Wednesday, 3 March 2010

AU - Get Off Of My Cloud

I like getting comments. It makes it easy to find something to post the next day. A couple more hit the blog in the last couple of days.

First, Mr Anonymous writes very constructively:

I think you should drop the AU from the dutching as in effect you backing a whole series of scorelines of which your predicted superiority is probably in the minority (say 4-2 and 5-3 in a 2 goal adv) Also as you discovered in the Chelsea game your also backing the diametric opposite of your prediction which presumably is poor value.
I think I’m in agreement for the most part. For sure I will be giving up a few winners, but the Any Unqutoed price is not looking like it’s value. Looking at Serie A, there have been 31 home teams expected to win by 2, and just 4 have hit the Any Unquoted. La Ligue has 6 of 42, Bundesliga 7 of 41 and the EPL 3 of 14. With the AU typically in the 4.0 to 6.0 range, it’s clearly poor value. However, in La Liga, the AUs hit 10 of 33, most due to Real Madrid and Barcelona who have some big wins at home. Anyway, with the Spanish exception, I think you are right that AU represents poor value.

Another comment was added yesterday to a post of January 17th, and read:
Try looking at the goalscoring abilities of the underdog. I'll try to calculate whether they are capable of scoring 2 goals (underdogs rarely are) in the game then go from there. Have a look at my blog for further discussion on correct score laying strategy:
I'm not sure I really understood what he was saying, and I rather lost confidence in any calculations when I read his latest post, edited here, but available on the blogroll in full:
A while ago there was a correct score laying thread on one of the betting forums by a man called Scotty who successfully layed some 350ish correct scores in a row.
Now a lot of people were in awe of Scotty after this but it really wasn't all that impressive and I'll explain why. Almost all his correct score lays were priced between 100 and 300. If we take the average as 200 then Scotty's expectation is to lose 1 in every 200 times and he actually lost 1 in 350. Good yes but given that it's less than twice his expectation it's hardly amazing. A good analogy would be a gambler getting two winners in a row at evens(2) which is actually more impressive than Scotty’s run of wins as it is running at double the expectation whereas Scotty was just less than double. This is not to belittle Scotty's thread in any way, just to say that it didn't quite deserve the awe it received.
For someone involved in probabilities, the poster actually demonstrates a worrying misunderstanding of how probability works. If Scotty’s winning run was 350, and he was laying at 200, then by my calculations, the probability of going 350 without hitting a loser is actually just 17%. He would be expected to have a loss after 139 bets, and while hitting a 5/1 shot isn’t really deserving of too much awe, it is a little more impressive than winning successive coin tosses at 3/1.

I also disagree with his assertion that “my point is that you should expect to win almost every bet” since expectations rather depend on the price you are betting at, but when he says that “the difference between winning and losing at correct score laying (and gambling in general) is the price you take” he’s correct – it’s another way of saying that you need to find value to be successful.

[Of course, my calculations above could be the ones that are wrong, since two nights ago I couldn’t correctly subtract 9 from 100…]

Finally, a rare prediction from the Conference National where the big game of the night sees Histon v Rushden and Diamonds predicted to end up as a draw at 3.55. Overall these hit at an average 24%, but 10 of the last 22 have been correct, with no less than four 2-2 results (which is of academic interest as there are no Correct Score markets at this level) but it does lead me neatly into letting you all know that Over 2.5 goals at the prices available (2.08+) is value as almost 60% of such games finish this way.


Anonymous said...

Is there any reason why you are not posting analysis of your selections?

I've grouped your selections into a spreadsheet and I can't see where you are making money? It would help clarify things if you put up your stakes as well as your selections.

Somebody posted further back that your ratings just seem to mimic the correct score market. I have to agree that they just seem to more or less match the underlying markets?

I just can't see how your ratings are making you anything?

CSL said...

Hi Cassini,
My comment regarding the underdog scoring 2 goals: 2-0 to the underdog is often considerably underpriced and in the 30's when it should be far higher.
A few recent examples of this:
Lay Fulham V Burnley 0-2 @ 36
Lay Tott V Everton 0-2 @ 30

Also I retracted my terrible analogy in my latest blog post and I thank you for pointing it out.

Anonymous said...

Admitting your error CLS means your now man enough to eat some Humble Pie!