Loading ...
Sorry, an error occurred while loading the content.

Re: [APBR_analysis] Re: Player Movement

Expand Messages
  • igor eduardo küpfer
    ... From: Kevin Pelton To: APBR_analysis@yahoogroups.com Sent: Sunday, August 31, 2003 11:11 PM Subject: [APBR_analysis] Re: Player Movement ... Link:
    Message 1 of 14 , Sep 1, 2003
    View Source
    • 0 Attachment
       
      ----- Original Message -----
      Sent: Sunday, August 31, 2003 11:11 PM
      Subject: [APBR_analysis] Re: Player Movement

      --- In APBR_analysis@yahoogroups.com, "Dean Oliver" <deano@r...>
      wrote:
      > So, though I didn't see JohnH's column, it would surprise me that
      > anyone is suggesting that competitive balance is good now.

      Link:
      http://sportsillustrated.cnn.com/2003/basketball/nba/08/18/hollinger.
      statitudes/index.html

      By JohnH's measure (# of teams with 60 or more wins or losses), the
      last two years have been two of the seven most competitive in league
      history.
       
       
       
      It just occurred to me that the topic of "competitive balance" is more subjective than I originally thought. For me, I think of a balanced league as one in which the outcome of any given game is close to 50-50, where picking winners unpredictable. I've always thought about it that way: if the result is a foregone conclusion, where's the competitiveness? My focus has always been on the game in front of me, at least as far as this topic goes.
       
      Other people like to see the absence of truly dominant teams (or truly putrid) during the course of the season as evidence for balance. JohnH seems to favour this view in his column. There is nothing inherently mistaken with this view. From this perspective it makes perfect sense to look at measures of the variability of winning percentages like standard deviations.
       
      Still other people focus not on individual games, nor individual seasons, but eras: frequent "dynasties" to them indicate an imbalanced league, as do horrendous sad sack franchises like the Clippers. A dominant team is one that manages to win across many seasons, not just one or two. Someone who favours this view of competitiveness would deny the standard deviation argument as irrelevant; they would claim that if the Celtics or the Lakers win every single season, it doesn't matter the variation in the league winning percentages. It's hard to deny the validity of this argument.
       
      What I like about the Competitive Balance Ratio is that it combines the two elements from the paragraphs above: it is the ratio of variation of team wins during one season to the variation across seasons. A pretty neat idea, but still needs a wee bit of tweaking IMO. For example, its reliance on winning percentages is too abstract for a concept as ephemeral as competitive balance, I think -- substituting some sort of ranks might give results closer to the everyperson view. After all, does it really matter whether the Bulls had 30 wins or 38 wins? In neither case are they in any danger of making the playoffs, and I think most fans see the primary difference between the Bulls and the Bucks as the difference between a crappy team and a postseason team, and not between a 30-win team and a 42-win team.
       
      Which brings up another thing: presumably the economists studying balance are doing so from their perspective as economists, and not necessarily as fans. Their definitions of "balance" may not be the same as the average fan's definition -- or yours and mine, for that matter. Maybe DeanO can add some thoughts here, but I'm guessing that econs study the topic by looking at its economic effects, such as attendance. This, however, is not my primary interest -- I am interested in the health of the league as I define it, not as it applies to those studying its economic health. For this reason I am interested in  finding a measure for competitive balance that assumes my definition. CBR looks like it does more to capture what I think of as balance.
       
       
       

      He also suggests in there that imbalance is good for the league - a
      question I have to imagine has been answered in some form by some
      economist at some point, but doesn't seem that difficult to do if it
      hasn't.
       
