Dutch Backing Longshots
 

Back at the beginning of the year, when as a newbie on this forum looking for ideas on Laying for a profit, I hung on to the "coat tails" of Speedy and MG, following their every move laying neural longshots. Now the boys did reasonably well, but one thing that was always in the back of my mind, who in their "right mind" is on the opposite side of the Lay bets Backing these 100:1 shots? Surely they're never going to see a profit!
As I got more involved I started to see the blown out dividends the Layers were accepting compared with tote prices. Where's the value, not for the Layer that's for sure! So where is the value, of course on the Backers side of the equation.
Now Beton provided a spreadsheet awhile back of the % chances of winning in order of favourites based on the 1st favourites price. Taking an example of R5 Mackay yesterday. The BF Back prices read: $1.75, $7, $9.20, $11, $14.5, $44, $100. Now if you refer to Beton's chart for a $1.70 favourite the actual long term percentages of each favourite winning are: 50%, 19%, 10%, 7%, 5%, 4%,2%. So as we're looking for longshot value if we exclude the first 4 horses, that represent 84% winning chances, and Back the rest. Well that leaves a 16% probability that one of the last 3 horses will get up and win. If we Dutch Back these 3 horses for a 16% chance of winning what price could we get? $6 is fair value, happy with a $50 profit??
Well for a $100 stake we put $68 on the $14.50 horse, $22 on the next and $10 on the last. The $14.50 horse "Help Me Rhonda" gets up and you help yourself to a $886 profit!!
NOW I know why Backers support the longshot end of the market. Sounds great but could you withstand the run of outs max 43 races?
Cheers RP

Looked at trying to get this working many times Rinconpaul & if you want a good example of how well it can work, take a look at Caulfield R4 last Saturday, the winner ~ $180bfsp, with every other runner pretty much given a chance (go figure).

Just for kicks I ran the following for last week -

BFSP $20-$999
Tab/BFSP ratio <1.0
Dutch stake $50

results +$6639
6/05/2013 -$600.0
7/05/2013 $1,302.2
8/05/2013 -$493.8
9/05/2013 $535.6
10/05/2013 -$496.8
11/05/2013 $7,341.9
12/05/2013 -$950.0


money for jam?;)

Maybe Speedy and MG were on the right side!
I just looked at 90 races last Saturday. If you Dutch Backed the last horses representing 15% chance of winning, you would have had 8 smiles ($4,580), 82 accidents (-$8,200) for a nett loss of $3,620!
RP

Stugots, we have a big discrepancy, I rechecked Saturday the 11th and found only one more winner using your min $20 BFSP. Still shows a loss of $2,939.
The winners were:
Caulfield R1 $935
Caulfield R4 $756
Kembla Grange R7 $370
Morphetville R2 $494
Morphetville R5 $481
Rockhampton R1 $439
Rockhampton R6 $447
Rosehill R8 $428
Toowoomba R4 $711
A total of $5061 nett after stake taken out.
Less 80 losing races @ $100 = $8,000
Nett loss $2,939
I included Belmont, Carnarvon, Caulfield, Cranbourne, Darwin, Eagle Farm, Gold Coast, Kembla Grange, Morphetville, Rockhampton, Rosehill & Toowoomba. Maybe you only targetted metro races?
Can you check, because I would love to think you are right and I'm wrong!
Cheers RP

Could be differences in our SP figures as I have double checked with the same results - try

BFSP $25-$999
Tab/BFSP ratio <1.0
Dutch stake $50

6/05/2013 -$600.0
7/05/2013 $968.1
8/05/2013 -$493.8
9/05/2013 $642.1
10/05/2013 -$446.8
11/05/2013 $7,391.9
12/05/2013 -$950.0


The trouble with this is of course the runs of outs & big drawdowns - it's a bit 'all or nothing' & why I didnt keep following it way back when.

week ending may5th -$2134 :(

AHH!! I can see what's going on now. Because I'm dutching horses with chances totalling 15%, in Caulfield R4 I have dutched 3 horses with divs of $17, $19, $190 which results in a payout of $756.
You're dutching any horse $20 plus so there's only one horse in that race that qualifies and it's "Primitive Man" @ $190, so you get a payout of $1,600 for that race.
I'll go back over the other races now and recheck.
Cheers RP

Actually the full $50 would have been on Primative Man as I am betting the full stake each race

All good mate, $7,204 using BF Back prices less commission $6,736. But hang on this is for a $100 stake not $50? What's going on?
Cheers RP

Whoops!!
 

