Confirmation bias is very real, and basically a very simple concept: as humans, we prefer to stick with our current views rather than changing them.

This can lead us to look at new pieces of evidence or information in a less than objective manner: we often subconsciously place more stock in information that confirms our current thinking, as opposed to that which challenges it.

Where the true meaning or impact of new information is up for interpretation, by default we’ll look at how it can support our current thinking rather than debunk it.

It’s just human nature.  In very simple terms, we like to think we’re right!

A simple, non-punting example can be made of a football player.  Anybody who’s ever supported any football club in any code will have encountered a team’s “whipping boy”: the poor soul who’s the target of his own fans’ wrath for his supposed incompetence.

During the game, the player will make a mistake, as all players sometimes will.  The fans will take that single error and interpret it as evidence that they’re correct (a quick shout-out to Essendon’s Brent Stanton if you ever happen to read this!).

The twenty other good things the player did well don’t rate a mention.  Why?  They don’t conform to the person’s existing thinking and so don’t add weight to their argument.  They’re ignored, despite being in the overwhelming majority and perhaps a much better guide to the player’s actual performance.

The coach continues to select this maligned player, as he (hopefully!) takes a more objective view on the topic and thus sees his true worth to the team.  This might perplex or anger fans, who can’t see past their own bias.

Not that the coach himself won’t have other biases of his own.  They’re a strange bunch, football coaches, though I’m informed they are actually human.

So what?

When you consider that the vast majority of our punting is done based on our opinion (or that of others), the dangers of confirmation bias to the hip-pocket become very obvious.

For example, ever done the form and thought you’d found something based on a race you remember?  Horse X is running today in the wet, and you remember backing it last campaign in a wet race and having a good win.  But what’s one race?  We need more evidence than that!

So you have a further look through at it’s form… and you don’t get far, because it’s very last start was a win – also in the wet!  That’s all the evidence we need, time to load up on this nag.

If you keep looking through the form, you might see other wet track performances (perhaps even at today’s track, or with today’s jockey) which are very poor.  But there’s a good chance you won’t pay as much attention to them, as your thinking is now set and they don’t support it.  There’s your confirmation bias in action, and it may well be leading you to a poor value bet.

Can you stop it?

Confirmation bias is extremely difficult, if not impossible, to eliminate completely.  But there are a few methods you can employ when it comes to your own punting.

Firstly, pay some attention to computer-calculated data.  It’s everywhere now, and although not many betting approaches rely on it entirely, they’ll almost certainly take it into account in some direct or indirect manner.  Computers don’t have emotions and process data objectively: they can break down a larger amount of results into a simple number, rather than placing too much emphasis on any individual example.

Secondly, if you’re doing your own punting, try to develop a range of sources of information which you trust, and take them all into account (to a point).  This is more likely to open your thinking to possibly conflicting points of view, particularly if you trust both sources.

Successful punters are a diverse lot, but almost all of them are very objective types.  If you’re going to be successful too, it’s a trait you’re going to have to develop.

1 COMMENT

  1. Got to hand it it to you Mark, you don’t shy away from the tough ones. LOL

    Now I’ve got a term for that damned thing that has pestered me for years and caused more frustration than I care to mention. Confirmation Bias, an insidious enemy just waiting to mug all your good hard work and turn logical process into confusion or just plain bad judgment. Thanks for labeling the dreaded thing.

    Totally agree with the computer generated data as a proving process. As I was a software developer for many years it is hardly surprising I look to that area for real confirmation. I tend to play devils advocate on my final few selections and any others which seem oddly over-supported in early markets. I look for the data supporting each for both win or loss, check it’s validity and applicability to the current event and often I find surprises. I like neurals ( a pretty straight data analysis of past performance) as a general guide and recently I’ve been looking at the Champion Bets daily ratings as another computer related data source.

    I’ll look for confirmation of my own selections and also take note of those who are highly rated on a data basis and check to see if I’ve missed something. Why are they not in my market? This may be a mistake in my analysis or it may well be alerting me to my beloved friend, a false favourite!

    Still, it’s hard to stop the ‘on a roll’ influence when you think you’ve uncovered something juicy. I think personal judgment and experience have a major role to play in successful and enjoyable punting but a purely analytical process should be employed to compliment that experience and judgment. We are human and we simply don’t get it right all the time. Only liars (and Shane Dye, just ask him ) can achieve that.