Guys,
This is going to be a bit of a long post, but, all week I've been banging my head agains a brick wall trying to figure out a price relationship between the horses.
We all know that if one is drifting,another should be coming in but which ones do we concentrate on, 1&2?, 1&3?, 2&3? etc,etc.
I then found the "Correl" function on excel which I think, is the answer to our prayers.
On the attached sheet you'll see 3 horses and the price correlation between horses 1&2, 1&3 and 2&3.
The price relationship between 1&2 is -1, between 1&3 it's
-.50 and between 2&3 it's .50. To cut a long story short, the nearer to -1 it is the better. So we can immediately see that we should concentrate on the first 2 horses. (The worst relationship is between 2&3, at .5 they can be put out with the washing).
So, concentrating on 1&2, their prices are exactly in tandem, one is coming in by 1 point each time as the other is going out by 1 each time, hence the result of -1 ( a perfect score!).
We can see that the average price for 1 is 1.73 and for 2 it's 2.02.
In Col J I've put a figure of 1.9 and we can see that it's 17 points overvalued from it's average.
So how do we know for definite that this is the one that's over priced and it's not the other that's underpriced?
Ah, don't forget the correlation, which is telling us that 1&2 has the strongest price relationship so, looking at number 2 the variance is only -0.02, it MUST be number 1 that's overpriced by a large amount.
If you put a figure of 1.75 in col J the variance is only 0.02, so, looking at both variances, the prices could go either way, maybe this is not the time to place an order.
Ideally, all the other horse in the race should be examined the same way, but it gives us food for thought.
Please note: The prices/book percentage were made up for the purposes of this exercise!
Thanks. Nigel
Price Relationships-Over/Undersold.
-
Timstertoo
- Posts: 9
- Joined: Wed Apr 15, 2009 7:56 pm
I've been thinking about the greyhounds a lot in this way and will be trying the same on them.
With the horses you sometime have enormous outsiders that give you crazy amounts if you add up all the odds. On the dogs you get a number mostly close to 36 if you add up the individual odds which makes sense to me. The number goes up a bit with a real big outsider but to me it feels that the relation between the odds on the dogs are a bit stronger due to the field always being exaclty 6 and maybe also that every dog has a real, albeit sometimes small, chance of winning.
Keep up the great work! I'll certainly let you know if I stumble on to something, or make a sheet that might help you.
With the horses you sometime have enormous outsiders that give you crazy amounts if you add up all the odds. On the dogs you get a number mostly close to 36 if you add up the individual odds which makes sense to me. The number goes up a bit with a real big outsider but to me it feels that the relation between the odds on the dogs are a bit stronger due to the field always being exaclty 6 and maybe also that every dog has a real, albeit sometimes small, chance of winning.
Keep up the great work! I'll certainly let you know if I stumble on to something, or make a sheet that might help you.
-
Timstertoo
- Posts: 9
- Joined: Wed Apr 15, 2009 7:56 pm
I might also be completely barking up the wrong tree but I'm still very new at this so even being very wrong is a valuable lesson. 
-
Nero Tulip
- Posts: 686
- Joined: Wed Apr 15, 2009 5:29 pm
Food for thought I think is the right way to look at this. Very often I start down one road, only to find many smaller ones branching off with lots of different sign posts. This I think is fairly similar, it's not a solution, but something to consider.
Book % is the obvious starter for this, the greater the share, the more influence that outcome will have over the rest. Using this as a starting point for over/undersold estimations is perhaps much harder than the maths might seem. As with anything in these markets, who are we to say anything is under/over, price is such a matter of opinion, and opinions can be way out, even if the majority share it.
I prefer to rephrase this line of enquiry as something that might indicate the likelyhood of the direction a price could move in. And as such, would allow us to manage a diversified book a little better, or gives us yet another indicator to add to an already burgeoning collection. Part of the bigger picture. Nice work again Nigel.
Book % is the obvious starter for this, the greater the share, the more influence that outcome will have over the rest. Using this as a starting point for over/undersold estimations is perhaps much harder than the maths might seem. As with anything in these markets, who are we to say anything is under/over, price is such a matter of opinion, and opinions can be way out, even if the majority share it.
I prefer to rephrase this line of enquiry as something that might indicate the likelyhood of the direction a price could move in. And as such, would allow us to manage a diversified book a little better, or gives us yet another indicator to add to an already burgeoning collection. Part of the bigger picture. Nice work again Nigel.