Mathematical Models Predicting Performance in Track and Field at the 2011, 2013 and 2015 IAAF World Championships

Ian Heazlewood, Joseph Walsh

    Research output: Chapter in Book/Report/Conference proceedingConference Paper published in ProceedingsResearch

    Abstract

    Mathematics and science are based on principles of description and importantly prediction. The ability to make substantive and accurate predictions of future elite sportsperformances indicates such approaches reflect “good” science. Often these predictions arepurely speculative and are not based upon any substantial evidence; however they providecriteria for athletes in terms of performance standards to be achieved to be competitiveinternationally, resetting international qualifying standards for IAAF sanctioned competitions such as World Championships, to focus on limits of human athletic performance andevaluating long term adaptations of athletes to chronic training. The research aim wasto further develop predictive mathematical models using IAAF World Championship datain athletics outdoor championships 1983 to 2009 (previous model) for men’s and women’s100m, 400m, long jump and high jump and then evaluate a revised model based on theaddition of 2011 and 2013 data to predict the 2015 performances. Data sets consisted ofaverage scores of the top eight performances in the finals of each event. The mathematical approach utilised regression-curve estimation time series analysis by evaluating datafits to linear, logarithmic, inverse, quadratic, cubic, compound, power, sigmoidal, growthexponential and logistic functions. The calculations were conducted using SPSS Version22 software and goodness of fit by R2, significance, accuracy of prediction and residuals.The results for mens and womens events indicated cubic functions consistently displayedthe best fits with R2 values for 100m (.723), 400m (.336), long jump (.221) and highjump (.920) for men and 100m (.366), 400m (.526), long jump (.692) and high jump (.496)for women. The inclusion of the 2011 and 2013 data improved the fits very marginally.The outcome was, although some variability existed in R2 the predictions for events wereaccurate indicating in the majority of events the cubic functions displayed good model fits.
    Original languageEnglish
    Title of host publicationProceedings of the 5th International Conference on Mathematics in Sport
    EditorsAnthony Kay, Alun Owen, Ben Halkon, Mark King
    Place of PublicationUK
    PublisherUnknown
    Pages60-65
    Number of pages6
    Publication statusPublished - 2015
    EventInternational Conference on Mathematics in Sport (2015 5th) - Loughborough University, U.K., Loughborough, United Kingdom
    Duration: 29 Jun 20151 Jul 2015
    Conference number: 2015 (5th)

    Conference

    ConferenceInternational Conference on Mathematics in Sport (2015 5th)
    CountryUnited Kingdom
    CityLoughborough
    Period29/06/151/07/15

    Fingerprint

    prediction
    resetting
    time series analysis
    mathematics
    logistics
    world
    software
    woman
    science
    calculation

    Cite this

    Heazlewood, I., & Walsh, J. (2015). Mathematical Models Predicting Performance in Track and Field at the 2011, 2013 and 2015 IAAF World Championships. In A. Kay, A. Owen, B. Halkon, & M. King (Eds.), Proceedings of the 5th International Conference on Mathematics in Sport (pp. 60-65). UK: Unknown.
    Heazlewood, Ian ; Walsh, Joseph. / Mathematical Models Predicting Performance in Track and Field at the 2011, 2013 and 2015 IAAF World Championships. Proceedings of the 5th International Conference on Mathematics in Sport. editor / Anthony Kay ; Alun Owen ; Ben Halkon ; Mark King. UK : Unknown, 2015. pp. 60-65
    @inproceedings{f6c8e8072d2241aab714871443f37cb6,
    title = "Mathematical Models Predicting Performance in Track and Field at the 2011, 2013 and 2015 IAAF World Championships",
    abstract = "Mathematics and science are based on principles of description and importantly prediction. The ability to make substantive and accurate predictions of future elite sportsperformances indicates such approaches reflect “good” science. Often these predictions arepurely speculative and are not based upon any substantial evidence; however they providecriteria for athletes in terms of performance standards to be achieved to be competitiveinternationally, resetting international qualifying standards for IAAF sanctioned competitions such as World Championships, to focus on limits of human athletic performance andevaluating long term adaptations of athletes to chronic training. The research aim wasto further develop predictive mathematical models using IAAF World Championship datain athletics outdoor championships 1983 to 2009 (previous model) for men’s and women’s100m, 400m, long jump and high jump and then evaluate a revised model based on theaddition of 2011 and 2013 data to predict the 2015 performances. Data sets consisted ofaverage scores of the top eight performances in the finals of each event. The mathematical approach utilised regression-curve estimation time series analysis by evaluating datafits to linear, logarithmic, inverse, quadratic, cubic, compound, power, sigmoidal, growthexponential and logistic functions. The calculations were conducted using SPSS Version22 software and goodness of fit by R2, significance, accuracy of prediction and residuals.The results for mens and womens events indicated cubic functions consistently displayedthe best fits with R2 values for 100m (.723), 400m (.336), long jump (.221) and highjump (.920) for men and 100m (.366), 400m (.526), long jump (.692) and high jump (.496)for women. The inclusion of the 2011 and 2013 data improved the fits very marginally.The outcome was, although some variability existed in R2 the predictions for events wereaccurate indicating in the majority of events the cubic functions displayed good model fits.",
    author = "Ian Heazlewood and Joseph Walsh",
    year = "2015",
    language = "English",
    pages = "60--65",
    editor = "Anthony Kay and Alun Owen and Ben Halkon and Mark King",
    booktitle = "Proceedings of the 5th International Conference on Mathematics in Sport",
    publisher = "Unknown",

    }

    Heazlewood, I & Walsh, J 2015, Mathematical Models Predicting Performance in Track and Field at the 2011, 2013 and 2015 IAAF World Championships. in A Kay, A Owen, B Halkon & M King (eds), Proceedings of the 5th International Conference on Mathematics in Sport. Unknown, UK, pp. 60-65, International Conference on Mathematics in Sport (2015 5th), Loughborough, United Kingdom, 29/06/15.

