tELO is mathematical model to compute relative rankings of players competing in the same tennis league. It is adapted from the ELO, which has historically been adoped in the chess community.

Unlike conventional pointing systems - which award a fixed number of points to the winner - the tELO looks at who the loser was before bestowin a number of points to the winner commensurate to the size of feat. 

Was the opponent bottom of the table? It'll be change points, if any. Was it the star player who was 10 and 0? Here are a lot more tELO points.

And of course, the score also matters. Bagel will award more points, whereas forfeits or TB wins less so.

Mathematically, the ELO first estimates the probability of player A winning given the relative ranking of both player A and its opponent, player B, as:

P(A) = 1 / (1 + 10^(-(R_A - R_B) / s))

P(B) follows as:

P(B) = 1 - P(A).

The log loss will then be:

-log(P(A))

[Need a math editor]