No algorithm involved, but when I want to try for a little extra spending money for the $1 state draw games, more often than not I can make a few quick bucks off of the daily keno game.
22 numbers are drawn from a pool of 80, and you get to pick 10 numbers to try to match them to any of the 22 that are drawn. Matching 6 out of 10 wins $7 on a $1 ticket.
I track the delta groups (1-10, 11-20, 21-30, etc or 1-9, 10-19, 20-29, etc). I track how many from each group are represented in each drawing.
Each group periodically will be represented with 4+ of the numbers drawn, such as 4+ from the 40s, 4+ from the 20s, 4+ from the 60s, etc.
Let's say that delta groups 30s and 50s have both gone a relative long while without being represented in 4+ of the numbers drawn. In that case, I play only 30's and 50's, 5 from each group, with the goal being to get a match of 6+ numbers should there be 6+ numbers from the 30s and 50s groups combined.
This effectively narrows my pool down to just 20 numbers, or 2 pools of 10: (30 31 32 33 34 35 36 37 38 39, 50 51 52 53 54 55 56 57 58 59). And then it is a matter of process of elimination, removing half of each group, whichever numbers I think are less likely than the others to be drawn based on various factors.
So let's say I play 30 31 33 35 38 51 52 56 57 59.
And the drawing results are 02 11 13 16 24 27 28 30 33 32 35 46 49 50 51 57 58 59 64 70 77 80.
Or perhaps you accurately predicted exploding 50s but there were only 3 from the 30s group, yet the results are 02 11 13 16 24 27 28 30 32 35 41 46 49 51 52 57 58 59 64 70 77 80.
$7 win on $1 ticket = $6 takeaway.
The only downside is that when playing 10 numbers from just two delta groups, statistically speaking in regards to the history of results, the chances of winning a huge prize are quite slim. Or as another way of putting it, it is a bit rare for 10+ of the 22 numbers drawn to come from just 2 delta groups, let alone for the two groups to be represented equally 5 each. So this isn't an approach I use while fantasizing about the top prize.
Now of course this method isn't reliable enough to risk huge bets by playing many lines at once. Again it works more often than not, which means a few extra plays for the month if I want to extend my lottery bankroll just a smidgen. However, it doesn't work all the time, because it first requires accurately guessing which 2 delta groups will be represented 4+ in the numbers drawn, and only if you guess that correctly can you then significantly reduce you pool of choices in order to significantly increase your chances to make $6.
When it works, it can feel like, "Gosh, what if I had bet $100 worth of this line, I would have made $600!" But when it doesn't work, you'll be glad you only lost $1. When you get a frustrating little unlucky streak and it doesn't work 2-3 times in a row, then you'll really be glad you weren't dropping hundreds but instead are down a whopping $3. After all, down $3 with 3 failed attempts, and then it working on the 4th try, is still a gain of $3 overall.
If you're in it for the thrill of the sport, you can also do $6 for 6 combination varieties from the 2 delta groups, with the goal being that at least one of them will get you the match of 6 and make at least $1 from the $7 prize. It's only a $1 takeaway but is still fun.