4 responses

  1. Gord

    Is Alex Scott still fit? Any chance on signing her temporarily?

  2. Andrew Crawshaw


    She wasn’t anywhere match fit at the end of last season, plus she’s far too busy now with all of her media work. I suspect we may well be looking for a midfielder in the January transfer window.

    I think we will be strong enough to get a result at home to Brighton this coming Sunday, even assuming that a couple of our youngsters play. I’m not sure that we have the ability to beat City on Sunday week without at least three of the top four fit to play the full 90 minutes

  3. Gord

    Well, let’s hope for the best this weekend, including no more injuries. Hopefully we’ll get some people back off the injured list soon.


    OT: Statistics (Men’s)

    I collected 23 “variables” out of the data up to the end of GameDay 12. One variable is time of possession, which is obviously strongly correlated to percent possession. I haven’t tried to find the amount of extra time in either half, so I am just assuming 90 minute games. With 23 variables, there are 506 possible relationships, except that comparing A to B is the same as comparing B to A. So the number of correlations examined is 253.

    A few of the correlations are either -1 (3) or +1 (2), and they are not investigated at all. There were 8 strongish negative correlations (at or below -0.70). There were 17 strongish positive correlations (at or above +0.70). The only correlation which seems to point to something real, is that the rate at which the home team commits fouls (home fouls divided by home possession) shows a strongish correlation with the amount of away possession. The other 24 strongish correlations are like comparing two measures of the same thing. The other 223 correlations are either not strong enough or are indistinguishable from no correlation (near zero is the most common situation).

    It is possible that the data constitutes multiple populations, and this would make analysis more difficult. One possible partitioning is:
    1. Top6 v Top6
    2. Top6 v ROTP
    3. ROTP v Top6
    4. ROTP v ROTP
    This latter group would be the largest subset, and so I looked at just ROTP v ROTP games. While many of the strong-ish correlations became stronger, none were other than trivial (in some way comparing a variable to itself).

    My guess, is that there just isn’t enough data yet, and we are almost 1/3 of the way through the season.

  4. Gord

    Der Spiegel has an interesting writeup on standard player contracts in the UK.


    I don’t think how the information is presented is well thought out.

    In any event, if a male player gets seriously hurt and is missing from the team for 18 months; the team at that point is allowed to terminate that player’s contract.

    The situation for females is different. They can be terminated after only 3 months.

    Apparently no female has been terminated for being injured, but among the examples in this Spiegel article are Arsenal players from Germany.

Back to top
mobile desktop