If you asked me to tell you the stat that is often used that annoys me the most, it is the assist, especially if you use it to criticise a player for lack of assists.
Let me give you two recent examples:
First, we have Daniel Podence’s goal against Chelsea:
Owen Otasowie assisted this. Now, Otasowie does well to nod the ball down, but Podence still has a hell of a lot to do to even craft a shot, let alone a goal - there are at least five Chelsea players between him and the goal. But, the assist goes onto Otasowie’s tally.
Conversely, let’s look at this chance:
Adama Traoré does everything that could be expected of him. He beats Kieran Tierney and stands up a great cross to the far post. But, because Leander Dendoncker does not score, Traoré does not get an assist. It is not Traoré’s fault that Dendoncker doesn’t score. And you could not convince me that Vitinha to Neto is more deserving of an assist than Traoré to Dendoncker.
So, assists do not really tell the full story - it is completely reliant on the teammate being able to stick the ball in the goal.
One measure that could be used is Shot Creating Actions. This, from the excellent FBREF/Statsbomb, is defined as ‘offensive actions directly leading to a shot’. Statsbomb even go as far as to credit up to two players - so in the chance above, both Podence’s pass to Traoré, and Traoré’s cross to Dendoncker get registered.
A second measure is Expected Assists, xA. This looks at the quality of the chance provided. In the examples above, Traoré would get xA of around 0.5, indicating that he created a chance that is scored about 50% of the time. Conversely, Otasowie’s xA is only around 0.05, because Podence would only score about 5% of the time.