exist.) For simplicity, keep sex male/female and race White/Black or, should you be neither, your own race/not your own race. Similarly, keep the chosen third characteristic on an “is vs. is not” basis. Make sure that each racer has a unique combination of characteristics.


The division into “binary” couplings is strictly a matter of combinatorics. As is, when each character has a distinct combination of three distinct “is vs. is not” characteristics, eight racers suffice. Add more values per characteristic and the number of racers explodes. Ditto, if more than one additional characteristic is chosen.

Now, which of the eight do you root for? Answer honestly.

Non-U.S. versions

Depending on the country at hand, a race-based, let alone specifically White/Black, juxtaposition might be much less interesting and the test might need corresponding adaption. Typical choices with a large risk of political loading and/or blanket judgments include population majority vs. large minority, (real or perceived) occupier vs. the occupied, and enemies of a recent war or of recurring warfare. In some cases, an even wider net might be needed, e.g. to include societal class.

Evaluating the answer

If you chose someone based on a combination of characteristics, regardless of whether and to what degree you share these characteristics, you have almost certainly failed the test. Neither race nor sex is a good criterion to be rooting for someone, and the same applies to most “third” characteristics likely to be chosen.

If you do happen to share the characteristics, it is the worse; if not, it can pay to investigate why some combination of characteristics was favored.

A partial exception is when the third characteristic is of a relevant nature, e.g. whether someone values “fair play” in a competition. Even then, a reduction to less than four racers is not possible, and the test will either be failed because one of the four was given undue preference, or see an answer of “any of the four”.

What then is a “passing” answer? A refusal to answer (!), with a motivation of e.g. “I would have to see the actual movie first, so that I can determine who is the most deserving” or “Neither race nor sex [etc.]”; or a reaction with a similar effect.

(However, a reasoning like “the character with characteristics A, B, and C is bound to be the most deserving” is a complete failure. Ditto, if in a more indirect manner, a refusal to answer for a spurious reason, e.g. an insistence that “Sex is not binary!!! The test is evil bigotry!!!” even in the light of the above side-note.)


  1. A similar test can be applied to real movies (and other works of fiction, not necessarily dealing with car racing), but with the dual disadvantage that (a) there will be less flexibility in combinations of characteristics and (b) impressions will likely be strongly influenced by actual worthiness to begin with. For instance, most racing movies that I have encountered have had some variation on the “Dick Dastardly” character—who will naturally receive few votes, regardless of issues like race, sex, and whatnot. (If you do root for such a character, and especially if you do so for reasons like race and sex, then some serious introspection is called for.)

    However, if sufficiently many works are considered, similar trends of “undue rooting” are often easy to find and some works can be reasonably suitable for the right persons even stand-alone (“Who is your favorite ‘Mod Squad’ character—and why?”). Moreover, such trends can be found for characteristics that are off-topic for this page, say, whether rooting is based on physical attraction. (If faced with the “Mod Squad” question, myself, I might give Peggy Lipton’s character the nod, simply because of her beauty.)


    Here I slipped: It is not a given that someone roots for a favorite character or vice versa. The outright opposite is not uncommon in works with a villain who is sufficiently interesting, well written, well acted, whatnot relative the hero. I let the above stand, however, as a formulation like “Which ‘Mod Squad’ character do you root for?” would be almost silly and as the formulation actually used does pose a valid test, if one slightly different from the main test.

    (Of course, the main test was very deliberately written to create a much more competitive situation than the one in the “Mod Squad” or the common match-ups of a single good guy/team vs. a single bad guy/team.)

    A particular benefit is the possibility to see whether more superficial or more substantial characteristics lead to rooting. For instance, rooting based on a shared skin color is highly superficial, based on e.g. a shared profession is more substantial (but still superficial), based on e.g. shared values is more substantial yet, and forgetting the “shared” part and looking at more abstract worthiness is quite substantial.

    Even so, some care must be taken, as there might well be systematic differences in portrayal or behavior in works of fiction, and the viewer might react to such differences. For instance and especially in recent works, characters that are Black, homosexual, whatnot, are often given a politicized twist, which tends to turn my sympathies against them. The cause is the politicized twist, not skin color, sexual preference, or whatever applies. Vice versa, White men are often given an unduly negative image in fiction, e.g. through an entirely disproportionate representation as wife beaters, cheaters, criminals, bigots, whatnot, and a negative reaction to such does not necessarily reflect an anti-White sentiment in the viewer. (But might well do so in the creative forces. Some more specific groups of predominantly White men, e.g. survivalists and MRAs, seem to be given an extremely distorted portrayal as a matter of course, with strong indications of an agenda to unfairly and misleadingly discredit these groups in the public eye.)

  2. Likewise, it can pay to look at rooting in the context of real sports, with the reservation that sports, from a spectator perspective, is usually something irrational and the rooting best viewed as something to add excitement and entertainment value. (Exceptions include family members of the athlete at hand.)


