In reading those types of accident reports, it seems that most of the
pilots didn't think about there takeoff performance charts at all.
They did not do a take off calculation. They just thought, "I've
allways made it before." Nor did they think, "Hey, I'm halfway down
the runway. I should be flying by now, better abort."

For a glider takeoff. It would be interesting to have good data on
expected takeoff rolls and climb rates. Each set of data would be
specific to towplane-glider pairs. Takeoff surface and winds would
have to be carefully recorded in addition to a GPS log with location
and airspeeds.

Any calculations would be suspect until experimentally verified. So
skip the calculations and start recording data. Get a flight log of
every takeoff. Try and quantify surface types, towplane ID and
winds. Then plot the data and make your conclusions.

And then use the data with a grain of salt.

Todd Smith
3S