It was recently shown in [4] that, for L 2 -approximation of functions from a Hilbert space, function values are almost as powerful as arbitrary linear information, if the approximation numbers are square-summable. That is, we showed thatwhere e n are the sampling numbers and a k are the approximation numbers. In particular, if (a k ) ∈ ℓ 2 , then e n and a n are of the same polynomial order. For this, we presented an explicit (weighted least squares) algorithm based on i.i.d. random points and proved that this works with positive probability. This implies the existence of a good deterministic sampling algorithm.Here, we present a modification of the proof in [4] that shows that the same algorithm works with probability at least 1 − n −c for all c > 0.Let H be a Hilbert space of real-or complex-valued functions on a set D such that point evaluationis a continuous functional for all x ∈ D, which are usually called reproducing kernel Hilbert spaces. We consider numerical approximation of functions from such spaces, using only function values. We measure the error in the space L 2 = L 2 (D, A, µ) of square-integrable functions with respect to an arbitrary measure µ such that H is embedded into L 2 . This means that H consists of square-integrable functions such that two functions that are equal µ-almost everywhere are also equal point-wise.