• lacanianmrxistB
    link
    fedilink
    English
    arrow-up
    1
    ·
    1 year ago

    There are statistical methods for determining the reliability of correlations as statistical predictors, as opposed to mere “coincidences”. It’s difficult to know without knowing sample size, receiver ability, tipped balls, “hurries”, etc. Really hard to gauge the usefulness of a graph like this without knowing whether ALL variables were taken into account (receiver sucks v. wide open in the end zone; ball tipped after thrown; scramble v. Pocket; etc.). When things like this are not explained in statistical terms, hard to say if they’re garbage or wishful thinking. I think getting a good first half would be an improvement. If we win, convincingly, it stop much of the critique of “the Howellitzer” (great tag, whoever thought of it). Still taking the boys straight up.