       
       
      An argument  could be made that the health of the league is increased by having the same recognizable stars achieving success on a regular basis. I can see some validity to this argument: the continuity of familiar names would be appealing to very casual fans, I think.
       
       
      ed
    • Jim Armstrong
      ... But the question is: who does this type of balance benefit? I might argue that if every game was so evenly matched that the outcome could just as well be
      Message 2 of 14 , Sep 4, 2003
      View Source
      • 0 Attachment
        On Mon, Sep 01, 2003 at 04:52:33AM -0400, igor eduardo küpfer wrote:
        > It just occurred to me that the topic of "competitive balance" is more subjective than I originally thought. For me, I think of a balanced league as one in which the outcome of any given game is close to 50-50, where picking winners unpredictable. I've always thought about it that way: if the result is a foregone conclusion, where's the competitiveness? My focus has always been on the game in front of me, at least as far as this topic goes.


        But the question is: who does this type of balance benefit? I might argue
        that if every game was so evenly matched that the outcome could just as well
        be determined by a coin flip, the interest of the average fan wouldn't be
        held very long. As the outcomes of games become more random, the win/loss
        standings become less relevant. Perhaps this type of balance would only
        appeal to the purist who enjoys the game itself, but has little interest in
        the outcome.


        > Other people like to see the absence of truly dominant teams (or truly putrid) during the course of the season as evidence for balance. JohnH seems to favour this view in his column. There is nothing inherently mistaken with this view. From this perspective it makes perfect sense to look at measures of the variability of winning percentages like standard deviations.
        >
        > Still other people focus not on individual games, nor individual seasons, but eras: frequent "dynasties" to them indicate an imbalanced league, as do horrendous sad sack franchises like the Clippers. A dominant team is one that manages to win across many seasons, not just one or two. Someone who favours this view of competitiveness would deny the standard deviation argument as irrelevant; they would claim that if the Celtics or the Lakers win every single season, it doesn't matter the variation in the league winning percentages. It's hard to deny the validity of this argument.


        Another common argument heard especially in baseball circles is the "hope and
        faith" type of competitive balance. Here fans of specific teams want to have
        some assurances at the beginning of the season that their team has reasonable
        hope and faith of qualifying for the post-season. To these fans, individual
        games don't seem to matter as much, nor do dominant teams. Even falling out
        of playoff contention isn't so bad as long as they have reason to believe
        their team is likely to contend within the next few years. I've suggested
        elsewhere that this might be measured by looking at standard deviations of
        preseason gambling odds.

        Other fashionable methods in recent years bring payrolls into the equation.
        High correlations between payroll and wins are deemed evidence of an
        unbalanced league, but often little effort is put into determining which way
        the causality runs or researching the complex factors that determine payroll.


        > Which brings up another thing: presumably the economists studying balance are doing so from their perspective as economists, and not necessarily as fans. Their definitions of "balance" may not be the same as the average fan's definition -- or yours and mine, for that matter. Maybe DeanO can add some thoughts here, but I'm guessing that econs study the topic by looking at its economic effects, such as attendance. This, however, is not my primary interest -- I am interested in the health of the league as I define it, not as it applies to those studying its economic health. For this reason I am interested in finding a measure for competitive balance that assumes my definition. CBR looks like it does more to capture what I think of as balance.


        I believe this is a fundamental problem of most measures of competitive
        balance. You start with someone's subjective view and work towards a
        formula that measures it. Too often the subjective view is rooted in
        the emotions of a sports fan who is unhappy with how his team is doing,
        or believes the league as a whole isn't as fun as in the "good old days".
        You come up with a measure that confirms that things are gloom and doom.
        Then someone else with an opposing subjective viewpoint will find a different
        measure that shows that the league is in fact balanced and that all is well
        with the world. This seems to happen in cycles over periods of years. At
        some point in the future the pessimistic fan's measure will show great
        balance, but if the fan is still dissatisfied with how his team or the
        league is doing, he'll simply come up with another measure to confirm his
        displeasure. And the cycle repeats.

        I believe looking at competitive balance from an economist's point of view
        is the more objective, and useful, approach. One can state your objectives
        up front, for example, to maximize overall league revenue or franchise values
        or to maximize fan interest as measured by attendance or TV ratings. Fans
        may not think this is important to them, but from a consumer's point of
        view it is, because fans don't want to see a league or team fold or the
        games not be televised. As with other products, in an efficient market, a
        financially healthy producer is good for consumers of the product, and it
        allows the producer the resources to give more of what the consumers want.
        Of course, you do have to be a bit careful with sports leagues that have
        monopoly power, which can be abused to the detriment of the consumers.


