Study says we’re over the hill at 24 - Public Affairs and Media Relations

/study-says-we_re-over-the-hill-at-24.ht

  • C’est à partir de 24 ans que se dégradent les performances motrices et cognitives. La preuve par #Starcraft

    Study says we’re over the hill at 24 - Public Affairs and Media Relations - Simon Fraser University
    http://www.sfu.ca/pamr/media-releases/2014/study-says-we_re-over-the-hill-at-24.html

    It’s a hard pill to swallow, but if you’re over 24 years of age you’ve already reached your peak in terms of your cognitive motor performance, according to a new Simon Fraser University study.

    SFU’s Joe Thompson, a psychology doctoral student, associate professor Mark Blair, Thompson’s thesis supervisor, and Andrew Henrey, a statistics and actuarial science doctoral student, deliver the news in a just-published PLOS ONE Journal paper.

    In one of the first social science experiments to rest on big data, the trio investigates when we start to experience an age-related decline in our cognitive motor skills and how we compensate for that.

    The researchers analyzed the digital performance records of 3,305 StarCraft 2 players, aged 16 to 44. StarCraft 2 is a ruthless competitive intergalactic computer war game that players often undertake to win serious money.

    Their performance records, which can be readily replayed, constitute big data because they represent 870 hours worth of strategic real-time cognitive-based moves performed at varied skill levels.

    Using complex statistical modeling, the researchers distilled meaning from this colossal compilation of information about how players responded to their opponents and more importantly, how long they took to react.

    La publication originale est accessible ici http://www.plosone.org/article/info:doi/10.1371/journal.pone.0094215

    3305 joueurs (dont 29 de sexe féminin)

    Avec incertitude, l’intervalle de confiance pour la dégradation de performance est entre 20 et 30 ans.

    Le déclin (inexorable…) ne dépend pas du niveau du joueur.


    Figure 5. Impact of aging on PAC Latency, and respective intercepts by League as described by the best fitting piecewise linear model.

    (soit, en calculant, de 10 ms de latence en plus par année au delà de 24 ans pour les meilleurs (400 ms, au top) à 25 ms par an pour les moins bons (1000 ms, au top)

    Néanmoins, les « vieux » conservent leur niveau global. Il est donc émis l’hypothèse qu’ils compensent la dégradation de performance, sans doute en « diminuant leur charge cognitive » .

    Experience nevertheless allows one to compensate for these declines indirectly. In our study, older players appear to hold their own despite their declines, perhaps by decreasing their cognitive load through the use of simplified strategies or improved use of the game interface.

    Les données individuelles (par niveau/ligue)


    Figure 3. Scatter plots of age and looking-doing latency by league.

    • Ca me semble encore un bel exemple de « big data hubris », cf cet article :
      http://arstechnica.com/science/2014/03/researchers-warn-against-the-rise-of-big-data-hubris

      Et notamment cet avertissement :

      The problem they identify resulting from this form of hubris is that it’s relatively easy to use big data to identify eye-catching and publicity-generating correlations. It’s much harder to turn these correlations into something that’s scientifically actionable, and harder still to do the actual experiments that reach a scientifically valid conclusion. (As the authors put this, “The core challenge is that most big data that have received popular attention are not the output of instruments designed to produce valid and reliable data amenable for scientific analysis.”)

      Accessoirement, avec « 3305 joueurs (dont 29 de sexe féminin) », on ne tire pas de conclusions générales sur les humains mais sur une population (au demeurant plutôt restreinte) de joueurs de sexe masculin dont on ne sollicite qu’une seule forme d’intelligence.

      #bigdata #hubris