In the lottery I participate in, the odds of winning the jackpot has been stated as being 1:5.4 million per row. Each coupon you hand in has 10 rows, where each row is a combination of 7 numbers out of the range 1 to 34.
The way I see it, if you chose the same combination 10 times, your odds are essentially 1:5.4 million, but if you choose 10 discrete combinations, your odds are essentially 1:540.000 instead.
So my question is, how do you select 10 combinations that are discrete enough to count as 10 unique combinations to maximize your odds?
If the lottery is fair then all results are equally likely. Just pick any $10$ different ones.
Suppose that the game is throwing a fair die and you can bet on any two numbers. Do you expect that some pairs are better than others? Or would $1$ and $2$ be just as likely as any other?
You know that there are approximately $5.4$ million possibilities. So, just label each result with a number from $1$ to $5379616$. Each of these labels will be equally likely. Now, do you feel the need to pick the $10$ by any rule or just pick any $10$?
The way that the result is generated is just mystique. In reality, it is just a way to generate a random number from $1$ to $5379616$.
One non-mathematical tip is: don't pick obvious patterns and avoid numbers that might appear in dates. These aren't any less likely to win but it is more likely that someone else has picked them and you would need to share the prize.
Another tip: don't play at all. In most places, it is easy to calculate the expected return as the numbers of tickets sold and the prize fund are usually published. 50% is quite good for a lottery. If this data is not published then definitely stay away.
If you must gamble then you might get much better value (less negative) elsewhere e.g. a casino.