I found a paper online by Elliott Hollield, Victoria Trevino, and Adam Zarn (link) that uses survival analysis to estimate how many Olympic records would be broken in the 2012 Olympic games. Here is the paper abstract:
We use recurrent-events survival analysis techniques and methods to analyze the duration of Olympic records. The Kaplan-Meier estimator is used to perform preliminary tests and recurrent event survivor function estimators proposed by Wang & Chang (1999) and Pena et al. (2001) are used to estimate survival curves. Extensions of the Cox Proportional Hazards model are employed as well as a discrete-time logistic model for repeated events to estimate models and quantify parameter signicance. The logistic model was the best t to the data according to the Akaike Information Criterion (AIC). We discuss, in detail, covariate signicance for this model and make predictions of how many records will be set at the 2012 Olympic Games in London.
Here’s what they predicted:
After prleminary measures, we estimated ve dierent models with signicant covariates. Four of these models were extensions of the Cox Proportional Hazards model while the other was a discrete-time logistic model for repeated events. In the end, the logistic model was the best t to our data based on the Akaike Information Criterion, and it depended primarily on the following covariates: SameAthlete, WorldRecord, PCC, MRI, and TRI. We also used survival estimates from three dierent recurrent event survivor function estimators to determine the number of new records that will be set in the 2012 Olympics. In 51 of the 63 events we considered in track & field, canoeing, cycling, and swimming, the estimated number of Olympic records that will be broken is somewhere between 20.12 and 31.14.
(Note: I added the part in italics)
This site lists the Olympic and world records broken in the Olympics:
Number of world records broken = 13 (4 in Track&Field, 3 in Cycling (Track), and 6 in Swimming)
Number of Olympic records broken = 29 (10 in Track&Field, 5 in Cycling (Track), and 14 in Swimming). This estimate is within the interval found in the paper. Even more impressive: the acknowledgments indicate that at least one of the authors was an undergraduate. Kudos!