$$(a^2 + b^2 + c^2)/3 \geq ((a + b + c)/3)^2$$ OR $$(a^2 + b^2 + c^2)/3 \leq ((a + b + c)/3)^2$$

for all $a, b, c \geq 0.$ The problem wants me to find which inequality is correct and then provide a proof for it.

I'm pretty sure the first inequality is correct (I assumed this after substituting random values as $a, b,$ and $c$). I've tried factoring the inequalities and this is what I ended up with:

$$9(a^2 + b^2 + c^2) \geq 3(a^2 + b^2 + c^2 - 2ab + 2ac + 2bc)$$

However, I don't know where to go from here. I'm also meant to be applying the AM-GM theorem to solve this problem but I'm unsure where and how to apply it in this situation. Any help would be extremely appreciated :)



$a,b,c \ge 0.$

The first inequality is equivalent to (Dr. Graubner):

$a^2+b^2+c^2 \ge ab+bc +ac$.


$a^2+b^2 \ge 2ab$; $a^2+c^2 \ge 2ac$; $b^2+c^2\ge 2bc$;

Adding LHS and RHS of these inequalities:

$2(a^2+b^2+c^2) \ge 2(ab+ac+bc),$

and we are done.

  • $\begingroup$ Can you please explain how you got to the equivalent first inequality? $\endgroup$ – Alexander B Jun 10 at 9:04
  • $\begingroup$ Alexander: $3(a^2+b^2+c^2) \ge (a^2+b^2+c^2+2ab+2ac+2bc)$. Subtract the RHS square terms : $2a^+2b^2+2c^2 \ge 2(ab+ac+bc)$OK? $\endgroup$ – Peter Szilas Jun 10 at 9:59

Your first inequality is equivalent to $$a^2+b^2+c^2\geq ab+bc+ca$$ and this is $$(a-b)^2+(b-c)^2+(c-a)^2\geq 0$$ which is true.


Yes, we can use AM-GM here: $$\frac{a^2+b^2+c^2}{3}-\left(\frac{a+b+c}{3}\right)^2=\frac{1}{9}\sum_{cyc}(2a^2-2ab)=$$ $$=\frac{1}{9}\sum_{cyc}(a^2+b^2-2ab)\geq\frac{1}{9}\sum_{cyc}(2\sqrt{a^2b^2}-2ab)=\frac{2}{9}\sum_{cyc}(|ab|-ab)\geq0.$$ Also, we can use AM-GM for three variables.

Indeed, $$a^3+b^3+c^3-3abc=(a+b+c)\sum_{cyc}(a^2-ab)=$$ $$=\frac{9}{2}(a+b+c)\left(\frac{a^2+b^2+c^2}{3}-\left(\frac{a+b+c}{3}\right)^2\right)$$ and since by AM-GM $$a^3+b^3+c^3-3abc\geq0$$ and $$a+b+c\geq0,$$ we are done!

  • $\begingroup$ Is there a simpler way to prove the inequality but still use AM-GM? $\endgroup$ – Alexander B Jun 10 at 8:12
  • $\begingroup$ @Alexander B I posted another way, but I think the first was much more easier. $\endgroup$ – Michael Rozenberg Jun 10 at 8:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.