[Jhob94]

What's the best? (and why)

enum _:Types
{
    Connected,
    Alive,
    Retarded
}

new bool:g_Bools[MAX_PLAYERS + 1][Types]

public client_putinserver(id)
    g_Bools[id][Connected] = true

or:

new bool:g_bConnected[MAXPLAYERS + 1], bool:g_bAlive[MAXPLAYERS + 1], bool:g_bRetarded[MAXPLAYERS + 1]

public client_putinserver(id)
    g_bConnected[id] = true

[Bugsy]

The first method is more organized and potentially could be faster but I'm not 100% on that.

My thought: In the first one, everything is stored in one array so when accessing data you only need to seek the address of the variable then offset it by ( id * size ) + item. Using separate variables, each variable is stored at a different memory address. In most cases this doesn't matter, but there are scenarios where it's applicable. Would you ever notice a difference, no.

[HamletEagle]

I prefer the first one, but this is a subjective opinion. I think that the difference, if it exists, is trivial because you initialize the same number of variables(even if in one array same ammount memory is allocated - someone correct me if I'm wrong).