I'll just say this
I'm still reading, learning and practicing, but here's what I think I figured out. When I first took a look at different counting systems, I had never heard of anything like an unbalanced count, but I liked the idea of not having to convert to the true count. I was only familiar with stuff like they had in the movie "21" (balanced counts). I read the KO Blackjack book. I started practicing in Casino Verite, and made my own custom KO strategy (I tweaked a few of the index numbers to make it fit me and my brain). My ups and downs were all over the place. It was tough to get ahead. I played thousands of hands with a $10 to $100 spread under a variety of (mostly good) playing conditions.
Then somewhere I got a tip to read Daniel Dravot's book (The Color of Blackjack), and it was a little pricey for a very small book, but it was basically a modification of KO where instead of completely ignoring the discard pile, what you did was compare your running count to pre-determined yardstick type numbers for the number of decks played. I'll give you an example of that in a minute. But bottom line, when I started taking the number of decks played into account, during my practice of thousands of more hands in CV, I was definitely making much more steady progress upwards, with not so many scary losses of bankroll.
To summarize my practice in Casino Verite: With straight KO, I too frequently went bankrupt with very wild swings. When I did as Dravot indicated to do, I climbed much more steadily.
Dravot pointed out that with KO, when you don't look at the discard pile, there are times in the early part of the shoe where you actually have an advantage, but don't even know about it. By the same logic, there are also times at the end of the shoe when you might think you have an advantage and make a big bet, but you don't actually have an advantage. The weakness of KO: early underbetting, late overbetting.
Dravot's idea didn't involve any pesky true count conversion, just that you had an idea that after 1 deck, your TC needs to meet or exceed a certain number, then after 2 decks dealt, your TC needs to meet or exceed a certain number. And so on, and so forth.
I started to see all the missed opportunities at the beginning of the shoe, and realized I was indeed sometimes placing big bets at the end that were not a good idea. Dravot's betting modifications also made me seem like I was playing more like a typical gambler, and less like a counter. If you use the KO Key Count for your threshold to decide to make bigger bets, you end up flat betting a lot until you suddenly ramp up at the end of the shoe.
After practicing this way for a long time, I got a lot more confidence than with plain KO. I started to get comfortable with looking at the discard pile and instantly knowing well enough how tall it looked. And I realized how important it is to take the number of cards played into account. And I realized how dumb it seems to completely ignore the discard pile. And for now, while I read and learn more, I think I'm going to change to a balanced system. I'm probably going to do Zen. I'm in the process of getting all the true count indexes into Excel, and then making my own rounded indexes for what is easy to remember for me. Zen tracks all the same numbers as KO. Two through Seven, and Ten-Ace.