School of Continuous Improvement researchers modelled the standing ovation as a seat-by-seat cascade through whatever part of a hall each patron can see, and found the floor rake accounts for most of the variation while the running order accounts for almost none.
A seat-level model calibrated on 96 end-of-year concerts across 21 community venues, and the running-order instrument a regional association has adopted from it
A regional concert association will print a predicted ovation likelihood beside every act on its programmes this season, reported to two decimal places.
The figure comes from STANCE, a paper released today by the School of Continuous Improvement, which treats a standing ovation as something that spreads seat by seat through whatever part of a hall each patron can actually see. Two observers per performance recorded row-level standing counts at 96 end-of-year concerts across 21 community venues, and the trajectories they brought back were used to fit the model.
What the model attributes the ovation to is the room. The slope of the floor carries the largest share of the variation in whether a hall gets to its feet; the running order, which is the thing committees revise each season, carries close to none of it. Removing the term for how strong the act was moved the modelled result by about a tenth of a second.
We set out to sharpen a decision and found the decision was not where the room believed it was. The School has since had cause to put the same question to one or two of its own.
— Dr Thandiwe Solberg, Senior Lecturer and Convenor, Demo Quarter
The paper adds to a body of work the University has been assembling one ordinary venue at a time, and its authors expect the instrument to travel further than the finding did. Dr Osei Vandermeer, Research Fellow and Convenor of the Review Cadence Observatory, said a model that returns a figure for every slot “will be asked for a figure for every slot, and the School is interested in what becomes of figures like that afterwards”.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.5dlngx.