Hi it's one of my old problem that I cannot solve let me propose it :
Prove that $$\Gamma\left(3^{-3}+1\right)<\Gamma \left(\operatorname{W}(3)^{-\operatorname{W}(3)}+1\right)\quad \tag{1}$$ where $W(\cdot)$ is the Lambert W function and $\Gamma(\cdot)$ is the gamma function.
The central function is :
$$f(x)=\Gamma\left(x^{-x}+1\right)$$
The derivative is :
$$f'(x)= -x^{-x}(\ln(x)+1)\Gamma\left(x^{-x}+1\right)\psi(x^{-x}+1)$$
Unfortunately it doesn't helps since $f(x)$ is increasing around $x=3$ while $f\left(\operatorname{W}(x)\right)$ is decreasing around $x=3$ .
Maybe I think we can do the same as the Claude Leibovici's answer here Show that $\Gamma(\Omega)\leq \Gamma\Big(\operatorname{W}\Big(x^{x}\Big)\Big)<2$ on $(0,1]$ .
My question :
How to show $(1)$ ?
Thanks in advance !
Mking the problem more general, considering $$\Delta(x)=\Gamma \left(\operatorname{W}(x)^{-\operatorname{W}(x)}+1\right)-\Gamma\left(x^{-x}+1\right)$$ around $x=e$, use Taylor expansions to order $O\left((x-e)^{n+1}\right)$
For example $$x^{-x}+1=\left(1+e^{-e}\right)-2 e^{-e} (x-e)+\frac{1}{2} e^{-1-e} (4 e-1) (x-e)^2+O\left((x-e)^3\right)$$ gives $$\Gamma\left(x^{-x}+1\right)=\Gamma \left(1+e^{-e}\right)-2 (x-e) \left(e^{-e} \Gamma \left(1+e^{-e}\right) \psi ^{(0)}\left(1+e^{-e}\right)\right)+\frac{1}{2} e^{-1-2 e} (x-e)^2 \Gamma \left(1+e^{-e}\right) \left(-e^e \psi ^{(0)}\left(1+e^{-e}\right)+4 e^{1+e} \psi ^{(0)}\left(1+e^{-e}\right)+4 e \psi ^{(0)}\left(1+e^{-e}\right)^2+4 e \psi ^{(1)}\left(1+e^{-e}\right)\right)+O\left((x-e)^3\right)$$
Similarly, $$W(x)=1+\frac{x-e}{2 e}-\frac{3 (x-e)^2}{16 e^2}+O\left((x-e)^3\right)$$ $$\log(W(x))=\frac{x-e}{2 e}-\frac{5 (x-e)^2}{16 e^2}+O\left((x-e)^3\right)$$ $$-W(x)\log(W(x))=-\frac{x-e}{2 e}+\frac{(x-e)^2}{16 e^2}+O\left((x-e)^3\right)$$ $$\operatorname{W}(x)^{-\operatorname{W}(x)}=e^{-W(x)\log(W(x))}=1-\frac{x-e}{2 e}+\frac{3 (x-e)^2}{16 e^2}+O\left((x-e)^3\right)$$
The formulae are already messy but workable. Computing the numerical value of $\Delta_n(3)$ as a function of the order $n$ of Taylor expansions, we have
$$\left( \begin{array}{cc} n & 10^6 \times \Delta_n(3) \\ 1 & -4892.218349 \\ 2 & +265.5544022 \\ 3 & +68.15191015 \\ 4 & -13.15976659 \\ 5 & +6.138334905 \\ 6 & +4.440578904 \\ 7 & +4.374006576 \\ 8 & +4.408250996 \\ 9 & +4.405368867 \\ 10 & +4.405111061 \\ 11 & +4.405194838 \\ 12 & +4.405189100 \\ \cdots & \cdots \\ \infty & +4.405188465 \end{array} \right)$$
In fact $\Delta(x)=0$ for $x=3.000042164$.
But, there is another to this equation close to $x=\frac 1e$. A first order Taylor expansion gives it equal to $0.44121$ while the exact solution is $0.42655$.