3am and can't sleep...

[pjcppts]

... so I've decided to write a few random musings...

So, been reading through the forums lately and interested in a few ideas, and also the psychology of trading. Seems to me, all the traders out there are likely to be in the camp of "it's never wrong to take a profit", which was Rothschild? I can see the logic. So...

My automation tracks certain traps in greyhound races, and lays 1 dog max per race provided certain criteria are met, which are basically the odds are shortening (check price v price x seconds ago) and are within a certain range (mid price, not too short and not outsider, say average odds are 4.0). The idea is to green out for 0.20p if possible using a lay to lose £6 per race. At 15 secs before race start green (or red) out. Theory is the market is very volatile on greyhounds and the chance of greening out is high.

It worked technically OK, but I had a fundamental psychological issue taking small losses when the odds moved against me, and would sometimes put in a minimum green out of 0.01p and just leave open red positions, which sometimes would cost me the full £6.00 and wipe out hours of good trading work.

So, a rethink. If my theory works, then my edge in being able to green out will be that I am laying at less than BSP. If BSP is taken as a close approximation to theoretical real odds of winning, then why bother with greening? Why not just leave an open position on each race. If you say my lays are at 4.0 I need 75% success to break even, but if the SP/true odds are 4.3 say, then I should get more than 75%.
Take 100 races laying £2 at 4.0 then 75% breaks even (ignore commission). 76% would mean +(76 x 2) - (24 x 6) = 152-144 = £8. Each % above 75 means £8 profit. That's a lot of 0.20p, even ignoring any reds...

So that's what I'm trying. I doubt you could call it trading anymore, but the results are certainly interesting.

Since Jun 19th (12 days):
595 winning lays out of 742 (80.19%)
Total profit £192 using £6 liability. I have my doubts in will scale up much beyond £20 per race but I want more results before ramping up. I'm perfectly happy with £15 quid a day average for now I've always been a very small punter.
Most losing bets in a row: 4 which was psychologically challenging
Most winning bets in a row: 23 (Monday morning!)
Winning days 9 (£40.95, £25.10, £5.20, £13.46, £2.61, £13.85, £43.34, £33.94, £38.75)
Losing days 3 (£10.82, £12.49, £1.44)

Now, may just be a random good run, time will tell, but "greening up is for pussies!"...

One last thing... every time I watch a dog race I'm now cheering when my lay gets knocked base over apex at the first bend. Does this make me a bad person?

Thanks for listening... oh, you've gone...

Pete

[Anbell]

You're thinking about the correct things

[Crazyskier]

The high volume, small margin stuff on the dogs is what I've been working on for years.

All I will say is that you need many tens of thousands of race results to determine anything approaching actual margins versus random variance. So keep at it. I personally always like to see the biggest loss days being rather smaller amounts than the average profit days figure - this for me is strong encouragement. I'd recommend using only trading profits to increase stakes, reducing back down if you see the results suffer. And always document everything, exporting to sheets so you can collate ALL the data and note every change to the strategy.

Good luck and keep us posted!

CS

[getagrip68]

An interesting read and good luck ... although on the latter point, yes.

[Derek27]

Back in the nineties on a night when I couldn't sleep and before I was on the internet, I ended up writing a program to track the real-time position of a fictitious satellite, with a basic graphical diagram of its orbit. It took me two weeks to complete but it certainly cured insomnia.

"greening up is for pussies!"...

I prefer hedges are for gardens.

[ShaunWhite]

which sometimes would cost me the full £6.00 and wipe out hours of good trading work.

So, a rethink. If my theory works, then my edge in being able to green out will be that I am laying at less than BSP. If BSP is taken as a close approximation to theoretical real odds of winning, then why bother with greening?

You're right to be thinking about the value of your bets, and that the BSP hedge bet is doing nothing positive other than reducing your commision a bit. You mention a loss aversion but that's probably down to lack of confidence about your general direction of travel. The losses aren't wiping out hours of work they are just part of the work, those will appear to be less significant if you look at your PL on a rolling 7 day basis or even a rolling month to stop you being bothered about the daily ups and downs.

I do OK on the dogs (aus more so than UK tbh), and 'trade' about 250 dogs a day (with about 1000 bets) , but still have losing days or runs of 3 or 4 days due to the ups and downs associated with the occasional missed hedge, They don't matter one iota, because the larger plus and minus amounts eventually even out over the months, they're 0ev (or close to) bets so that will be the case.

On a £6 stake you might well be up or down 60 a day so just get used to that rather than trying to micromanage the pennies. Many people don't even look at the cash pl because it's misleading (I certainly don't) all I look at is my theoretical pl vs bsp, the idealised world where the hedge is done and clean across the card. If that's positive then longterm you WIlL be positive longterm. Looking at the cash is unnecessary stress when it's down and unreasonably optimistic if its up.

[ShaunWhite]

Many people don't even look at the cash pl because it's misleading (I certainly don't) all I look at is my theoretical pl vs bsp.....They don't matter one iota, because the larger plus and minus amounts eventually even out over the months, they're 0ev (or close to) bets so that will be the case.

And as if by magic an example occurs, 2 horses just either side of evens that didn't do the 'right' thing when the hedge failed. Couldn't care less, one had a theoretical green value of £25, the other about £32. Couple of months ago in a very similar situation both won, made about £1200 and that didn't please me anymore than this disappointed me.

