Hi everyone,
I'd appreciate some opinions and advice on outright betting, value, and dutching.
Question 1 – Value betting
Let's say I'm looking at a horse race and I identify a horse at 10/1 that I believe is significantly overpriced and represents great value.
If my assessment is correct, does that automatically imply that the favourite is underpriced and therefore offers little or no value? In other words, if I've found value elsewhere in the market, is it reasonable to view the favourite as a potential lay bet, or can both the favourite and the outsider still represent value at the same time?
Question 2 – Dutching strategy
I've become fairly confident at finding outsiders and dutching selections, but I'd like to develop a more consistent long-term strategy.
One thing I'm unsure about is how many selections I should include in a dutch. Sometimes I narrow a race down to three horses, while in other races I may have five or six.
Should I:
Limit myself to a maximum number of selections in each race?
Aim to cover only a certain percentage of the book?
Focus on maintaining a stake-to-potential-return ratio?
Or simply back any runners that I believe represent value, regardless of how many there are?
When dutching, is it generally better to
Stake to a fixed target profit if any selection wins?
My goal is to determine whether my approach can be profitable over a large sample of bets, so I'm trying to build a disciplined strategy rather than just relying on instinct.
I'd be interested to hear how experienced dutchers approach this and what has worked for you over the long term.
Many thanks
Value betting & Dutching
-
ShaunWhite
- Posts: 10829
- Joined: Sat Sep 03, 2016 3:42 am
Q1. That's only true in races with 2 runners. If there's 8 then more then 1 could be value.
Q2. Include all the ones you see as value.
"dutching" is just having more than one bet, great for punters who like to get something back regularly but not a serious method imo unless you happen to be betting on 3 horses because all 3 horses are value. Otherwise you're just paying a premium to smooth your returns (at best) or making a loss because you're including some that are really poor value.
All trading/betting is about value, whether that's backing more than one horse, or having more than one bet on one horse (eg a back closed out with a lay). If you back with a 5% expectancy and lay with a -6% expectancy you'll sometimes make money and feel like a 'trader' but the wins won't cover the losses. You've just got two bets that net to -ev.
Q2. Include all the ones you see as value.
"dutching" is just having more than one bet, great for punters who like to get something back regularly but not a serious method imo unless you happen to be betting on 3 horses because all 3 horses are value. Otherwise you're just paying a premium to smooth your returns (at best) or making a loss because you're including some that are really poor value.
All trading/betting is about value, whether that's backing more than one horse, or having more than one bet on one horse (eg a back closed out with a lay). If you back with a 5% expectancy and lay with a -6% expectancy you'll sometimes make money and feel like a 'trader' but the wins won't cover the losses. You've just got two bets that net to -ev.
Thanks for the reply.
There are quite a few races where I identify several horses that I believe represent value, so I end up dutching them.
Do you think I'd be better off narrowing it down to just one selection per race and trusting that, if my value assessment is accurate, the results will even themselves out over the long term?
The reason I ask is that there have been quite a few occasions where I've identified three horses, decided to back just one, and then one of the other two has gone on to win while my selection has finished second. It's happened often enough to make me question whether I'm making the right decision.
I suppose the answer is that if I'm consistently selecting the horse that offers the best value, then either way over a large enough sample the results should take care of themselves. Would you agree with that, or do you think there's still a strong case for dutching multiple value selections in the same race? Opinions Anyone else ??
There are quite a few races where I identify several horses that I believe represent value, so I end up dutching them.
Do you think I'd be better off narrowing it down to just one selection per race and trusting that, if my value assessment is accurate, the results will even themselves out over the long term?
The reason I ask is that there have been quite a few occasions where I've identified three horses, decided to back just one, and then one of the other two has gone on to win while my selection has finished second. It's happened often enough to make me question whether I'm making the right decision.
I suppose the answer is that if I'm consistently selecting the horse that offers the best value, then either way over a large enough sample the results should take care of themselves. Would you agree with that, or do you think there's still a strong case for dutching multiple value selections in the same race? Opinions Anyone else ??
-
ShaunWhite
- Posts: 10829
- Joined: Sat Sep 03, 2016 3:42 am
Oh ffs I typed you a decent reply and lost it.... Basically, back/lay anything that's value, might be one in a race or might be several of them.
I'm sure some others will chip in with whatever they think/do.
I'm sure some others will chip in with whatever they think/do.
-
Aberdeentrad
- Posts: 28
- Joined: Sat Jan 06, 2024 7:07 pm
Agree, back or lay anything that has value as this will increase overall EV of the race.
Difficult part might be knowing you have value, for me I place my bets pre-off and then check for value EV against BSP. My take is that if my backs are higher than BSP and lays are lower then those are good value and I calculate EV based on that, overtime profits should follow EV if BSP is fair.
If you are backing / laying at BSP based on a model using form / other factors then I guess you just need to compare the long term results against expected outcome to confirm that the model is finding value, i.e winning more often than the odds suggest.
Key thing is to know you are definitely getting value as anything can look good with some positive varience.
Difficult part might be knowing you have value, for me I place my bets pre-off and then check for value EV against BSP. My take is that if my backs are higher than BSP and lays are lower then those are good value and I calculate EV based on that, overtime profits should follow EV if BSP is fair.
If you are backing / laying at BSP based on a model using form / other factors then I guess you just need to compare the long term results against expected outcome to confirm that the model is finding value, i.e winning more often than the odds suggest.
Key thing is to know you are definitely getting value as anything can look good with some positive varience.
Kelly-Optimal Counterexample (Haigh, 2003) for Reflection?
Even though only Charlie and Alpha are positive Expected-Value (EV) bets, Kelly adds the slightly negative EV runner, Bravo, to the mix. The additional stake on Bravo reduces the overall variance of outcomes from this single event and simultaneously reduces the long-term Risk-of-Ruin (RoR) for the sports trader.
Note: the maximum total stake should be scaled from 22.8% to no more than 5% of bankroll (RoR)!
Even though only Charlie and Alpha are positive Expected-Value (EV) bets, Kelly adds the slightly negative EV runner, Bravo, to the mix. The additional stake on Bravo reduces the overall variance of outcomes from this single event and simultaneously reduces the long-term Risk-of-Ruin (RoR) for the sports trader.
Note: the maximum total stake should be scaled from 22.8% to no more than 5% of bankroll (RoR)!