    Mathematical Models Predicting Performance in Track and Field at the 2011, 2013 and 2015 IAAF World Championships. / Heazlewood, Ian; Walsh, Joseph.

    Proceedings of the 5th International Conference on Mathematics in Sport. ed. / Anthony Kay; Alun Owen; Ben Halkon; Mark King. UK : Unknown, 2015. p. 60-65.

    Research output: Chapter in Book/Report/Conference proceedingConference Paper published in ProceedingsResearch

    TY - GEN

    T1 - Mathematical Models Predicting Performance in Track and Field at the 2011, 2013 and 2015 IAAF World Championships

    AU - Heazlewood, Ian

    AU - Walsh, Joseph

    PY - 2015

    Y1 - 2015

    N2 - Mathematics and science are based on principles of description and importantly prediction. The ability to make substantive and accurate predictions of future elite sportsperformances indicates such approaches reflect “good” science. Often these predictions arepurely speculative and are not based upon any substantial evidence; however they providecriteria for athletes in terms of performance standards to be achieved to be competitiveinternationally, resetting international qualifying standards for IAAF sanctioned competitions such as World Championships, to focus on limits of human athletic performance andevaluating long term adaptations of athletes to chronic training. The research aim wasto further develop predictive mathematical models using IAAF World Championship datain athletics outdoor championships 1983 to 2009 (previous model) for men’s and women’s100m, 400m, long jump and high jump and then evaluate a revised model based on theaddition of 2011 and 2013 data to predict the 2015 performances. Data sets consisted ofaverage scores of the top eight performances in the finals of each event. The mathematical approach utilised regression-curve estimation time series analysis by evaluating datafits to linear, logarithmic, inverse, quadratic, cubic, compound, power, sigmoidal, growthexponential and logistic functions. The calculations were conducted using SPSS Version22 software and goodness of fit by R2, significance, accuracy of prediction and residuals.The results for mens and womens events indicated cubic functions consistently displayedthe best fits with R2 values for 100m (.723), 400m (.336), long jump (.221) and highjump (.920) for men and 100m (.366), 400m (.526), long jump (.692) and high jump (.496)for women. The inclusion of the 2011 and 2013 data improved the fits very marginally.The outcome was, although some variability existed in R2 the predictions for events wereaccurate indicating in the majority of events the cubic functions displayed good model fits.

    AB - Mathematics and science are based on principles of description and importantly prediction. The ability to make substantive and accurate predictions of future elite sportsperformances indicates such approaches reflect “good” science. Often these predictions arepurely speculative and are not based upon any substantial evidence; however they providecriteria for athletes in terms of performance standards to be achieved to be competitiveinternationally, resetting international qualifying standards for IAAF sanctioned competitions such as World Championships, to focus on limits of human athletic performance andevaluating long term adaptations of athletes to chronic training. The research aim wasto further develop predictive mathematical models using IAAF World Championship datain athletics outdoor championships 1983 to 2009 (previous model) for men’s and women’s100m, 400m, long jump and high jump and then evaluate a revised model based on theaddition of 2011 and 2013 data to predict the 2015 performances. Data sets consisted ofaverage scores of the top eight performances in the finals of each event. The mathematical approach utilised regression-curve estimation time series analysis by evaluating datafits to linear, logarithmic, inverse, quadratic, cubic, compound, power, sigmoidal, growthexponential and logistic functions. The calculations were conducted using SPSS Version22 software and goodness of fit by R2, significance, accuracy of prediction and residuals.The results for mens and womens events indicated cubic functions consistently displayedthe best fits with R2 values for 100m (.723), 400m (.336), long jump (.221) and highjump (.920) for men and 100m (.366), 400m (.526), long jump (.692) and high jump (.496)for women. The inclusion of the 2011 and 2013 data improved the fits very marginally.The outcome was, although some variability existed in R2 the predictions for events wereaccurate indicating in the majority of events the cubic functions displayed good model fits.

    UR - http://www.mathsportinternational.com/previous.html

    M3 - Conference Paper published in Proceedings

    SP - 60

    EP - 65

    BT - Proceedings of the 5th International Conference on Mathematics in Sport

    A2 - Kay, Anthony

    A2 - Owen, Alun

    A2 - Halkon, Ben

    A2 - King, Mark

    PB - Unknown

    CY - UK

    ER -

    Heazlewood I, Walsh J. Mathematical Models Predicting Performance in Track and Field at the 2011, 2013 and 2015 IAAF World Championships. In Kay A, Owen A, Halkon B, King M, editors, Proceedings of the 5th International Conference on Mathematics in Sport. UK: Unknown. 2015. p. 60-65