AD's daily selections

[gerry]

Well that was interesting out of the 5 races i picked the only one that let me down yet had the second at big price was the race over further than mile as i said earlier the other four races had the winner in there i had three f/casts also and you had tricast in one of the races also AustinDillon75 that was the longer race to.
i had good profit day of 10pts and slow steps but right ones i think will be interesting how your day went all in all to.
But i think those longer races and small fields really do have poor impact on results.

 

[AustinDillon75]

Not a good day there, 10 winners from 42 (expected 11.61) and a loss of 14.58 to BSP. So the profit for the week since Monday is back down to 8.47 pts, and a return of 4.1%. As per usual, fluctuation but it is feeling a bit like we are turning corners then giving it back. That's fine.

Regarding Hambelton, it seems like he got butchered early by the eventual winner which came home at a BSP of 48.

Some of our value picks ran very creditably.

Resemblance 2:20 Haydock ran 2nd at 25/1
Pershaada 2:32 Goodwood won at 7/1
Naqeeb 4:17 Goodwood ran 2nd at 14/1
Kings Crown 7:22 Pontefract ran 3rd at 12/1

But overall, not the best. Picking up the pieces for tomorrow.

 

[AustinDillon75]

Well that was interesting out of the 5 races i picked the only one that let me down yet had the second at big price was the race over further than mile as i said earlier the other four races had the winner in there i had three f/casts also and you had tricast in one of the races also AustinDillon75 that was the longer race to.
i had good profit day of 10pts and slow steps but right ones i think will be interesting how your day went all in all to.
But i think those longer races and small fields really do have poor impact on results.

Our posts have just crossed but I did just look and saw you had 4 of 5 above and in the race you missed out on, I actually had the first four home in the wrong order!

It all needs bigger sample sizes I think before too many conclusions are to be drawn.

 

[gerry]

Anything that shows not loosing is good and great way to just slowly change what's needed here was my thoughts other night and might sound bizarre to some but it was my thinking after i had few poor months i had to look at it more safer and relaxed approach as we get older retired here its more hobby so has to be some kind of fun to.
So i put a hundred in betfair and thought right do what i have been doing pick certain races like mile or under and at least 8 runners thats my new solid guide and if 8 to 14 runners pick 3 horses and over 14 pick four horses if needed just eliminate one if not getting profit from all horses bet no matter how small as a win is a win.
Then my mind got thinking stick to the 1pt bet a horse for the full month even if i am down 60 thats 2 pts a day every day for the month i will have learnt a lot of what's going on so put other hundred in betfair would give me now 140 do same for next month but bets now of 1.4 pts a horse see how that goes if had to loss 60 again as i am betting more points a horse that would actually have been small improvement from last month as bet more per race then i would have 80 add other hundred and now got 180 and bet next month 1.8 per horse so building trust in system and hopefully get the loss of 60 a month down say lost 40 add other hundred would now have 240 so only adding 1hundred a month but getting system better building bank and hopefully now not loosing per month and using that would feel i am actually doing better than i have lately and now got system where turning small profit every month but bank going up stop putting money in finally and betting same system sounds weird i know but i am hoping that's worse case scenario.
Getting away from multi betting and just solid dutching in races i know well, always betting for month at 1 percent of starting month bank then rescheduling it every month as i would be amazed if losses are what i planned for at start but its more a feel good not worry strategy at start which we do a lot just like you feel a lot of the time when you think its going know where yet your up just i think work out the races you need keep and get rid of ones that don't suit your thinking.
If i start getting months not loosing and then turning any kind profit i would be happy you only need look at the finishes of a lot of the races so tight that racing is tough so dutching is giving you that safe guard of the dangers coming to beat it, in fact some or most time i am now shouting on the danger to come win as it usually slightly better odds.

 

[AustinDillon75]

Today's ratings and full cards for York and The Curragh. I've now added the racecard number and number of runners to the tables too and I have looked at field size results - more soon, but quite revealing stuff emerging.

 

[Outlander]

Today's ratings and full cards for York and The Curragh. I've now added the racecard number and number of runners to the tables too and I have looked at field size results - more soon, but quite revealing stuff emerging.

Just an observation , no data to back it up, was interested in your value picks shaded yellowish and I noted you mentioned sometimes the model picks up the 3rd rated horses in 3 runner races and this not ideal, what about a rule somewhere along the lines of there must be at least 2 horses of a lower forecast price that are lower in the ratings than the value pick.
This would rule out horses in the bottom 2 of the ratings and also horses that don’t hold an advantage over fancied runners according to you ratings. Simple rule but might weed out undesirable value picks ?

 

[AustinDillon75]

Just an observation , no data to back it up, was interested in your value picks shaded yellowish and I noted you mentioned sometimes the model picks up the 3rd rated horses in 3 runner races and this not ideal, what about a rule somewhere along the lines of there must be at least 2 horses of a lower forecast price that are lower in the ratings than the value pick.
This would rule out horses in the bottom 2 of the ratings and also horses that don’t hold an advantage over fancied runners according to you ratings. Simple rule but might weed out undesirable value picks ?

You are right, the script still needs that revision and I am yet to do it. The principle is that for it to count it must be in the each way placings (so the top rated if 4 or less, in the top 2 for 5-7, top 3 for 8-15, top 4 for 16+.

