The rock sequence itself. If all sedimentary rocks were laid down by a global flood, there would be an expected sequence - clastics, sandstones, siltstones, shales - and maybe a smattering of limestone at the top - all over the world, at the same time. Instead, we have a broad range of intermixed rock sequences indicated countless transgressions and climatic changes. And then we have sedimentary rocks uplifted, folded, and then eroded, with subsequent additional sedimentary rocks, sometimes with a younger sequence deposited upon an eroded, angled sequence (such as the Great Unconformity in the Grand Canyon). And not all sedimentary-rock is based upon marine deposition - you have terrestial dune sandstones that are hundreds to thousands of feet thick out in the Colorado Plateau - you have fossil coral reefs that are hundreds of feet thick, and, since we can study the rate of coral growth and accumulation in the present, it isn't hard to determine that such thick limestone sequences are millions of years old.
So how do you propose that the flood could deposit a sequence of sedimentary rocks, tilt them, erode them UNDERWATER (a depositional, not an erosional environment), and then deposit more sedimentary rocks of a completely different type.