Separating the Survivors from the Casualties


This article originally appeared in the December 1986 issue of UltraRunning Magazine

by Fred Pilon

About a year ago Bill Tharion, a Research Psychologist with the U.S. Army Research Institute of Environmental Medicine in Natick. Massachusetts, was explaining to me some research that he was conducting on the psychological factors that limit endurance performance. Besides just finding out what physical factors determined how well. and for how long, people would perform. the Army was investigating psychological factors that might help them to better train soldiers for “sustained operations,” those missions lasting longer than six hours. I happened to mention to him that ultrarunners faced many of the same factors that soldiers would encounter: extreme. physical fatigue, injuries. loneliness, boredom. etc. I suggested that he should extend his investigations to ultrarunners to help speed up his research.

Consequently, Tharion, Shelley Strowman, Terry Rauch and Barb Shukitt, all Research Psychologists, attended two major ultras —— The Mountain Masochists Trail Run (50 miles) and the Old Dominion 100 Mile Endurance Run. both in Virginia. At both races they administered demographic and psychological questionnaires to the runners before and after the race. With the results of the Mountain Masochists races in hand, Tharion and Rauch did a discriminant analysis to determine which factors best predicted an individual’s likelihood to be a “survivor,” i.e., a finisher. They also did a multiple regression analysis to allow them to predict a finisher’s time.

When they developed their formulas so that had achieved a post-race ability to correctly Classify over 90% of the runners, they headed off to the Old Dominion to administer their questionnaires and to see how well their weighted factors would work for the 100 miler.


How well did their formulas work? Not that great -« they were about 60% correct in predicting whether a runner would finish or not. and they were off an average of three hours in predicting a runner’s time. Still. that’s a lot better than chance. Tharion and Ranch pointed out a few problems in their pilot study. First, the number of participants was extremely small (they hereby request all readers to run the same race at the same time). Second, the psychological questionnaire was too long (people’s minds tend to wander when all the questions seem the same) and too unfocused. And finally, perhaps the questionnaire needs to be administered sometime other than 10-12 hours before the race.

Nonetheless, their questioning did bring out some significant differences between the two groups — survivors (finishers) and casualties (nonfinishers). First, finishers tended to train at a faster pace than nonfinishers. And, secondly. neither group ran more training mileage than the other. However, DNFers had an average longest run of 38.2 miles compared to the of 30.7 miles for the finishers. (Obvious questions still needing answers: Did the better quality, i.e., speed, of the finishers’ runs preclude their running as far as the nonfinishers? Were they at their glycogen limits? Did the casualties tire themselves out too much in their longer training runs?) Third, previous ultra experience counts; those with more experience were more likely to finish than those with less. Among the finishers, ll participants out of 14 had run 10 or more ultras before the Old Dominion. Among the nonfinishers. only 9 out of 25 had 10 or more ultras under their belts.

Rauch observed that self motivation does not seem to be a factor classifying finishers vs. nonfinishers since both groups had similarly high levels of motivation. He also noted that both groups were equal in terms of pre-race tension, anxiety, fatigue and vigor levels. Ranch noted, though. that as individuals the runners reacted quite differently to pressure — some performe.d well while others cracked. Rauch hopes that future questionnaires will allow for better classification of participants.

Good news for all ultrarunners is that an ultra, or even a good part of one, does reduce your level of tension whether you finish or not, according to Tharion and Ranch.

After they refine their questionnaires and formulas Rauch and Tharion hope to reach at least a 75% probability of correctly predicting failure or success and finishing times. Hopefully they will have isolated some factors that ultrarunners can use to improve the odds of achieving our personal dreams.


1 Comment