Thought I would start a topic for value. I have been creating my own betting forecast for my trading for a few months and have found it really helpful in finding value and good trades which help support those moves a little bit more and think a collection of knowledge can help us all.
I always create my book at 100%. My book is always created with no current price or book influence and entirely my own opinion.
I am using this as an experiment and hopeful we can all follow along,
I will use my trading knowledge using the same stakes to trade these races on betdaq with only one trade per runner allowed and any price that is priced over my forecast will be staked at sp at 1 minute to post time for the same stakes.
Obviously, it may not be possible to do this with 100% attention as I will be focusing on trading every day as normal but will chose a handful of races each day that I have no intention of trading and today I have two.
My stakes will be as follows. 1.01-2.00 £25 stake.2-3s £17.50 3-4s £12.50. 4-6s £6. 6-10s £3 anything over £10 will be £2 stake.
I stick with flat and AW racing (well currently just AW racing).
Chosen Lingfield today:
1.30 Lingfield.
1. Miss Munity 4.0
2. Assanilka 3.0
3. Lava Light5.0
4. Highland Pass10
5. French Mix 10
6. Hipz 10
2.00 Lingfield.
1. Clement 6.5
2. Iron Ryan 5.0
3. Outer Space 3.0
4. Varsovian 5.5
5. Captain Lars 10
6. Mr Andros 15
Section for Value
- SeaHorseRacing
- Posts: 2893
- Joined: Fri May 20, 2016 7:06 pm
Last edited by SeaHorseRacing on Fri Feb 02, 2018 6:59 pm, edited 1 time in total.
- SeaHorseRacing
- Posts: 2893
- Joined: Fri May 20, 2016 7:06 pm
Iron Ryan was 15 to 20ish when I posted my forecast. 6s now. Pretty pleased with that assessment.
- SeaHorseRacing
- Posts: 2893
- Joined: Fri May 20, 2016 7:06 pm
About £2.80 profit on betting but made just short of £5 trading. Will post results when I can. Interesting.
Not heard of Kelly?SeaHorseRacing wrote: ↑Fri Feb 02, 2018 11:50 amMy stakes will be as follows. 1.01-2.00 £25 stake.2-3s £17.50 3-4s £12.50. 4-6s £6. 6-10s £3 anything over £10 will be £2 stake.
Backing horses to return the same amount (which approximates to SHR's staking plan) is a form of the Kelly principle that assumes that the true chances of winning is very slightly greater than the odds. It's useful for testing strategies without spending excessive time calculating true chances.LinusP wrote: ↑Fri Feb 02, 2018 2:24 pmNot heard of Kelly?SeaHorseRacing wrote: ↑Fri Feb 02, 2018 11:50 amMy stakes will be as follows. 1.01-2.00 £25 stake.2-3s £17.50 3-4s £12.50. 4-6s £6. 6-10s £3 anything over £10 will be £2 stake.
It makes me laugh when people test systems/tipsters to level stakes, using the same stake for a 33/1 shot as they would an even money favourite.
- SeaHorseRacing
- Posts: 2893
- Joined: Fri May 20, 2016 7:06 pm
Derek27 wrote: ↑Fri Feb 02, 2018 3:17 pmBacking horses to return the same amount (which approximates to SHR's staking plan) is a form of the Kelly principle that assumes that the true chances of winning is very slightly greater than the odds. It's useful for testing strategies without spending excessive time calculating true chances.LinusP wrote: ↑Fri Feb 02, 2018 2:24 pmNot heard of Kelly?SeaHorseRacing wrote: ↑Fri Feb 02, 2018 11:50 amMy stakes will be as follows. 1.01-2.00 £25 stake.2-3s £17.50 3-4s £12.50. 4-6s £6. 6-10s £3 anything over £10 will be £2 stake.
It makes me laugh when people test systems/tipsters to level stakes, using the same stake for a 33/1 shot as they would an even money favourite.
Ah ok, I didnt know it was called Kelly principle. Yes I agree I have just done a rough method to try and make stakes abit more cleaner but this is just a test to compare difference between trading and betting and hopefully learn something.
Yeh your stake should be a function of your edge/bankroll.Derek27 wrote: ↑Fri Feb 02, 2018 3:17 pmBacking horses to return the same amount (which approximates to SHR's staking plan) is a form of the Kelly principle that assumes that the true chances of winning is very slightly greater than the odds. It's useful for testing strategies without spending excessive time calculating true chances.LinusP wrote: ↑Fri Feb 02, 2018 2:24 pmNot heard of Kelly?SeaHorseRacing wrote: ↑Fri Feb 02, 2018 11:50 amMy stakes will be as follows. 1.01-2.00 £25 stake.2-3s £17.50 3-4s £12.50. 4-6s £6. 6-10s £3 anything over £10 will be £2 stake.
It makes me laugh when people test systems/tipsters to level stakes, using the same stake for a 33/1 shot as they would an even money favourite.
