You are wrong, in a computer "Today is Monday" uses 14 bytes = 92 bits Hufman uses 49 bits, so its shorter. You have to check "how much it really cost" to the computer your code you cannot base it on the preset it seems to use only. But is true that in some cases will not be useful where there is NOT too many bits. Also it works better when you have 2^n amount of different posibilities. In such case it is better to use base 10 (because you have only 10 different characters) and then you will have a number of 10^14 where you need 47 bits. And that means hufman encoding is not always the best (that is why there are so many compression algorithms)