Misunderstanding 12: An animal is expected to dole out to each relative an amount of altruism proportional to the coefficient of relatedness.
This is a serious error, and one that Dawkins himself committed in the first edition of The Selfish Gene – see the endnotes to the second edition. It arises from a misinterpretation of Hamilton’s Rule. This states that altruism towards relatives can evolve if br > c, where b is the fitness benefit to a relative, c is the cost to the altruist, and r is the appropriate coefficient of relationship between them. It does not imply that the amount of altruism given to relatives should be proportional to their degree of relationship. As Dawkins points out in his retraction, ceteris paribus any altruistic benefit should be given to the closest relative available, and none at all to any more distant relatives. It may be that diminishing returns or other special circumstances would make it worthwhile to divert some benefit to the latter, but there can be no general rule for this.