Wednesday, 15 August 2012

Catch 22

While SoccerDude stopped short of calling in the authorities, he has nonetheless told all his friends (both of them) and neighbours that I have crashed his domain, and behaved appallingly.

His full reply, worth reading, although obviously not as entertaining as this blog, can be found here, but a couple of comments in my defence before you head off over there. (Between starting this reply, and posting, I see that SoccerDude has been feeling bad for being "somewhat hard and mean-spirited towards Cassini". Not a problem for me at all. Signor Cassini has a think skin, and was mildly amused by it all. I was pondering, for example, how a bogey can have an Italian appearance, although it's snot something I want to dwell on really.)

It was actually a novel experience to be corrected about something, because my first comment is an admission that I did indeed go the wrong way when suggesting the sample size of 22 was cherry-picked. I could say that my sub-editor accidentally removed the word ‘not’ from the sentence, which should have read suggesting the 22 figure was somewhat 'not' cherry-picked, but since such a badly worded sentence would never appear on my blog, I will just have to admit that the cherry-picking suggestion was wrong. SoccerDude’s already weak assertion that Community Shield draws are rare would have only been weakened further by the choice of 22 matches.

As for going back to 1908, I think this is a very lazy approach. This match can trace its roots back to 1898 and the Sherriff of London Charity Shield, so I would suggest including that first game, which coincidentally was a draw between Corinthian and Sheffield United – which further strengthens my case that draws are far from rare in this match. As an historical side-note, not a lot of people know that the Sherriff of London Charity Shield was actually last competed for as recently as 1983, when it was won by Watford who beat Corinthian Casuals 6-1 (not quite a draw).

Having cleared up that little mess, I now turn my attention to SoccerDude’s mocking of my suggestion that “with probably thousands of people 'trading' this game, it seems to me that identifying in-play trading opportunities is at best difficult”. Now I should have added the word value before opportunities, but I think most people assumed this.

My point is that when a football match is in-play, there are so many traders in the market, that to think you can somehow spot value ahead of those others, is something of a reach. I am talking here about the more active of in-play markets – some of the less populated markets may well offer a reasonable shot at finding value, but if you are trading Match Odds or Under/Over 2.5 goals in a big game, I think this is a stretch. There is nothing wrong with having confidence, but a reality check is good too.

For me, the reality is that when we are watching a football match, we are all seeing the same thing. With some exceptions, I suggest that once a game is in-play, the market moves in a predictable direction at a predictable speed, interrupted only by rare, and key, random events. If an Under 2.5 price is 2.0 at kick-off, the rate at which that price will decay is, for the most part, known. If the game proves to be more defensive than expected, then it is true that when this becomes apparent, the Under price may briefly become value, but this change has become apparent to everyone watching the game, and unless you are at the front of the line, the value will have gone.

I have some more thoughts on this for another day, but essentially, my feeling is that the term 'football trader' is abused. , I believe that the traders who can profit long-term from football are few and far between. SoccerDude may well be one of them, but I would like to know more about this ‘qualified and defined edge’.

This is a bold claim, and as the late Carl Sagan put it – “Extraordinary claims require extraordinary evidence”. The odd large loss should be a red flag. Losses should be more frequent, but smaller in size, than wins.

Successful trading is about letting winners run, and cutting the losses. When the ‘odd large loss’ appears, you can certainly be forgiven for suspecting that this is a result of gambling, rather than trading.

I'll wrap up for now with some high praise from SoccerDude, who as I have mentioned before, has himself written some great posts on football ratings and markets:
Let's be clear, Cassini's blog is one of the best out there. It's well-written, full of decent content and his output is prolific. It's also one of the longer-lasting blogs discussing our favourite subject, with many others having long fallen by the wayside since he's been going. My own little corner of the blogging universe is also in its infancy compared to his, and so I should be a little more humble, I suppose.
By the way, for anyone who is genuinely interested in the analysis-side of betting, can I advise that you read the four-part series on Applying Elo Ratings to Football, written by Cassini on It's clearly written and should provide a decent uplift in your strike-rate if used well. In these articles, he gives a good grounding on how to use / rate form.
Where he says 'one of the best', it's clearly a mistake, and should read 'the best'.

And I was joking about going back to 1898 for the Community / Charity Shield data. 1908 is quite far enough.

1 comment:

George said...

Liked your Elo articles. Very skilful navigation of the topic!