You can't make money if you're going to worry about cash from week to week or month to month, do it right and do it enough and 'value' never lies.

Screenshot_3.jpg

[Crazyskier]

Many people don't even look at the cash pl because it's misleading (I certainly don't) all I look at is my theoretical pl vs bsp.....They don't matter one iota, because the larger plus and minus amounts eventually even out over the months, they're 0ev (or close to) bets so that will be the case.

And as if by magic an example occurs, 2 horses just either side of evens that didn't do the 'right' thing when the hedge failed. Couldn't care less, one had a theoretical green value of £25, the other about £32. Couple of months ago in a very similar situation both won, made about £1200 and that didn't please me anymore than this disappointed me.

You can't make money if you're going to worry about cash from week to week or month to month, do it right and do it enough and 'value' never lies.
Screenshot_3.jpg

I get your point about the previous £1200 win countering today's losses. but I'd be VERY concerned if failed hedges had cost me north of £700 in a single day.

I know you're familiar with Guardian's max capacity on both streaming and the older API-NG modes, and also understand refresh rates versus number of loaded markets, so why would the hedge fail Shaun, assuming they are 'take SP'?

CS

[ShaunWhite]

I get your point about the previous £1200 win countering today's losses. but I'd be VERY concerned if failed hedges had cost me north of £700 in a single day.

I know you're familiar with Guardian's max capacity on both streaming and the older API-NG modes, and also understand refresh rates versus number of loaded markets, so why would the hedge fail Shaun, assuming they are 'take SP'?

CS

I don't use Guardian so no resfeshing issues, and the non-hedges are mostly down to me betting right down to the wire and sometimes having matches with the start milliseconds later. It's no big deal. I've been doing it for over 4yrs and 1.4 millions animals. I'm staking about 500 on UK horses, 3 or 4 per race so ups and downs are fairly chunky compared to when I thought a tenner was a lot.

Plus or minus 4figures in a day is a 2nd std deviation event though tbh

[Crazyskier]

I get your point about the previous £1200 win countering today's losses. but I'd be VERY concerned if failed hedges had cost me north of £700 in a single day.

I know you're familiar with Guardian's max capacity on both streaming and the older API-NG modes, and also understand refresh rates versus number of loaded markets, so why would the hedge fail Shaun, assuming they are 'take SP'?

CS

I don't use Guardian so no resfeshing issues, and the non-hedges are mostly down to me betting right down to the wire and sometimes having matches with the start milliseconds later. It's no big deal. I've been doing it for over 4yrs and 1.4 millions animals. I'm staking about 500 on UK horses, 3 or 4 per race so ups and downs are fairly chunky compared to when I thought a tenner was a lot.

Plus or minus 4figures in a day is a 2nd std deviation event though tbh

I imagine it would be with those stake sizes. Interesting!

CS

[ShaunWhite]

I imagine it would be with those stake sizes. Interesting!

You know how it grows without you realising it if you do it long enough. It's just a shame everything has a point of diminishing returns.

[pjcppts]

Thanks for all the input!

So, the Law of Sod invoked and I had my worst day ever using this method:
44/61 (72.13%) for a net loss of £16.28 Whoopy do... I'm OK with that.

On the other hand, I got the Donny placepot up for 12p and made £326 !

Roundabouts and swings! L2L £8 from tomorrow!

Cheers all.
Pete

[stueytrader]

You're right to be thinking about the value of your bets,

This post, and some other discussion on the horse forum, has got me thinking.

Some little while back I altered my approach to my trading fundamentally (pun not intended, but relevant!). I realised I was always including many fundamentals in my analysis about specific horses in races, yet in trading alone these fundamentals could often be less important - they became more important in the actual races and outcomes themselves.

So, for some while now, I've been what I'd call a hybrid trader and punter - I no longer look only for a winning trade, but also a losing or winning race selection. Many advise against doing this in the trading world. But, I have found it to be far more beneficial to my results than trading alone.

Just a question though Shaun - I see you have mentioned the idea of the value a few times - do you actively analyse for edges in value (i.e. likely outcomes now), or just assume value from the prices as roughly level for ev?

[ShaunWhite]

Just a question though Shaun - I see you have mentioned the idea of the value a few times - do you actively analyse for edges in value (i.e. likely outcomes now), or just assume value from the prices as roughly level for ev?

Pre-race it's just vs bsp. That part is fairly obvious, ie green at bsp means you're on for a profit for no cost which is undeniably value. But it's also necessary in-running although harder to gauge, but the assumption is that in an efficient market the prices are correct which makes taking prices -ev and offering prices is +ev.

But the overall 'message' is just that trading is betting (lots of bets) and when you add them up and average them out then you only make money with value. Not something that has to be on your mind all the time, but it explains situations where lots of small wins get wiped out by big losses, because you've got a stop loss that's laying the horse in 2nd place at 500/1 instead of 5/1. That's an obvious bad bet but when you're betting so much even the small % amounts add up and why traders should be more likely to offer than take.

So no huge ground breaking idea just that you don't learn much from market totals, you need to dig down and look at the bets that made it up.

[stueytrader]

Agree with all that Shaun.

Though there is one small clarification issue - which is what is the definition of value. That seems like it can be defined in terms of what the outcome of the event is (i.e. win or loss of a selection), or value in relation to other prices - that latter one is more of a trading value definition of course.