How long does it take to judge whether a system is working?

Trading is often about how to take the appropriate risk without exposing yourself to very human flaws.
Post Reply
FrogThimble
Posts: 124
Joined: Mon Dec 11, 2017 6:26 pm

Saturday was brutal... My system showed its flaws at around the 100 trades point when something I hadn't anticipated happened and the previous day's profits got wiped out. Back to the drawing board for me again. :D
xitian
Posts: 457
Joined: Fri Jul 08, 2011 2:08 pm

"Back to the drawing board" sounds a bit drastic. There could still be legs in your strategy if you can account for the "unanticipated" situation that caused the losses. If you can prevent those without minimising the amounts of all the previous wins then you could potentially still be quite positive overall.

If you plot a chart of cumulative total P&L for all the markets (or trades) what does it look like?

Re: the £400 of losses testing previous strategies - consider it the start up expense like any business. Commit as much as you are happy to a start up fund and use it sensibly. It needs to be money you can lose and not worry about, since there's likelihood any/most businesses don't get off the ground. But don't be afraid of using it. Just be sensible. Even better - learn the skills to collect data and backtest stuff without risking money. Admittedly this is a bit harder with in-play strategies.
FrogThimble
Posts: 124
Joined: Mon Dec 11, 2017 6:26 pm

xitian wrote:
Sun Apr 15, 2018 4:17 pm
"Back to the drawing board" sounds a bit drastic. There could still be legs in your strategy if you can account for the "unanticipated" situation that caused the losses. If you can prevent those without minimising the amounts of all the previous wins then you could potentially still be quite positive overall.

If you plot a chart of cumulative total P&L for all the markets (or trades) what does it look like?
The chart is this:
graph.jpg
It would only have taken one more bad trade to wipe out all profit. I have a few ideas about ammending the method to try to reduce the risks - which is what I mean about the drawing board really. Major tinkering rather than a total rethink. I'll try that from tomorrow going back to minimal stakes.
You do not have the required permissions to view the files attached to this post.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

It's obviously a high strike rate strategy and that usually goes hand in hand with larger losses. If you can stop the big ones then all well and good but you still have a positive trend.

What was the thing that happened that you didn't expect? Technical failure? Or a genuinely very unusual event?
User avatar
marksmeets302
Posts: 527
Joined: Thu Dec 10, 2009 4:37 pm

Shaun is right, you have a strategy that has the occasional big loss, and they determine how well you do. After about 30 to 40 of these big losses you will have a good idea if your system is worth the effort. Right now there is no telling if they occur too frequently.

You might want to trade in some of the upside to lower the risk, As is this will be very hard to trade emotionally. You *will* encounter multiple back to back big losses en you're almost guaranteed to throw in the towel at that point. Another option is to trade this with low stakes in a portfolio of strategies.
FrogThimble
Posts: 124
Joined: Mon Dec 11, 2017 6:26 pm

ShaunWhite wrote:
Sun Apr 15, 2018 11:14 pm
It's obviously a high strike rate strategy and that usually goes hand in hand with larger losses. If you can stop the big ones then all well and good but you still have a positive trend.

What was the thing that happened that you didn't expect? Technical failure? Or a genuinely very unusual event?
I attempted a trade on a horse at a very short price which then fell the second after I'd backed it before I could get a chance to hedge it. Obviously this can and will happen but I was already exposed in that race due to the poor performance of the first horse I'd tried to trade on and I should have just let the original loss ride - even though it would have broken my then rule to do so. That's one of the rules I'm changing... I have to be willing to accept medium sized losses to cut the risk of huge losses even if it means fewer wins.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

FrogThimble wrote:
Mon Apr 16, 2018 2:52 pm
I attempted a trade on a horse....... which then fell the second
Perhaps the strategy isn't suitable for situations with a high degree of jeopardy, like fences, or sprints or amateur riders. Or maybe that's what's driving your gains? It's so easy to just say weed out the losers, but you really have to mind you don't throw the baby out with the bath water.
FrogThimble
Posts: 124
Joined: Mon Dec 11, 2017 6:26 pm

ShaunWhite wrote:
Mon Apr 16, 2018 3:09 pm

Perhaps the strategy isn't suitable for situations with a high degree of jeopardy, like fences, or sprints or amateur riders. Or maybe that's what's driving your gains? It's so easy to just say weed out the losers, but you really have to mind you don't throw the baby out with the bath water.
All these things are possible. Today I'm playing all of the UK and Irish races regardless of distance or format. One day won't give me much insight but it's certainly possible I've been choosing some of the wrong races and missing some of the ones I ought to have played.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

FrogThimble wrote:
Mon Apr 16, 2018 3:33 pm
All these things are possible.
I once heard the Sage of Hook say that it's not enough to know that something is working in case A but not in case B, you also need to understand why it's working. Or maybe I got that from a fortune cookie, I can't remember.
FrogThimble
Posts: 124
Joined: Mon Dec 11, 2017 6:26 pm

ShaunWhite wrote:
Mon Apr 16, 2018 4:22 pm

I once heard the Sage of Hook say that it's not enough to know that something is working in case A but not in case B, you also need to understand why it's working. Or maybe I got that from a fortune cookie, I can't remember.
I'm not doing too well today in terms of net results. Across the first 20 races my bank is down to 90.7% on the start of the day.