That's what happens when you get up at 3.00am, brains fried! That figure should be $18,900 plus the other wins equals $24,504 less commission = $22,911. less losses $8,000 = Nett profit of $14,911.
I'm booking my round the world trip NOW!!
Cheers RP

my fried brain coming to the party?;)

Stugots, would have been a good day Monday at Dubbo, 240:1 Brave Jay got up. $1,306 for the day. Tuesday no good, a loss of $2,300. All $100 stake.
RP

Rinconpaul, if I get the time tonight I will run the last 12 months, but if my previous endevours in this area are anything to go by, it will not make for good reading.

Cheers mate. I'll tell you how lucky I am. I ran a test today with $5 bets Dutching >=$20. I had a couple of early wins but also a lot of losses so after losing $100 the bot was set to stop betting. It stopped betting just before a 42:1 shot got up at Sandown R8 and then a 100:1 Canterbury R7!!
Can you beleive it?
RP

Dutch backing longshots
 

Hi Riconpaul

Can you provide the link to Betons spreadsheet if possible please?
Like you I am fairly new to the forums. I reckon your inputs have created a lot of new interest with different slants and ideas on dutching and lay betting. I am sure fellow forum members would agree.

Age has caught up with me but I am still having fun dabbling with dutching and lay betting following ideas from the more experienced punters like Stugots, Speedy and MG. I am sure I have missed numerous others.

Thanks Ted

ooo yeah, the punting gods are a pack of **********;)

will post last 12 months results in the morning

It's a thread called "first five statistics" January 2013. I'm going to embellish it tomorrow by inserting $ values in lieu of % to make it quicker to spot an under/overlay, which I'll post for anyone interested. Thanks against to Beton for providing the original.
Cheers RP

Please remember that the sheet is "to tote" BF is different.

Your much better off dutching all runners under $30 then all those over $30.

UB, are you referenceing BF SP prices in this statement? What do you do if the dutching calculator runs out of profit before it gets down to the 29:1 shot, NO bet?
Why not Top & Tail it by Dutch Back < $30 and Lay >= $30?
I perceived there was more value in Backing the longer priced horses because they are massive overlays to the tote price or true chance of winning %?
Interested to hear your thoughts or confirmed stats UB, thanks. RP

No not the SP just the fixed prices as then you know what prices you are getting.

I ran back a few days quickly looking at races where all horses under $30 could be dutched for < 90%. Using the lower available fixed odds. It won 25 out of 26 races. Profit on dutching to win $100 returned $311.05

So that's a 3.45 unit profit (311.05/90) and a 96% strike rate over the short time frame I looked.

It needs further investigation but its a promising start.



Ever heard of the fav long shot bias. Well the massive overs on betfair just bring this closer to remove the bias. For example a $100 horse on the tote should win 1% of races. It might be $400 on betfair. That's still only a difference of 0.75% chance difference. The horses real chances are more likely closer to 0.75% i9n reality then 1% due to the fav long shot bias.

The difference between a horse at $1.70 on the tote and $1.75 on betfair is 2.9%. That's a whole lot more then 0.75% value that your chasing in the long shots.

Work on percentages... not price.

Gotcha! Thanks for the reply UB, I'll check it out further. It will be interesting to see the results Stugots comes up with too.
Cheers RP

Morning UB. Is there a 'quick n dirty' way to get to the '<90%' mark for back testing purposes?

Cheers LG

Ok, well not as bad as I expected with a 7.5% loss on t/o for 12 months to last Sunday.

If your not immune to a bit of filtering, the following produced a profit in 9 of 12 months, but not a lot of action with an avg of 2 bets per day.

BFSP $20-$999
Tab/BFSP ratio <1.0
Non maidens only
Tatts codes br cr mr nr sr vr
Distance >=1200 to <=1550
No starters >=9 <=12
Race starts average >=4
Favourite bf price <1min to jump >$3.50

GOOD WORK Stugots!!
To clarify, do the filter conditions that you detailed, result in the 7.5% loss, profit 9 in 12, or was that applicable to no filters BFSP>=$20?
Thanks mate, RP

Hi RP, the 7.5% loss was for the lot ~ 6400 races excluding maidens.