        > He also suggests in there that imbalance is good for the league - a
        > question I have to imagine has been answered in some form by some
        > economist at some point, but doesn't seem that difficult to do if it
        > hasn't.
        >
        >
        >
        > An argument could be made that the health of the league is increased by having the same recognizable stars achieving success on a regular basis. I can see some validity to this argument: the continuity of familiar names would be appealing to very casual fans, I think.


        IMHO, the real key is to determine optimal levels of competitive (im)balance
        that would generate the most interest among fans worldwide, and ultimately
        generate the most revenue for the league. Then of course, figuring out
        the factors that lead to such levels is a whole different story.

        A brief but informative economic discussion of this topic is included in
        a paper published online by John Sigfried of Vanderbilt University.
        http://www.cato.org/pubs/journal/cj14n3-4.html
        The section titled "The Welfare Effects of Balanced Competition"
        (about halfway down the page) is most relevant.

        Jim
      • Michael Tamada
        ... From: Jim Armstrong [mailto:jarmstrg@world.std.com] Sent: Thursday, September 04, 2003 8:57 PM ... I have not looked deeply into the literature, but at the
        Message 3 of 14 , Sep 4, 2003
        View Source
        • 0 Attachment
          -----Original Message-----
          From: Jim Armstrong [mailto:jarmstrg@...]
          Sent: Thursday, September 04, 2003 8:57 PM


          >Other fashionable methods in recent years bring payrolls into the equation.
          >High correlations between payroll and wins are deemed evidence of an
          >unbalanced league, but often little effort is put into determining which way
          >the causality runs or researching the complex factors that determine payroll.

          I have not looked deeply into the literature, but at the Western Economics
          Association meetings this summer, most of the economists who looked at the role
          of payroll and winning used simultaneous equations techniques, which permit
          measuring not just the impact of payroll on winning, but also the impact of
          winning on revenue (and then payroll). The economists who presented at the
          Society for American Baseball Research meetings also did so, when appropriate.

          What was interesting to me was that, in the small number of presentations that
          I saw, the simultaneity effects were modest. The payroll ==> winning effects
          looked much larger than the winning ==> payroll effects, so much so that
          ignoring the simultaneity didn't seem to affect the results much.

          That was only a small handful of studies, I don't know what most studies have
          shown.

          [...]

          >IMHO, the real key is to determine optimal levels of competitive (im)balance
          >that would generate the most interest among fans worldwide, and ultimately
          >generate the most revenue for the league. Then of course, figuring out
          >the factors that lead to such levels is a whole different story.

          Based on what I saw at the meetings, economists have only started to try
          to come up with such measures. "What do fans want?" is another way of
          stating the question, and though economists have for quite awhile been
          trying to estimate how much fans respond to winning, it appears that the
          research on how much fans respond to "competitiveness" is in its infancy.

          >A brief but informative economic discussion of this topic is included in
          >a paper published online by John Sigfried of Vanderbilt University.
          >http://www.cato.org/pubs/journal/cj14n3-4.html
          >The section titled "The Welfare Effects of Balanced Competition"
          >(about halfway down the page) is most relevant.

          A nice paper, utilizing pretty current economic research but
          presenting the results without a single equation or graph. Which
          is rare for economists, though more common for the ones who tend
          to write stuff for the Cato Institute. A teensy bit of warning
          there: the Cato Institute has an ideological axe to grind, one
          in favor of markets and against government intervention. In
          particular, whereas most average citizens and for that matter most
          economists will have some concerns about the possible negative effects
          of monopolies, cartels, and monopsonies, the Cato Institute will tend
          to pooh-pooh such concerns.

          So I think it's not a coincidence that this article acknowledges the
          monopsonistic power that sports leagues have, but suggests that
          this leads to an overall benefit to social welfare.