However... if I'd only played the races of 10 furlongs or longer my bank would be at 102.8% of the start of the day.

On the 12 races of 10 furlongs or greater I've won on 11/12 with an average 23p profit
On the 8 races of less than 10 furlongs I've won on just 2/6 with an average loss of -£1.63

Very small sample but I can see why it's happening that way and it's enough to persuade me to give the short races a miss.
Charon
Posts: 70
Joined: Tue May 25, 2010 1:09 pm

Just curiosity from my side; is your strategy based on historical stats? Is there any race-selection involved?
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

FrogThimble wrote:
Mon Apr 16, 2018 5:46 pm
Across the first 20 races my bank is down to 90.7% on the start of the day.

Very small sample but I can see why it's happening that way and it's enough to persuade me to give the short races a miss.
Test, refine, repeat ad nauseum.

Tiny, tiny, tiny stakes. I'd have my liability at about 0.01% of my R&D bank for any one loser.
10% in an afternoon is going to bust you before you find anything, but I think you're realising that.
FrogThimble
Posts: 124
Joined: Mon Dec 11, 2017 6:26 pm

Charon wrote:
Mon Apr 16, 2018 6:12 pm
Just curiosity from my side; is your strategy based on historical stats? Is there any race-selection involved?
No, it's not based on any historical stats. It was (or is) based purely on the mathematics. I understands maths rather more than I understand horses, to be honest. There is race selection but, again, that's on mathematical grounds too.
FrogThimble
Posts: 124
Joined: Mon Dec 11, 2017 6:26 pm

ShaunWhite wrote:
Mon Apr 16, 2018 6:31 pm


Tiny, tiny, tiny stakes. I'd have my liability at about 0.01% of my R&D bank for any one loser.
10% in an afternoon is going to bust you before you find anything, but I think you're realising that.
I've taken the last 6 days off in order to think through and refine my strategy. I've come back today to begin again with my new rules in place and a really tiny experimental bank to see if I really can do this before I play with bigger money.

The things I've concluded in my time off are:

1) The need for a stricter escape safety Servant that will cut my strike rate but also cut the size of my losses on the bad trades.

2) Forget about Irish races altogether - for some reason which I can't work out they seem to defy the mathematics of my strategy in a crazy way. So it's just UK races from now on. I might be totally wrong on this but I'll not lose many chances by swerving them, I suspect.

3) Be more selective in which races I try to trade on. I now have a new mathematical rule for that which prevents me trading on horses that are either side of a specific price range.

4) I was wrong about short races being a problem as such... it was other factors at play in them which made me think they were a problem.

5) Increase stake size at the start of each day only rather than at the start of each race (Effect being that during any one day each race should see me staking a slightly smaller percentage of my bank before reverting to my default percentage to begin the next day).


So on my first day back, with this new more conservative approach, my rules only allowed me to trade on 8 of today's races. My old rules would have seen me attempt 26 trades today... So these refinements will seriously limit my activities but, hopefully, not my overall profit if I am basically right... Though with sitting out so many races I am a bit concerned about the boredom factor and breaking of rhythm.

Anyway, here is today's results:

Wincanton 2:15pm: +£0.27
Wincanton 3:20pm: +£0.10
Market Rasen 3:30pm: +£0.09
Market Rasen 4:00pm +£0.09
Stratford 4:10pm +£0.09
Wincanton 4:25pm +£0.25
Market Rasen 4:35pm -£0.02
Wincanton 5:30pm +£0.09
Day 1.jpg
I will update this thread each day in order to keep myself honest and within my rules.
You do not have the required permissions to view the files attached to this post.
User avatar
ruthlessimon
Posts: 2094
Joined: Wed Mar 23, 2016 3:54 pm

FrogThimble wrote:
Sun Apr 15, 2018 10:41 pm
The chart is this:
Thought you might be interested in this Frog (following strategies all created with 1mth of data)

The following image is a pre-off strategy - with 10 variations - each variation is an extra tick require to enter when the criteria are met. (i.e. 1 = criteria met & 1 tick extra, 2 = criteria met & 2 ticks extra etc etc)

Image

What's interesting is how similar my initial strategy looks to yours - & how a solid looking methodology quickly degrades when small changes are made - hence a big query as to whether the initial strat was even an edge - or I'm just proving certain pre-off markets are inherently mean-reverting :lol:
Post Reply

Return to “Trading Psychology”