All Bets Are Off

This is often my motto when it comes to Derby matches, such as the Liverpool v Everton match this evening.

As you probably know, I like my stats and often spend time going over them before I decide to get involved in a match (depending on whether I'm trading or scalping).

I had already decided that I wouldn't get involved in the Liverpool match, but still had a look at the very basic stats to see what they looked like - and although this is undoubtedly awful after-timing, it does show that if you rely on just the basic data then you'll probably be heading for bankruptcy in the long term. So what did the basic data have to say?

Well it showed that Liverpool had drawn more games at home than any other team in the league, a whopping 61%. It also showed that their matches at Anfield had only averaged a paltry 1.92 goals.

And for the visitors, before this match, they had seemingly struck a little run of form, but their away matches had also only seen 2.15 goals scored on average.

If I was going to have a stab at a prediction on this game based only on these stats - putting to one side the fact that I already know the result - then I would have guessed at a low-scoring draw, possibly 0-0 or 1-1. But of course had I trusted these rather rudimentary methods, I would have lost money.

I should point out here, that I didn't bother to look at any of my advanced methods for this match, so there is the possibility that these too could also have pointed to the same kind of prediction. Generally, however, they would prove more profitable.

Many people avoid derby matches, so I realise that I'm not saying anything particularly new in this respect. I only mention this game because the raw data does (apparently) throw-up an opportunity. But looking at the data in this way provides absolutely no edge whatsoever. Every man and his dog is not only capable but probably also willing to analyse the data with such rushed brushstrokes, so we must go further if we're to make any serious headway.

This follows on from what I was saying in my I See Deep Patterns post last month, in which I suggested that we should try to look at the data in a different way. Whether betting/trading/scalping is a profession for you, a hobby or just a bit of a laugh, we all want to win money (or it's not a laugh, is it?). And in order for us to achieve this goal, we need to put a little bit of work in.

We shouldn't be afraid of this. Generating the data is not never-ending work. Often, once you've sorted the data into the format that you require, then it's just a case of keeping the raw data up-to-date. The football-data website can provide most of what you require in this respect, and what's missing can be gleaned from other sites. But once you have gathered it and formatted it, then analysed properly, it should prove it's worth over time.

And then, all bets will be back on again :-)