1: \begin{abstract}

2: A family of random matrices $\boldsymbol{X}^N=(X_1^N,\ldots,X_d^N)$ 

3: converges strongly to a family $\boldsymbol{x}=(x_1,\ldots,x_d)$ in a 

4: $C^*$-algebra if $\|P(\boldsymbol{X}^N)\|\to\|P(\boldsymbol{x})\|$ for 

5: every noncommutative polynomial $P$. This phenomenon plays a central role 

6: in several recent breakthrough results on random graphs, geometry, and 

7: operator algebras. However, strong convergence is notoriously difficult to 

8: prove and has generally required delicate problem-specific methods.

9: 

10: In this paper, we develop a new approach to strong convergence that uses 

11: only soft arguments. Our method exploits the fact that for many natural 

12: models, the expected trace of $P(\boldsymbol{X}^N)$ is a rational function 

13: of $\frac{1}{N}$ whose lowest order asymptotics are easily understood. We 

14: develop a general technique to deduce strong convergence directly from 

15: these inputs using the inequality of A.\ and V.~Markov for univariate 

16: polynomials and elementary Fourier analysis.

17: To illustrate the method, we develop the following applications.

18: \vspace*{.5mm}

19: \begin{enumerate}[leftmargin=*,label=\arabic*.]

20: \addtolength{\itemsep}{.5mm}

21: \item We give a short proof of the result of 

22: Friedman that random regular graphs have a near-optimal spectral gap,

23: and obtain a sharp understanding of the large deviations 

24: probabilities of the second eigenvalue.

25: \item We prove a strong quantitative form of the strong convergence

26: property of random permutation matrices due to Bordenave and Collins.

27: \item We extend the above to any

28: stable representation

29: of the symmetric group, providing many new examples of the strong 

30: convergence phenomenon.

31: \end{enumerate}

32: \end{abstract}

33: