Between the Lines

"Of all unimportant subjects, football is the most important" – Pope John Paul II

The enigma of the negative persistence effect

Posted by hakanrylander on September 29, 2008

For some time now I’ve suspected that a number of hardcore visitors to this site share a soft spot for a scientific approach to the good things in life; football, sports betting, medical products, you name it. Those of you who fall into this category, please help me out. While reading a statistical analysis of English football I came across the following statement:

“Analysis of sequences of match results reveals evidence of a negative ‘persistence’ effect: a recent run of good results appears to create either pressure or complacency, increasing the risk that the next result will be bad.”

Does this really make sense? It seems to me that a run of good results would in fact increase the chance of the next result being good. I wouldn’t e.g. bet against Chelsea at least drawing their next home game.


6 Responses to “The enigma of the negative persistence effect”

  1. Oh, that is an interesting topic! Still, I believe that all attempts to scientificly prove that “a sequence of good results will increase the chance of a failure” are doomed. Yes, common sense tells us that the chance of a perfect EPL season is slim and that is just as much science we need.

    I have seen a few attempts to prove the “due to fail theory” and most of them fail to convince a grumpy old geezer since they tend to be based on assumptions that would implicate the theory itself. A way of reasoning good enough for Descartes but I beg to differ.

    Suppose that Chelsea of today would play in the conference (Oh, joy!). It is very likely that they will win a lot of games. Probably all of them! Anyone who claims that Chelsea would have to loose some games in the conference due to the “all good things come to an end theory” are invited to call me. I have some land for him to buy!

    “The Trend is Your Friend” is the mantra of the stock market but I think that mantra has more substance than the “due to fail theory”. The Key to Sports Betting is to succesfully evaluate when the betting market has over reacted to the friendly trend. There might reason to bet againts, say, Chelsea because the betting market has a upward bias in the estimate of Chelsea’s winning chances and not because they have won so-and-so many games in a row (which of course would feed the bias).

  2. hakanrylander said

    Thanks, Farbror! You have just succeded in raising the intellectual level of this blog by around 400 percent (possibly a conservative estimate). The way I understand it, the theory that a long winning run moves a team closer and closer to a defeat can only be valid if we assume that God (the one in Heaven, not Cantona) has already decided when e.g. Chelsea shall lose their next home game. If he has, I hope it’s soon.

  3. BillyIdle said

    This idea must be kin to the American theory (prevelent in predicting baseball winners) of the “due team”. Meaning due to win because they have lost too many games, or too many games to a rival team, or to a rival starting pitcher. This seems to be an inverse “due to win” idea here. The “revenge” idea that one team will try harder. Doubt it has much real basis in fact. The less talented team may lose no matter what the effort.
    However as in other sports, teams can become complacent and fall back on the idea of floating in with the tide. The we are out in front, we don’t need to try hard syndrome. If there is any tie in to the statistics you refer to it would amount to psychological factors concerning lack of effort. Also known as human nature. So we are speaking of the emotional factors of wanting to prove something versus coasting toward a championship. You be the judge.

  4. BillyIdle said

    “Analysis of sequences of match results reveals evidence of a negative ‘persistence’ effect: a recent run of good results appears to create either pressure or complacency, increasing the risk that the next result will be bad.”

    Giving this a little more thought I can’t buy into the “pressure” idea either. In professional sport their is always pressure. In fact, the phrase “either pressure or complacency” sounds to me as though the person is guessing. What evidence? Which is the result, pressure or complacency?

    “Does this really make sense? It seems to me that a run of good results would in fact increase the chance of the next result being good”.

    Have to agree with hakanrylander. If there is a “persistent effect” it would have to be a persistent effort to try to win games not lose them. DUE TO LOSE would make a good Country and Western song though.

  5. hakanrylander said

    The quote that started this discussion “Analysis of sequences…etc” was taken from “The Economics of Football” by Dobson and Goddard (Cambridge University Press). I’ve now tried to look a bit closer at how they come to this conclusion.
    It turns out that they do in fact agree with me that a good run of results increases the chance of the next result being good. But they argue that this to a large extent is explained by team heterogenity (in English this means that good teams tend to beat bad teams), rather than the length of the spell itself and its effect on confidence/complaceny affecting the probabilities of the outcome of the next game.
    To seperate these two factors they use a simulation approach to establish how many and how long sequences (wins, losses, unbeaten or win-less) that should be observed in a season “if the only systematic determinant of match results is team heterogenity”.
    They find that the actual numbers of good/bad spells of any duration are almost always smaller than the numbers predicted by the simulation. Their conclusion is that there are negative short-term persistence effects (such as pressure or complacency) that tend to curtail the durations of both good and poor spells. This means that a team enjoying a run of good results has a higher probability of a poor result in the next game, relative to the team’s underlying strength.
    I’m still unconvinced. Mainly because I think it well nigh impossible to accurately simulate the number of sequences during a season.
    I also believe that the hidden parallels between sports betting and country & western songs need to be further explored.

  6. […] In previous posts I’ve touched upon two of the segments that I find most interesting; the increasing inequality between rich and poor clubs and how the length of a spell of wins/losses affects the probabilities of the outcome of the next game. […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: