Should $3\!-\!$pointers be worth $2.5{\it ?}$ 
I was seeking an alternative scoring rule sets instead of three points for a win (gained more engaging and balanced) to cancel the theory "Banking a draw meaning a new kind of loss".

It's kind of unreal to assess the consequences of individual changes on game dynamics and appeal. The changes allow for novel strategic and tactical patterns to emerge, while keeping the games close to the original. Overviewing Premier League's seasons from 2013 through 2022, I realized this subtle relationship between Predicted points and Goals differential in cases:
. 3.0 points for a win
$\texttt{Predicted points}= .63\cdot\texttt{Differential}+ 52$
. 2.5 points for a win
$\texttt{Predicted points}= .52\cdot\texttt{Differential}+ 45$
. 2.0 points for a win
$\texttt{Predicted points}= .41\cdot\texttt{Differential}+ 38$
Because $\sum\texttt{Differential}= 0$ so in the system of 2.5 points for a win, its slope meaning the position of the "median" team (52th percentile, $52\%$), its intercept meaning the average of the "median" team (48th percentile, $= 45/95= 48\%$). That's why I doubted it would be, maybe much better. (The probability of a draw is different from a quarter according Skellam distribution.) I wish to see your discovery demonstrate the rich possibilities lying beyond the rules of modern football.
 A: I am also not sure whether I understand the point of the question, so I reply to both possibilities.
It should be noted that the rating system is not independent from the teams' performance. In other words, changes to the rating system would also impact the outcome of each match.
Therefore, in my opinion, there is no really fair way to predict outcomes to decide the winner of an non-complete competition, as it would have changed their performance if the teams would have known about the modified scoring system. (at least if we can agree that for a competition to be "fair", all participants must know the rules of the system)
Similarly, 2.5 points for a win are not more "fair" or "better" than 3 points. It simply gives a different incentive for winning/drawing.
A: We could compute the team performances based on an ELO ranking system. If you are considering non-complete competitions (each team has not seen exactly all of the other teams) then such a scheme would be more fair.
One such ELO ranking derived system could be fitting a logistic regression model to the points scored.
Aside from its use to fairly decide on a winner in a non-complete competition, one can also use such a system to come up with a better ratio for the points of a win:draw which is currently 3 in soccer. We can add this ratio as a parameter in the fit and find out which ratio maximizes the likelihood. That ratio would be a balance that makes the score of a match more 'predictable' and more inline with the difference in the two team abilities.
