BJ,
Its wrong to assume by backing every runner in every race that the money return equals the total less Takeout.
Try this exercise, total the number of runners at each meeting then total the Payoff for each winner, then do this for a number of meetings an see what your results tell you...this is an important subject when one studies the Market Efficiency of Wagerers at each track(if you take it a step further, you can do it by each day of the week).
On your other query,
If you produce ratings that show individual Standard Deviations based on LTD performances, it will become apparent that races that have a tigher overall average STDev are Golden, it doesn't matter on the races actual Classification, its the runners ability to be consistent, the STDev is your tool in this cluster type of analysis