The filters produced a pot of 45%, a bit high for my liking but I don't think any of the filters are outrageously back fitting, so who knows.

A quote of mine from the Lay the Neural Outsiders thread with reference to Speedy & MG,"If I were you guys I'd let the systems die off, lock the formula in the same vault that the CocaCola recipe is held in. Just post annual results to us from your Swiss chalets. In years to come forumites will be trolling the net looking for these legendary systems....lol"

Stugots, how about we meet up in Zurich for a drink in 6 months time, when we go visit Speedy & MG in their chalet??...lol
have a great day, RP

might be a struggle RP as I think I will be still on the QE2 round the world cruise;)

Well 90% can only be determined on a race to race basis. Its 100/odds and making sure all those under 30 add up to less then 90. You don't get heaps of bets (about 8-10 a day in Aus).

I'll check todays figures in the next 20mins.

7 from 8 today. Resulted in +20.01 on $100 dutch plan.

Useless Bettor
Could you give an example of a race that fits this criteria. I,m not sure I've got it correct. Thanks

I don't know if this interests anyone who backs longshots but yesterday's Ipswich 5/12 paid $54 on Unitab and $74 on Betfair - it ran third at it's last start.

Scone 8/9 paid $31 on Unitab and $40 on Betfair - it ran second at not last start but at it's previous start.

I'll try and help Clive. Look at Betfair's Racing Form and Results page, a race from yesterday Albany Race 6. If you were Dutch backing the divs read:
3.80 4.50 6.80 7.40 9.20 21 27 30 40
If you enter these figures up to 27 in a Dutching calculator (www.oddschecker.com/betting-tools/dutching-calculator.html) for 100 staked, you will get the following profit $4.07. In other words if you dutched the first 7 horses in order of price and one of those horses won, you would get a profit of $4.07 less commission = $3.80. What you need to get is at least $10 so this bet, although in profit, is poor value.
Remember too that you haven't got time to be entering figures manually into a calculator as they change all the time so you need a bot that only fires a bet if your desired profit is met.
Interestingly, in this particular case, if you dutched the first two or three horses there was good value but that's another story.
Hope that helps, RP

Thanks Rinconpaul

I get it now

Yes MG, it would have been a good day for longshot backers yesterday with Taystar $85, Polar Blitz $95 & Get That Indya $32 getting up. You should have seen the Quadrella $20,392.
Don't worry my brain's been ticking over, the two longshots were identified by my Longshot DNA filters as potential winners but you still had a mix of short priced horses in the other two legs, which ones to pick, the age old problem??
Have a great day, RP

AUSDubb (AUS) 13th May	R5 1600m CL3	13/05/13 14:58	7. GRAND IAM	12	8.2	8.8	12.1951219512
AUSDubb (AUS) 13th May	R5 1600m CL3	13/05/13 14:58	4. RUN LEGLESS	12	17	17.5	5.8823529412
AUSDubb (AUS) 13th May	R5 1600m CL3	13/05/13 14:58	2. BOARDWALK	12	10.5	11	9.5238095238
AUSDubb (AUS) 13th May	R5 1600m CL3	13/05/13 14:58	3. LAUGH ADAM	12	4	4.2	25
AUSDubb (AUS) 13th May	R5 1600m CL3	13/05/13 14:58	11. JUDGE SMAILS	12	20	24	5
AUSDubb (AUS) 13th May	R5 1600m CL3	13/05/13 14:58	1. AKAMAS	12	3.1	3.15	32.2580645161

Here is an example from a few days ago. hopefully it formats ok. The odds are the 2nd and 3rd last columns. The percentage is the last column which is 100 divided by the bet price offered (lower odds).

In this case they added up to 89.859 so it was a race to bet. The result was AKAMAS winning returning 100 meaning a profit of approximately 10.14

Thanks Useless Better
I,ve worked through your example and I can follow it. I think it might be a bit frustrating looking at lots of races, but finding very few that qualify. Still it is worthwhile if it is longterm profitable.

Clive,

Yes it would be frustrating. You could make your cut off higher if needed. But if its automated then its of no concern looking through it ... it just needs to be coded and then watch the dollars roll in.

Sadly, I'm not that computer literate and need to do everything manually, although I do have a dutching calculator