Hoping to Learn From my Error

So I designed and built two custom PCBs of my own design with no previous background in electronics before starting this project only a couple months ago. My main PCB would have worked just fine if I didn’t somehow get dislexic and switch the GND and power out pins on my voltage regulator…but a couple quick jumpers and scratched out traces later it works fine.
However for my second PCB, which goes along with the same project, the process of figuring out the problem was much more difficult for me to discover and eventually fix. I don’t however totally understand why it is working now. I used FreePCB to make my boards and I made a couple bitmaps of the original board (bad one) and how it should look (good one).

Bad One

Good One

I’m using a 12V battery to supply power and that’s getting stepped down to 9V on my first board, then the 9 volts goes to some LEDS and also wires to this second board where a Really Bare Bones Board (Head) steps it down to 5V and I use the 5 volts to power a RF receiver (labeled RF) and then use wires to send that 5 volts back to another portion on my main board to power my shift registers.
I scratched out the GND trace lower right on the board, and removed it from the PCB in the pic. This solved the problem and my set-up works perfectly now, which is a great feeling. I’m just lloking to learn from this for furture projects hopefully to avoid making the same mistakes. Why did cutting that GND trace solve the problem? Aren’t GND runs supposed to be together?

Many thanks for anyone who can relay any insight.

It rather difficult to understand what you're trying to do from just the PCB layout. Do you have a schematic?

Anachro you probably will remember this schematic, and yes I remember that my LEDs are backwards in the schematic layout.

So 12v from the battery, into the rgulator with an on/off switch (SW), putting out 9v. One line of 9v goes to the LEDs, through the transistor when the transistor base is triggered, and back to ground. The other line of 9v goes to the RBBB which is doing all the controlling of the light programs. The RBBB knocks the 9v down to 5v with its on board regulator, and you can see the 5v from the RBBB feeds the RF receiver and gets sent to power the 74HC595's. All of the traces labeled D5, D6, ect. are short for digital 5, digital 6 of the Atmega 168. I have the RF receiver connected to digital pins 2 and 3 of the Atmega.

I hope that's enough info, that's all I can think of right now that would be of use.

Yes, I would have thought that all Grounds should be wired together in a circuit like this. Are you sure you have the correct pin-outs for all the connectors? And when the circuit didn't work, in what way did it not work?

Yes all the pins are correct as it is working great now, just as it worked great on my breadboard earlier. I have signal A and B from the RF receiver going to pins 2 & 3 of the Atmega to trigger a ISRs that advance and reverse the light patterns with each press of the button...A for forward, B for reverse. When the receiver senses the button being pressed (it's a momentary remote so it puts out 5v as long as the button is held) it will put out 5v to the respective pin and the Atmega will sense this and trigger the appropriate ISR. While in the breadboard (it worked) the only thing I can see different it that the GND trace that I removed in the pic and scratched out on the PCB was not involved anymore. While troubleshooting the RF receiver would not put out 5v on either the A or B pins when plugged into the PCB. However place the receiver into the breadboard and that problem GND trace was effectively removed via not placing a jumper wire there, the thing lit up and worked great. So as of right now by scratching out that trace the LEDs light up and animate according to the code in the Atmega, and the RF receiver puts out 5v on A or B pins just great depending upon which button is pressed. It works so I shouldn't complain I guess but it still leaves me wondering and saying WTH!? :(

Need more info on your RF Receiver module. Datasheets?

So here are scans of the datasheet supplied to me. I didn't scan page 3 as it had one line of non-relevant text on it. There is some of my writing on the paper FYI. And the other.

So how did you determine the pinouts that you hand-wrote along the photograph, if the pin number fields on the second page were completely blank?

For a moment there how I had the pin outs written down had me drawing a blank...thought I was a damn magician! Then I remembered that on the underside of the PCB they had labled them with a silkscreen which pin was which. I transferred this info on to the sheet when I was making my own PCB as it was much easier than flipping the thing over to look at the silkscreen all the time.

And any clue what pin VT-receiving instruction is about or does? I don't have it connected to anything and the system works, I did note however when troubleshooting that I mixed up my pins when testing with the multi-meter (wanted to see if any volts came out of pin A or B but instead had it on VT) and whenever any button was pressed the VT pin would put out 5v.