Cheers, that's the nugget. I sense more reading on the way.
How many of you also gamble (deliberately)?
- ShaunWhite
- Posts: 9731
- Joined: Sat Sep 03, 2016 3:42 am
- ruthlessimon
- Posts: 2094
- Joined: Wed Mar 23, 2016 3:54 pm
Although the question is, where is the initial value (calculating the mean) taken from?
If something is backed off the boards late, perhaps it becomes a value lay bet; but not necessarily a value lay pre-off trade. I'll have to look at my stats again but I'm sure the large majority of "big movers" simply don't have time to reverse their full distance - although there's a high chance I'm defining the "start point of the big move" differently to how'd you measure it - hence why I think confusion of definitions/terminology is a bigger component of failure vs psychology.
& if the price doesn't move after laying it - surely that means the price was fair & not value (pre-off)?
- ruthlessimon
- Posts: 2094
- Joined: Wed Mar 23, 2016 3:54 pm
For May (so far) it's actually come out as a positive correlation
The more the fav steams, the smaller the reversal (correlation of 0.2) - & it can just about be seen in the graph:
.. but I've taken fair value as 5mins (which is a very very big assumption)
The more the fav steams, the smaller the reversal (correlation of 0.2) - & it can just about be seen in the graph:
.. but I've taken fair value as 5mins (which is a very very big assumption)
- ruthlessimon
- Posts: 2094
- Joined: Wed Mar 23, 2016 3:54 pm
I'll shut up after this one
but if we plot Peter's latest video on the graphs; here's where it'd be for May (even though the trade was Sept):
I think this confusion leads to so much unnecessary heartache - & such a large reason for people failing!!
but if we plot Peter's latest video on the graphs; here's where it'd be for May (even though the trade was Sept):
I think this confusion leads to so much unnecessary heartache - & such a large reason for people failing!!
I think the question is, how does one know that the two horses are of equal ability, by looking at the graphs, in the first place?
If a horse drifts markedly for no apparent reason one may speculate that it's got too big and it's now value, but it's nothing more than speculation. On the other hand if you have a strategy of backing big drifters in certain circumstances then you're just looking for value on average.
-
- Posts: 863
- Joined: Tue Dec 15, 2015 6:47 pm
One thing I do believe trading has taught me about gambling, is how inefficient the market can sometimes be about the fundamental value for some selections.
The huge drifters that then go on to win/run very well in the race. The hammered favourites or steamers that run awfully in the race. After trading for any length of time you soon realise the market can often be way off in it's statistical estimation of fundamental value.
The huge drifters that then go on to win/run very well in the race. The hammered favourites or steamers that run awfully in the race. After trading for any length of time you soon realise the market can often be way off in it's statistical estimation of fundamental value.
...Or perhaps a new BA variable; where we can test for the harmonic mean---- in the last x minutes/seconds selection's harmonic mean was x <,> or = than/to [insert back/lay/last traded price here]ShaunWhite wrote: ↑Mon May 21, 2018 3:49 pmCheers, that's the nugget. I sense more reading on the way.
- Crazyskier
- Posts: 1166
- Joined: Sat Feb 06, 2016 6:36 pm
Re: How many of you also gamble (deliberately)?
Me (sadly).
Lack of discipline and loss acceptance is the bane of my life (like so many others I'm sure!)...
CS
Me (sadly).
Lack of discipline and loss acceptance is the bane of my life (like so many others I'm sure!)...
CS
- ShaunWhite
- Posts: 9731
- Joined: Sat Sep 03, 2016 3:42 am
It's generally accepted that SP is efficient in a large sample. Have you got any stats that say the SP of the big movers isn't? The danger is that these are just memorable rather than statistically significant. The value probably existed before the move but if we knew that we'd all be minted.stueytrader wrote: ↑Thu May 24, 2018 11:10 amThe huge drifters that then go on to win/run very well in the race. The hammered favourites or steamers that run awfully in the race.
WRT the topic of the month 'harmonic mean', is this proven to be even closer to the actual winning likelihood than SP? If deviation represents value then I assume that's the case.
- ruthlessimon
- Posts: 2094
- Joined: Wed Mar 23, 2016 3:54 pm
I'll take up the challenge Shaun Gives me a good excuse to finish merging the BF promo data to my current capture sheets
I was gonna say "should be easy".. but even a simple statement such as "huge drifters that then go on to win/run very well in the race" quickly leads to a lot of complexity if we're not careful (morning WAP, Preoff WAP, SP, Big recoveries vs flat markets, race types, courses, speed of move etc)!!
I was gonna say "should be easy".. but even a simple statement such as "huge drifters that then go on to win/run very well in the race" quickly leads to a lot of complexity if we're not careful (morning WAP, Preoff WAP, SP, Big recoveries vs flat markets, race types, courses, speed of move etc)!!
Samples are often too broad and generic.
Is the earth round, or flat?
If you are looking from space, it is clearly round, but if you are on the American plains, it's clearly flat. It's just a matter of perspective. If you plot data from a general data set and apply general rules to it, it will produce general results. Finding an edge there is incredibly hard.
On average the each is perfectly round and flat, you are looking for some Himalayas or Rockies and selling flat by waiting for something to walk over them and eventually return to mean. You may also want to buy flat when they are walking through somewhere like death valley which is 86m below sea level.
I've run a value lay bot since 2011 but realised it results were erratic in terms of going on winning or losing runs. So I know also run a dutching bot which squashes individual variance by covering more than one selection.
On a final point, anybody who pays premium charge should also be gambling.
Is the earth round, or flat?
If you are looking from space, it is clearly round, but if you are on the American plains, it's clearly flat. It's just a matter of perspective. If you plot data from a general data set and apply general rules to it, it will produce general results. Finding an edge there is incredibly hard.
On average the each is perfectly round and flat, you are looking for some Himalayas or Rockies and selling flat by waiting for something to walk over them and eventually return to mean. You may also want to buy flat when they are walking through somewhere like death valley which is 86m below sea level.
I've run a value lay bot since 2011 but realised it results were erratic in terms of going on winning or losing runs. So I know also run a dutching bot which squashes individual variance by covering more than one selection.
On a final point, anybody who pays premium charge should also be gambling.
Do you mean traders should become gamblers to avoid PC or gamblers have a high chance being taxed with a PC? If last than it's a very controversial statement.Euler wrote: ↑Fri May 25, 2018 8:35 amSamples are often too broad and generic.
Is the earth round, or flat?
If you are looking from space, it is clearly round, but if you are on the American plains, it's clearly flat. It's just a matter of perspective. If you plot data from a general data set and apply general rules to it, it will produce general results. Finding an edge there is incredibly hard.
On average the each is perfectly round and flat, you are looking for some Himalayas or Rockies and selling flat by waiting for something to walk over them and eventually return to mean. You may also want to buy flat when they are walking through somewhere like death valley which is 86m below sea level.
I've run a value lay bot since 2011 but realised it results were erratic in terms of going on winning or losing runs. So I know also run a dutching bot which squashes individual variance by covering more than one selection.
On a final point, anybody who pays premium charge should also be gambling.