What to bet?
It took me six complete months of not stepping foot in a casino to answer that question and about a thousand others. The answer depends on like at least a hundred variables.
In fairness, I'll set up a scenario for you. You know basic strategy (BS) to the T because you've researched or simulated what it should be EXACTLY for your specific game. You wrote it over and over again on blank paper until it was burned into your brain. Congradulations, now the house edge is only 0.5% against you (considering we're talking about playing in decent gaming towns).
You went out and bought 6 or 8 decks of cards and started simulating real casino situations while trying to successfully conduct a Hi-Lo running count (RC).
Then you practiced deck estimation via spacial judgement as to whether 1 or more decks were played; 2 or more decks were played; 3 or more decks were played; etc. Then you practiced the RC but now you start dividing it by the whole number of decks left UNSEEN. This quotient is called your true count (TC or T).
Then you researched the "Illustrious 18." These are your 18 most valuable BS deviations. For example, insure when the TC is >= 3. This is number one. Another example might be to double-down (D) an 11 vs. A when the TC is >= 1.
We'll assume you're making at least $25K per year, you can currently afford to lose the $5K you have in your account right now and you can afford to lose an additional $5K per year, hereafter. Thus, your bankroll (BR) is $10K of which, 50% is replenishable and 50% is liquid.
Your TC is in essence an estimate of edge. For every Hi-Lo TC added or subtracted from your default edge represents 0.5% added or subtracted from -0.5%. Thus, you should always stop betting when your TC is <= -1 because the edge against you is 1%. Now, you have part of your answer...
Bet 0 @ TC<= -1. If you have to bet when -1<TC<+1, bet $10 (your BR/1000). When your TC is +1, the house edge is 0%. In terms of how much money you make per hour, it doesn't matter how much you bet at this point. The more you bet, the worse your standard deviation (SD) will be, the worse your risk-of-ruin (RoR) will be and the worse your overall edge will be because your average initial bet (AIB) will be higher. However, zero to $20 is reasonable.
At TC=+2, your edge is 0.5%. 0.5% of $10K is $50. I would suggest less. I'm suggesting $40. From now on, make it simple on us both: Always subtract one from your true count and multiply by $40. For example, the TC is +4. Your edge is 1.5% and you are betting $120.
If you do this EXACTLY, you’ll make a gross average of $20 per 100 hands. This does not include tipping which should be thought of as “camouflage.” Your RoR will be less than 5%. This means that you will have a 5% chance of losing your entire BR