    An interesting twist on rooting is my take on tennis over the last few years (time of writing: 2024): I was well ahead of most others on Djokovic and GOAT-hood. Every additional success that he achieves proves me the more correct, which gives me a reason to root for him.

    Outside of sports, however, rooting-for-poor-reasons can be an outright problem, especially as there often is room to actually act in favor, or to the disadvantage, of the involved. Consider political support (voting, donations, volunteering, whatnot) based on something as superficial as a common race (as was often the case with Blacks and Obama) or sex (women and Hillary). Politicians should be judged exclusively on more solid criteria like (prior or expected) policy choices and (proven or estimated) competence level.

  3. Similar issues exist in other areas (and can sometimes be used for similar tests). For example, if someone looks for a role model, is that role model chosen for excellence in the field at hand or based on some “commonality”? (Note e.g. the common Feminist obsession with having specifically female role models for young women, e.g. “to get women into science”. Better would be for the young women to pick a role model based on worthiness as, e.g., a scientist, to release themselves from this obsession with “gender”, or, even, to question whether they need a role model in the first place—chances are that the best thing to do is simply to strive for excellence in one self.)

  4. The test is not easy and even those who somewhat legitimately see themselves as unbiased can stray. (Note my request to “Answer honestly.”—not just based on what the perceived-as-correct answer might be.)

    However, I expect a great proportion of those who illegitimately see themselves as unbiased to fail. Indeed, my motivation for the test is largely Leftist hypocrisy, as my observations over the years show Feminists to be among the most sexist groups in existence, Black activists and whatnots, maybe Blacks in general, to be far more racist than typical Whites, etc.


    Leftist bullshit like “unconscious bias” is not bullshit because we have no biases of which we are not conscious —virtually all of us do and the observation borders on the trite. It is bullshit because it is abused for attempts to bend observations to fit theories that must not be questioned, to discredit opponents, to attribute to “evil” motivations even when perfectly legitimate motivations exist, and similar; and because it is given much too much room.

    The point of “unconscious bias” is not to have a serious discussion of such biases—it is to further an ideological agenda.

  5. In the formulation of the test, I spoke of “angled to make any given character seem artificially more sympathetic”.

    Here, I needed to be somewhat vague, because being more explicit might have given a strong clue to what the correct answer was. Examples of such angles, in my intent, include a romantic subplot given to the “angled” racer but not to the others, the angled racer coming back from a bad injury, and similar. Not included is more objective deserving–undeserving criteria, that play more on true merit and less on sympathy-creating tricks. Consider a two-man race with prize money, a competent driver who races fairly to save an orphanage, and an incompetent driver who tries to win by sabotage in order to become famous. Preferring the former over the latter is certainly legitimate. (But note that the border can be hard to draw and that particular care should be taken when several racers are deserving/whatnot but the depiction is asymmetric. What if the one is depicted as racing for the orphanage while another is racing for his mother’s hospital bills, but this is only tangentially mentioned in the movie?)

    Of course, such “angles” can be a considerable issue in real life too, and awareness of the risk can reduce a similar problem to the one tested above. Chances are that the real-life sympathies of most are based less on merit and more on factors like who is better known, whose story has actually been heard, who has a more sympathetic behavior, and similar—corresponding to the “angling” of movies. For instance, if two parties are in conflict, it is extremely foolish to hear the complaint of the one and immediately take sides without bothering to hear the side of the other—but this is exactly what most seem to do.

    Could this vagueness have put some test takers at a disadvantage? Conceivably, if they assumed that no merit-based criteria were available either, and picked by the three characteristics only based on the lack of such merit-based criteria. Even then, a choice based on the three characteristics would have been flawed and the test takers should still, ideally, have refused to make a pick.

  6. A related idea to “root for” is “identify with”. While I personally see the common mentions of “identify with” in discussions of fiction as naive, paying attention to what characters we do or do not identify with can be similarly useful.


    What is wrong with “identify with”? At least two things:

    Firstly, it implies a much to strong connection between viewer (reader, whatnot) and character. If the connection, with that strength, is there, it points to a weakness in the viewer that is problematic and, with some reservations for children, should be addressed accordingly. If not, the formulation is unfortunate and misleading, and something more appropriate should be chosen. (Whether the replacement necessarily would be “root for”, or something equivalent to that phrase, I leave unstated.)

    Secondly, “identify with” is an area where some differences that would be illegitimate with “root for” either are legitimate, would be very hard to overcome, or would be pointless to overcome. (The last, again, with an eye at the weakness involved in “identify with”.) For instance, I might well “root for” a 16 y.o. blond valley girl in a fight with a vampire, but “identify with”? It simply would not make sense.

Follow-up exercise

Spend some time thinking on that third characteristic, why you picked specifically that characteristic, whether it is really something that made sense, etc.

(Here I can say next to nothing about what answers are, in some sense, correct, as this will vary far too much from person to person, characteristic to characteristic, etc.—to the point that the same choices by two different persons need not be equally worthy. If in doubt, thinking and self-reflection are often more important than answers, per se.)