Nope - thats your 'random, out of your control' conclusion.
If you we just consider a simple example of a player making a shot. His relevant attributes will determine a result as to whether its on target or off - there will also be a random generated factor (actually there will be more than one) which will be applied to that initial result to modify it. I'd imagine that as in PNP DnD, iirc, that factor can produce a definite fail or success regardless of the initial result - like on a d20 rolling 1 or 20. Ofc its a small chance but a chance nonetheless
In games these exist to ensure there are no guaranteed wins or losses, success or failure - whether its 1/100, 1/20 or 1/16/ or 1/10 or another will affect the frequency of these things.
(There will also be other hidden factors such as the one that determines whether two identical players (attributes-wise) are different performers - I'd call it performance factor. It could be a smaller range from 0.75 to 1.25 (or even smaller) or big from 0.5 to 2.0 (or bigger). That means using the bigger number any player could be half as good or twice as good as his average - good 3*, bad 5*, great 7*, bad 7* - its all relative.)
Simply put, there is always a chance that a 1* succeeds in testing a 9* GK and similarly that a 9* GK fails to stop that 1* shot. Once you get an unfavourable run of chance you get a bad beat.
However, I do think these unexpected things happen too often in this game - too many exulted David's and dead Goliaths.
You yourself have spoken about a higher/lower chance of winning home/away. If that was true they'd achieve it my factoring everything by a say 1.2 or 0.8 factor. It offers an improved chance to win/lose based on location but not too much to rule out the opposite.
PS. all the numbers here are for demonstration purposes