I have written a little program that draws league tables as a sort of graph that helps predict the final standings. Each team has a line that begins at the teams current number of points (marked "0") and ends at its maximum number of points ("3"). I also assess crude probabilites for the remaining matches using those to guess each teams most likely number of finishing points ("X"). In a few hours the last two matches for this weekend kicks off. The probabilities assessed for them are as follows...
Those should seem not too far off to most. They are tendencies deducted from the current standings. Looking at a bookmakers site I get the probabilities of 63%, 24%, 14% and 44%, 28, 27% respectively. Not very far from each other except that the bookies seem to have more confidence in Leverkusen than the past season suggests. The graph is a bit more interesting than usual because all teams haven't played the same number of matches. Leverkusen are placed 8th but it is immediately clear that they are likely to be placed 7th tonight. It also appears that from Dortmund an up relegation is extremely unlikely. Medals could theoretically end with teams as lowly placed as Köln currently - but that many changes in the standings are always unlikely. Those conclusions are among the more cautious one could make. Four teams (from 2nd to 5th: Bayern München, Hoffenheim, Wolfsburg and Hamburger SV) are level at 42 points. But Hoffenheim is predicted to finish with 59 points, the others at 60. While this is just the "most likely" finishing points and the distribution really should be regarded as a hint then it does suggest that Hoffenheim has a bit harder opposition left. There is a lot of room for improvement. A lot. And I do change the script every now and then. Suggestions appreciated.