According to the documentation of HighwayHash, for finding a collision are expected $m \over 2$ guesses, being $m$ the message.
By contrast, 'strong' hashes such as SipHash or HighwayHash require infeasible attacker effort to find a hash collision (an expected $2^{32}$ guesses of $m$ per the birthday paradox) or recover the seed ($2^{63}$ requests). These security claims assume the seed is secret. It is reasonable to suppose $s$ is initially unknown to attackers, e.g. generated on startup or even per connection.
They are talking above about a keyed hash function with input/output of 64-bits, having the key the same size.
But what if a hash a message of 32-bits in size and a 32-bits key, and keep the message and the key secret, does the effort for finding a collision is $m+k \over 2$? $m$ is the message, $k$ is the key?
What would be the formula to calculate the expected numbers of guesses of key + message for finding a collision?
