Gambling gets a bad rap. There are good reasons for this. The most important reason why gambling is bad is because it makes it hard to tell when you are getting cheated. Even if you are perfect at assessing odds, which most people are not, you have to do a lot of transactions to make sure you are not getting cheated.
In most cases, the "house" takes a cut as well. Whether you think this is fair or not, is up to you. But this is true of savings accounts as well, without gambling.
Our entire banking system is predicated on "trust" that you trust someone to store your money. This requires a complex legal and political structure to prevent fraud, abuse or simply mismanagement.
Let's imagine you are trying to save to buy an electric bicycle for $1,000. We will also imagine, that you have $10/day to set aside for savings.
At that rate, it will take you 100 days to save up for the bicycle.
Now let's imagine that instead, you take that money each day, and bet it at fair odds, to win $1,000. How long should you expect to take to get the money? Well, typically, people use averages, or "expected values" to compute the time it will take, in which case, the time is exactly the same: 100 days. But if instead we look at the median time, we get a completely different answer.
The easy way of computing this, is recognizing, that each day we have a 99% chance of losing the money. To get the odds of successive losses, we simply multiply those values together. So the chance of losing 2 days in a row is:
0.99 * 0.99 =
(1 - 0.01) * (1 - 0.01) =
1 - 2 * 0.01 + 0.0001
1 - 0.02 + 0.0001 =
0.9801
The chance of losing n days in a row is:
0.99^n
So to compute the median time to get $1,000 if you save $10/day, we must simply find the value of n such that
0.99^n < 0.5
Let's try some values
[user@computer ~]$ bc -l
bc 1.07.1
Copyright 1991-1994, 1997, 1998, 2000, 2004, 2006, 2008, 2012-2017 Free Software Foundation, Inc.
This is free software with ABSOLUTELY NO WARRANTY.
For details type `warranty'.
0.9900000^100
.36603234127322950493
0.99000000^50
.60500606713753665044
0.99000000^60
.54715664239076147619
0.99000000^70
.49483865960020725986
0.99000000^69
.49983702989919925238
0.99000000^68
.50488588878706995190
As you can see from those calculations, 69 is the first value for which 0.99^n is less than 0.5, so the median time to win $1,000 dollars betting $10/day, is 69 days! In the median case, you will save 31 days, or an entire month, by betting your money instead of saving it.
The fact is median and averages are different beasts. If you simply save your money, you are guaranteed to have $1,000 by the 100 days. But if you gamble, some times you will win that money on the first day, and sometimes after 1000 days you still won't win the money.
In fact, according to the above, 37% of the time, you won't win the money even after 100 days. So that's the downside. If you square 0.37 again, you get about a 14% chance that you still won't win after 200 days. Going out from there, it gets worse.
Evaluating when gambling is an effective strategy, tells you a lot about how to invest money in general. Basically, gambling will take a certain event, and smear it out over time, so it could happen sooner, later, or never.
If you absolutely need something to happen, then saving is a much more effective strategy, provided you have sufficient time to save the money you need. If you don't have time to save or resources to borrow, your best bet, literally, is to save as much as you can, and then gamble for the rest.
If you only have 99 days to save $100 at $1 a day, then you can save $99 and then you have a 99% chance of winning the last slice of money you need.
So gambling, for certain financial applications, can be a viable alternative to banking. This substantially decreases counterparty risk and moral hazard. In this case the "house" does not need to store any money long term. They only need to find people willing to bid odds on a certain pot size. The "house" doesn't even have to handle any money directly, they can simply require people bidding on the next pool, to pay their bids to the winner of the last pool.
By not handling any money themselves, the "house" protects themselves from both financial regulation and theft. The house can simply work for tips or a small percentage of pot bids go to the house directly instead of the last pool winner. The house may collect the payments from the initial pool to pay for startup costs or "reserves" in case one of the "pools" does not attract sufficient participants.
In this way, by only handling a very small amount of money directly, the house can facilitate many many transactions.
And while this may be prone to corruption or "fixing" schemes, it is also possible to cryptographically ensure results are fair and random.
This kind of system is perfect for use with bitcoin or other cryptocurrencies.
While traditional financial institutions have an important role in financial systems, there are much simpler and slimmer institutions possible, that can take over many of the roles of these traditional institutions. How you use this and what you decide to do with your money is up to you.