What I am not doing is claiming anything on here to be a value pick if it's 3rd in a race of any less than 8 runners recognising the output still flags it. I think I mentioned it a page or two back.

For avoidance of doubt here are the non-top rated value selections today. I am looking at potentially different metrics for what comprises value but below.

 

[gerry]

hopefully learning something else today about amount of horses in race i bet 12 races today only one race over a mile again and that lost but i also bet four races where there was more than 20 runners and they all let me down to so looks like to many runners can be as much damage as to few runners for wide draws likely the reason to spread out so that was 5 races to many runners or to far then out of the other 7 only 1 let me down so at end of day had 50% winners and a lose of 7 pts so nearly back to square one but that's what saturdays do in tough races so now i just need really stop the longer races and cut out with horses over 14 runners as new addition and be confident that should cut out all bad luck races so we now have must be at least 8 runners no more than 14 and no further than mile.
Only thing is not sure what you can classify as to long or short over jumps next season maybe just stick to 8 to 14 runners and being hurdles makes sense to but i enjoy my chases so cant completely ruin the fun side but i am away ahead of all summer so plenty time to think about it.

 

[AustinDillon75]

Results and total since 27 April. I've set the results sheet up now so I can just produce this each day moving forward.

The script is going to start taking a closer look at the larger field races because thanks to gerry I decided to look at them and in races of 10 runners of less, it delivers 103pts from 770 races but conversely we lose around 36pts from 11 runners or more. I don't think it's a case of just binning those races off - the model looks at pace and ELO in isolation and is not properly looking at either factor relative to the rest of the field so is a point that now needs addressing.

It's also the fourth Saturday in a row where we are hitting good strike rates etc but the wider public are probably proving wise to the goings on in a way that they may not be on a wet Tuesday afternoon at Hexham. So we are going to ask the model to pay much closer attention to how the racing world functions on that day of the week. We've lost 32pts on Saturdays in the above so far.

 

[AustinDillon75]

Today's top picks and a card for the Curragh.

 

[gerry]

These are the 6 races i have looked into today with no uk flat races on.
3 20 CURRAGH 8FUR 12 RUNNERS. 7-8-4.
3 28 UTTOXETER 2M 4FUR 11RUNNERS. 4-3-5 HURDLE.
5 12 UTTOXETER 3M 8 RUNNERS. 5-4-8 CHASE.
3 37 KELSO 2M 4 FUR 12 RUNNERS. 2-6-1 HURDLE.
5 22 KELSO 2M 9 RUNNERS. 1-2-3 HURDLE.
2 35 FONTWELL 2 M 8 RUNNERS. 4-6-1 HURDLE.

Not sure what they will through up but a stable small selection of races and horses to keep it well in control will be interested how it adds up.

 

[AustinDillon75]

28 runners, 7 wins and a further hammering.

 

[gerry]

4 nice winners among mine gave me a 12pt win today i see the chase let me down as i said hurdles likely better as every little detail why your horse can get beat even with best form .
So far we have small fields slow pace, large fields unluck in running and also poor draw, also chases with falling possibility or even bad error and hurdles to maybe lesser extent but still jumping hurdles, i am not keen on all weather to as a lot of bunching up in bend, and obviously the races where no form worry me as just an other guess to make on little form. Novice form also not as reliable so again were thinking is that improving or not. All in all looks like solid handicaps with minimum 8 to max 12 or 14 horses is the sweet spot for my dutching but still in progress is having three selections right well i think it likely is as usually as they go 123 in selection more times than not there getting bigger in price to and cutting out 3 could kill the better odds thats why possibly not saying is but why your feeling your prices are not big yet AustinDillon75 but i think you have had some nice winner odds there and even in Chesham bowers thread he mentioned maybe not getting horses over 8/1 was a bowers point so let your profit loss tell you how its doing and don't chase bigger odds when i looked my recent winners they where spread over all three selections 1 2 and 3 so i would have likely done better leaving out shortest odds of the three but it can change so i am sticking to my plan i have set for 3 months building it slow add funds to get it up to reasonable amount per horse bet and learn along way if its working.

 

[gerry]

Have you though of how it would do AustinDillon75 if you where betting selections with money back offer on bet365 you would have 25% winners today and 43% placed so money back in them ok shorter odds but only had 33% loss. Even first two places if dont want to cut to short you had 25% won and 25% second money back.

 

[AustinDillon75]

gerry I've designed it to go via the exchanges because over time I think it will be the most profitable and if I use the likes of Bet365 anything that works will probably see me restricted in the end. I'd argue that BSP is really the true genuine measure of viability. Perhaps if it comes to a clear and genuine point of profitability

I was hoping to instigate a real £200 bank / 1% stake on 1 June but it is increasingly something I'm questioning as the profit seems to just go up and then be taken away again. I've run model v10 this morning having used v11 yesterday and it removed one of the winners, and v11 flopped in one of the four horse races from one well fancied runner that didn't finish to the other well fancied runner that didn't finish. v9 also gave largely the same selections (2 different and both lost anyway).

So I am just going to put it down to variance, stick with v11 for today, and see how we go. v11 is the one that supposedly, understands the larger fields and considers all the differences between Saturday racing and the remainder of the week.