Top metrics I look for to evaluate an algorithmic system are the following;

Max Drawdown:

From your highest equity peak to the lowest trough. Anything greater than 30% is a no go for me. Exceptions here apply, but you really want to minimize this value. It's important that when the system isn't making money, it's either not trading or treading water. Prolonged periods of drawdown will be hard to sit through once it goes live. You'll have to watch your money dwindle away and hope the robot can course correct. It will also make it incredibly difficult to determine when the strategy has stopped working.If you have to lose 60% of your equity before you know 'forsure' it's not functioning anymore, that's a tough spot.

Profit Factor:

The profit factor is defined as the gross profit divided by the gross loss (including commissions) for the entire trading period. I call this the "comfort level". Anything less than 2 will be uncomfortable to watch trade. It'll be losing and winning all over the place and it suggests your edge is fairly small. It should come as no surprise that a small profit factor will have larger drawdowns. Anything less than 1 and you're not profitable at all.

Number of Trades & Win/Loss Ratio:

Profit factor will tell me a lot about how much bigger my winners are to my losses, however win/loss ratio as well as the total number of trades will tell me about how the system behaves. You say you have a 64% success rate? That's phenomenal and definitely profitable. This is also a good metric to consider when you think about commissions. If I'm trading 20 times a day and winning 12/20 of them, my broker is going to be my best friend and probably take most of my profits along the way.

This is a game of:

Idea Creation > Testing & Implementation > "Success" & Euphoria > Inevitable Recognition of Error > Despair > Idea Creation.

Every quant dreams of breaking that cycle and developing something that can generate alpha on it's own. A lot of people will try, few will persist, even fewer will truly succeed. I have yet to break this cycle. Every quarter Nami is rebalanced and the capital allocation is reset. We've had quarters of 65% drawdown and we are intent on taking risky strategies. Many of our strategies from years passed would be losing money hand-over-fist or bit-over-byte if they were left to trade today. Every strategy Nami executes is entirely technical and will routinely take egregious amounts of leverage on its trades. Don't put you HELOC on these strategies, put your 500$ in craps winnings here instead.

My best "Risk ON" strategy has a max drawdown of 12-45% and a profit factor of 1.7-2.2. It varies based on how much leverage I allow it to give itself. It trades 6 different crypto currencies.