Impact of stake on automation

A place to discuss anything.
Post Reply
User avatar
jamesedwards
Posts: 2234
Joined: Wed Nov 21, 2018 6:16 pm

Having spent several years working on hundreds of automations across horses, dogs and football I am coming to the conclusion that even tiny adjustments in stake can have a devastating impact on ROI%.

Time after time after time I have something that roughly returns 100% or better only for it to plunge into loss the moment I dial up the stakes. So far I've always moved on to the next attempt figuring it was just my bad luck but I've reached the point where I figure I just can't be this unlucky.

Here's my latest example. 2,500 football markets over 11 days. Everything to the left of the red line is based on £5 stake, everything to the right is £10 stake.
Capture.PNG

This is just one example of dozens through the years. I'm eager to get the views of others. Is this an epiphany that I should have come to a long time ago or am I just doomed to eternity of bad luck? Can such small increases in stake really make that much difference to profitability?

Views please.
You do not have the required permissions to view the files attached to this post.
User avatar
gazuty
Posts: 2547
Joined: Sun Jun 26, 2011 11:03 am
Location: Green land :)

My views on this are in the scaling up section of my longer guidance post - viewtopic.php?f=17&t=23699

I find most markets quite sensitive to stake size, generally in proportion to volume and the Goldilocks principle tends to apply. Obviously experimenting is required to find the Goldilocks points for a particular strategy for a particular type of market.
LinusP
Posts: 1871
Joined: Mon Jul 02, 2012 10:45 pm

Yes, it’s the hardest part of the game :D

However you need to confirm it was the stake increase, can you analyse your matching rates (winners and losers) against your stake size?
dm1900
Posts: 71
Joined: Sun Jan 15, 2017 10:02 pm

This is one of the most interesting discoveries one ends up making. The problem shifts slightly from "how do I identify value" to "how do I execute in meaningful size while retaining match rates and thwarting adverse selection as much as possible"
User avatar
napshnap
Posts: 1189
Joined: Thu Jan 12, 2017 6:21 am

Maybe you triggered other's bots to hop in front of your orders by increasing your stakes. Happens a lot in illiquid bigspreaded early markets.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

Have a crossover period when you're staking between £5 and £10 randomly rather than switching from one to the other, and produce a chart sorted by stake>date rather than just by date.

Date sorted PLs like your chart are ok for showing the effect over time but sorting by other things such as stake, price, time of day, day of week etc etc will tell you about those factors. Eg a PL sorted by price should trend evenly across the x-axis price range, but you might find those above or below a certain price point trend the opposite way. Date is interesting to a point and everyone produces date sorted PLSs, but they don't dig into what's actually making or losing money in the same way putting something other than date on the x-axis can.
User avatar
jamesedwards
Posts: 2234
Joined: Wed Nov 21, 2018 6:16 pm

ShaunWhite wrote:
Mon Jun 07, 2021 3:05 pm
Date sorted PLs like your chart are ok for showing the effect over time but sorting by other things such as stake, price, time of day, day of week etc etc will tell you about those factors. Eg a PL sorted by price should trend evenly across the x-axis price range, but you might find those above or below a certain price point trend the opposite way. Date is interesting to a point and everyone produces date sorted PLSs, but they don't dig into what's actually making or losing money in the same way putting something other than date on the x-axis can.
Done all this already to get a profitable automation as currently stands. But then up the stakes and wallop! :lol:

ShaunWhite wrote:
Mon Jun 07, 2021 3:05 pm
Have a crossover period when you're staking between £5 and £10 randomly rather than switching from one to the other, and produce a chart sorted by stake>date rather than just by date.
That's a plan. Will try this.
User avatar
jamesedwards
Posts: 2234
Joined: Wed Nov 21, 2018 6:16 pm

Thanks for all comments and suggestions so far. Very interesting!

Stake back down to £5 today and lo-and-behold it has been profitable again so far.
LinusP
Posts: 1871
Joined: Mon Jul 02, 2012 10:45 pm

jamesedwards wrote:
Mon Jun 07, 2021 6:31 pm
Thanks for all comments and suggestions so far. Very interesting!

Stake back down to £5 today and lo-and-behold it has been profitable again so far.
Now it comes down to how can you get more of those £5 stakes matched ;) (you might find you can make more money by reducing that stake further)
User avatar
jamesedwards
Posts: 2234
Joined: Wed Nov 21, 2018 6:16 pm

LinusP wrote:
Mon Jun 07, 2021 7:01 pm
jamesedwards wrote:
Mon Jun 07, 2021 6:31 pm
Thanks for all comments and suggestions so far. Very interesting!

Stake back down to £5 today and lo-and-behold it has been profitable again so far.
Now it comes down to how can you get more of those £5 stakes matched ;) (you might find you can make more money by reducing that stake further)
Problem is I'm much more interested in churning volume (to generate Commission towards Premium Charge) than making profit.
LinusP
Posts: 1871
Joined: Mon Jul 02, 2012 10:45 pm

jamesedwards wrote:
Mon Jun 07, 2021 7:08 pm
LinusP wrote:
Mon Jun 07, 2021 7:01 pm
jamesedwards wrote:
Mon Jun 07, 2021 6:31 pm
Thanks for all comments and suggestions so far. Very interesting!

Stake back down to £5 today and lo-and-behold it has been profitable again so far.
Now it comes down to how can you get more of those £5 stakes matched ;) (you might find you can make more money by reducing that stake further)
Problem is I'm much more interested in churning volume (to generate Commission towards Premium Charge) than making profit.
Chasing 40% instead of straight 60?
User avatar
jamesedwards
Posts: 2234
Joined: Wed Nov 21, 2018 6:16 pm

LinusP wrote:
Mon Jun 07, 2021 8:04 pm
jamesedwards wrote:
Mon Jun 07, 2021 7:08 pm
LinusP wrote:
Mon Jun 07, 2021 7:01 pm


Now it comes down to how can you get more of those £5 stakes matched ;) (you might find you can make more money by reducing that stake further)
Problem is I'm much more interested in churning volume (to generate Commission towards Premium Charge) than making profit.
Chasing 40% instead of straight 60?
Indeed!
User avatar
jamesedwards
Posts: 2234
Joined: Wed Nov 21, 2018 6:16 pm

Amazing. Exact same automation now back in consistent profit since moved back to £5 stakes.
Capture2.PNG
You do not have the required permissions to view the files attached to this post.
Post Reply

Return to “General discussion”