Code: Select all
edge = odds/prediction-1.0 # can be clipped as full kelly is a bit scary
kelly_size = edge/(odds-1.0)*bankroll
- SeaHorseRacing
- Posts: 2893
- Joined: Fri May 20, 2016 7:06 pm
Anyone have a spreadsheet on that formula or could pinpoint me in the direction of doing so in excel?LinusP wrote: ↑Fri Feb 02, 2018 4:01 pmYeh your stake should be a function of your edge/bankroll.Derek27 wrote: ↑Fri Feb 02, 2018 3:17 pmBacking horses to return the same amount (which approximates to SHR's staking plan) is a form of the Kelly principle that assumes that the true chances of winning is very slightly greater than the odds. It's useful for testing strategies without spending excessive time calculating true chances.
It makes me laugh when people test systems/tipsters to level stakes, using the same stake for a 33/1 shot as they would an even money favourite.
Code: Select all
edge = odds/prediction-1.0 # can be clipped as full kelly is a bit scary kelly_size = edge/(odds-1.0)*bankroll
The old Timeform phone service used to stake to return 10% of bank, if betting on the true chance odds. i.e. 5% stake on an even money chance, 2% on a 4/1 chance, 1% on a 9/1 chance.
This is exactly the same as using the Kelly criterium where you assume the odds you get are 11% higher that the true odds, 10/1 for a 9/1 chance or 5/1 for a 9/2 chance.
This is exactly the same as using the Kelly criterium where you assume the odds you get are 11% higher that the true odds, 10/1 for a 9/1 chance or 5/1 for a 9/2 chance.
- SeaHorseRacing
- Posts: 2893
- Joined: Fri May 20, 2016 7:06 pm
8.45 Wolverhampton
1. Gold Flash 10.0
2. Chaucer's Tale 8.0
3. Joey's Destiny 8.0
4. Awesome Allan 23.0
5. Lucky Lodge 5.5
6. Kafoo 4.0
7. Corporal Maddox 12.0
8. Prince Jai 13.0
9. Frank The Barber 34.0
1. Gold Flash 10.0
2. Chaucer's Tale 8.0
3. Joey's Destiny 8.0
4. Awesome Allan 23.0
5. Lucky Lodge 5.5
6. Kafoo 4.0
7. Corporal Maddox 12.0
8. Prince Jai 13.0
9. Frank The Barber 34.0
-
- Posts: 3140
- Joined: Sun Jan 31, 2010 8:06 pm
Why will you only be backing? Surely a lot of value will be available to lay also, i.e. you have Gold Flash at 10 and it's currently 4.5. Also what 'edge' will you be taking , backing at any price over your 100% forecast book would give you little worthwhile edge on many runners.SeaHorseRacing wrote: ↑Fri Feb 02, 2018 11:50 amI will use my trading knowledge using the same stakes to trade these races on betdaq with only one trade per runner allowed and any price that is priced over my forecast will be staked at sp at 1 minute to post time for the same stakes.
I'm not having a dig btw as I've done similar things in the past making my own tissues and all it showed me was I was useless at it, but also the providers of newspaper tissues were just as bad The other thing I always noticed with out of line tissues is that the value to be had is usually early doors as markets do start to correct the more people that enter them. That's why it's very easy for people like Hugh Taylor to show long term profits by picking off and quoting bets on bady priced runners where little if any real money is actual laid at the prices.
Last edited by spreadbetting on Fri Feb 02, 2018 8:15 pm, edited 1 time in total.
- SeaHorseRacing
- Posts: 2893
- Joined: Fri May 20, 2016 7:06 pm
Thanks for your input buddy.spreadbetting wrote: ↑Fri Feb 02, 2018 6:49 pmWhy will you only be backing? Surely a lot of value will be available to lay also, i.e. you have Gold Flash at 10 and it's currently 4.5. Also what 'edge' will you be taking , backing at any price over your 100% forecast book would give you little worthwhile edge on many runners.SeaHorseRacing wrote: ↑Fri Feb 02, 2018 11:50 amI will use my trading knowledge using the same stakes to trade these races on betdaq with only one trade per runner allowed and any price that is priced over my forecast will be staked at sp at 1 minute to post time for the same stakes.
I'm not having a dig btw as I've done similar things in the past making my own tissues and all it showed me was I was useless at it, but also the providers of newspaper tissues were just as bad The other thing I always noticed with out of line tissues is that the value to be had is usually early doors as markets do start to correct the more people that enter them. That's why it's very easy for people like Hugh Taylor to show long term profits by picking off and quoting bets on bady priced runners where little if any real money is actual laid at the prices.
One other thing would be remove any reference to the wife's account, they do read forums.
I just want to try and build a comparison to see if trading based on a tissue price. For example. Only interested in a selection to trade if I believe it has been priced incorrectly.
I am also really curious to know if and how many winners are winning much shorter then what i think they should be, there's no point backing value in every race if 75% of races are won by horses that are priced too short. Just found this way of thinking really interesting lately.
After your input it has alswo got me thinking.
Is it easier to find a value bet or is it easier to find a value lay? Are most winning horses overpriced or are they underpriced. Is their a bias in any way?
For example, On the AW Charlie Apple Godolphin newcomers, are always favourite and are always short. Surely these are priced too short..
- SeaHorseRacing
- Posts: 2893
- Joined: Fri May 20, 2016 7:06 pm
I didn't even think about laying until spreadbetting said so
But thinking about it, taking value it should all be about turnover and laying would increase that... wouldn't it?