          On the other hand, professional sports are one of the areas where
          quite a few average citizens and economists will have that same
          viewpoint -- that monopsony, a draft, etc. are not bad things in
          pro sports. Also John Siegfried's been the Secretary for the
          American Economic Association for years, so he's not exactly some
          wild guy off the fringe (although even some of the AEA's presidents
          have had some fairly extreme views, from Galbraith on the left to
          the Chicago School guys on the right).


          --MKT
        • Jim Armstrong
          ... Some other studies I ve read show similar results, though perhaps not as clear cut as suggested above. Also, the winning == payroll effects are more
          Message 4 of 14 , Sep 5, 2003
          View Source
          • 0 Attachment
            On Thu, Sep 04, 2003 at 10:06:25PM -0700, Michael Tamada wrote:
            > I have not looked deeply into the literature, but at the Western Economics
            > Association meetings this summer, most of the economists who looked at the role
            > of payroll and winning used simultaneous equations techniques, which permit
            > measuring not just the impact of payroll on winning, but also the impact of
            > winning on revenue (and then payroll). The economists who presented at the
            > Society for American Baseball Research meetings also did so, when appropriate.
            >
            > What was interesting to me was that, in the small number of presentations that
            > I saw, the simultaneity effects were modest. The payroll ==> winning effects
            > looked much larger than the winning ==> payroll effects, so much so that
            > ignoring the simultaneity didn't seem to affect the results much.
            >
            > That was only a small handful of studies, I don't know what most studies have
            > shown.

            Some other studies I've read show similar results, though perhaps not as
            clear cut as suggested above. Also, the winning ==> payroll effects are
            more indirect and more difficult to measure. I also haven't yet seen many
            good arguments discussing which causalities are desirable for a sports
            league.

            I haven't heard much about the recent WEA/SABR meetings. Is there a summary
            of proceedings and/or listing of research presented available online anywhere?

            > >IMHO, the real key is to determine optimal levels of competitive (im)balance
            > >that would generate the most interest among fans worldwide, and ultimately
            > >generate the most revenue for the league. Then of course, figuring out
            > >the factors that lead to such levels is a whole different story.
            >
            > Based on what I saw at the meetings, economists have only started to try
            > to come up with such measures. "What do fans want?" is another way of
            > stating the question, and though economists have for quite awhile been
            > trying to estimate how much fans respond to winning, it appears that the
            > research on how much fans respond to "competitiveness" is in its infancy.

            Yeah, what fans as a group say they want don't always show up in their
            actions. They may say they want lower ticket prices, but they continue
            to be willing to pay for higher prices. The Clippers vs. the Grizzlies
            might be a very evenly matched game of professional teams, but such a
            contest isn't in relative high demand.

            > to write stuff for the Cato Institute. A teensy bit of warning
            > there: the Cato Institute has an ideological axe to grind, one
            > in favor of markets and against government intervention. In
            > particular, whereas most average citizens and for that matter most
            > economists will have some concerns about the possible negative effects
            > of monopolies, cartels, and monopsonies, the Cato Institute will tend
            > to pooh-pooh such concerns.
            >
            > So I think it's not a coincidence that this article acknowledges the
            > monopsonistic power that sports leagues have, but suggests that
            > this leads to an overall benefit to social welfare.
            >
            > On the other hand, professional sports are one of the areas where
            > quite a few average citizens and economists will have that same
            > viewpoint -- that monopsony, a draft, etc. are not bad things in
            > pro sports. Also John Siegfried's been the Secretary for the
            > American Economic Association for years, so he's not exactly some
            > wild guy off the fringe (although even some of the AEA's presidents
            > have had some fairly extreme views, from Galbraith on the left to
            > the Chicago School guys on the right).

            Agreed. I'm not an economist, but what I've learned is that economists
            tend to have reputations, fairly or not, based on the institutions they
            represent. Perhaps part of this is that individuals with similar
            ideologies tend to gravitate together. To some extent this is true of any
            academic field, but seems to be particularly true for the field of economics.

            Jim
          Your message has been successfully submitted and would be delivered to recipients shortly.