Field size

[DR RON]

A general question to all. Whatever your method of selection, be it ratings, system or throwing a dart, at what field size does your particular method start to reduce the ROI. Generally speaking as fields get bigger, strike rates decrease but are offset by larger average dividends, however there must be a point at which the actual return on investment as a percentage starts to significantly reduce.

[moeee]

Well if you checked my tips,you would see that mine would be a field size of 2 or more!

[KennyVictor]

I reckon the question should be "below what field size do you start to reduce ROI". Don't know if it's the rounding effect or what but small fields (less than 5 certainly) don't work for me with quinellas which is what I'm betting at the moment.


KV

[Bhagwan]

What I have found is the profit starts to kick in with fields of 9-15 runners .
Anything less & one will find that the selection SR may be the same but the value just aint there to show a profit.

If one wanted to test this , target all races with exactly 8 runners .
One will be supprised to see that no matter what one tries to do to filter the edge , the race will be won by really short payers or really long payers, frustrating stuff.
One will also notice is that the real inform horses will often not perform up to expectations.
There could be a number of reasons for this
e.g. the race could possible have an uneven pace to it ,which allows anything to swamp the field from the rear.

One may devise a system for these 8 starter races , but one will see that it changes dramaticly month to month & thats not really what one needs if one is trying to target some consistancy into their punting strategy.

Cheers.

[DR RON]

Thanks for the replys guys. I spent a lot of time today thinking about what the answer would be and I guesses that as the field size got bigger in most cases the pot would decrease. This then leads me to ask the question at what field size would return you the most $ in actual profit rather than pot. Obviously you may have a method that produces50% pot in fields of 4 but only 2 % in fields of 16, the 50% looks better on paper, however if thhere is only 5 bets for a year as opposed to say 300 for fields larger fields then the actual $ amount you win may be lower with the reduced turnover, so at the end of the day is it POT or actual $ that matter more, and in what size field does your actual $ return reach its peak. I hope you guys understand what I am trying to get at. ITs been a long day and I'm very tired.

p.s I cant believe some of the times that people post at, I've noticed a few from around the 3 am mark recently???

[Bhagwan]

One will find that it varies month to month because there is ususally one big payer that changes the average for that field size group.

If one wants a general figure for field size, the answer is 10 ,because there are more of them, than say 11 only or 15 only.

One will note that there will be some months that show a loss , then a month with some strong payers , it becomes a patience sort of thing.

Cheers

[DR RON]

Thanks for the input Bhagwan, I guess i was hoping for those who keep a record of every bet they make in a spreadsheet would be able to go through their records and say , my best actual profit comes from races with x number of runners. Maybe there arn't too many people who keep such detailed records of their bets.?? I would assume that people who don t keep records of every bet dont really want to knowthe truth about their betting? And then when asked whether they win or lose can honestly say they dont know.

[KennyVictor]

Dr Ron,

I tried to post a table of results of number of runners against ROI but it looked so crap in the preview I gave up. Someone told us the tag to stop the silly software reproportioning your text a few days back (could have been jfc but I'm not sure) but I can't find it. If they could tell us again I'll try to post the table.

KV

[KennyVictor]

Hi Dr Ron,

OK Found it.

These figures aren't actual bets but returns by my ratings system over past years. They are percentage returns betting my best rated horse to win with various field sizes on races since 1994. Strike rate is about 29% if it is of any use to you.






<PRE>Runners <=7 8 9 10 11 12 13 14 15 16 <=17
WA 96.2 108.3 92.7 104.6 104.2 105.0 91.0 104.0 116.3 94.9 -
NSW 101.8 100.3 99.7 101.1 106.8 105.1 105.6 107.1 97.7 100.3 65.3
QLD 97.1 98.0 98.4 105.0 110.8 98.4 90.6 97.5 113.2 98.0 114.3
</PRE>
</PRE>This includes about 8000 races overall for WA and 32000 for NSW so it's quite a significant sample. I can't see any great correllation between field size and success rate but I have to say I didn't expect to. I would expect less wins but at a correspondingly higher price in larger fields.

KV

[joelance]

Hi Kenny

Only by anecdotal evidence...as i dont break them up into various elements to analyse..(i just use three columns...output..returns..profit.)....in my experience the better class races especially melbourne and sydney offer the greatest profits. In the last 10 Melbourne cups, my ratings have had 8 winners on top and they are definately big fields. Do you by any chance break down your figures by class?

[KennyVictor]

Hi Joelance,

Not by class but I do break them into 5 prize money bands which is a similar thing I suppose. Not really any significant results there. NSW I do a bit better in the top 5% by prize money but worse in QLD. Victoria the second to highest band is best. Figures pretty much all over the place really.
KV

[DR RON]

Thanks for posting those figures Kenny, just going on those raw figures seems like the 15 starters produced your best return overall, I would have guessed that the 8- 14 fields would have the best returns. You mentioned your strike rate was about 29%, are you able to work out the strike rate just for the 15 runner fields? As you say it would probably be lower but with a higher average dividend. Are your bets usually top 2 or 3 in favouritism or do they vary quite a bit?

[KennyVictor]

Don't know how significant the figures are around the 15 races mark, as you can see below the number of 15 horse races is a fair bit less than the average for all fields.

Strike rates:
Qld 24.1% on 15 runners against 28.8% overall. (430 races out of 20,000)
WA 24.8% against 27.4% (455 races out of 8500)
Nsw 22.1% against 29.1%, (440 races out of 32,000)

I honestly don't think the 15 races being best is significant, more chance. I just ran Victoria and it returns 95.7% on 15 runners, strike rate 19.10% against 26.4% (730 races out of 25000).

Yes generally the selections are in the top three of favoritism, I have had a $20 winner or two but they are red letter days indeed.
By the way, if I bet winners I bet to a price and find that this cuts the bets by about 50%, cuts the strike rate by 5 to 8% and boosts the ROI by 7.5 to 11% (depending on the state). Doesn't give any more meaningful figures by field size though.

KV

[DR RON]

Thanks for that KV, I guess it still comes down to getting value for your selections or at the very least a minimum acceptable price for your selections.

[crash]

Hi Kenny,

How have you found ROO for various field sizes? I tend to stick to field sizes of 8 to 14 for this reason. The odds might balance out the books but ROO potential on bigger fields of 15 and up always worry me. It's a psychological thing obviously as far as your figures go, but 'feeling' uncomfortable psychologically does make me think it is having an impact on my selection process [handicapping] and my bet amounts.

[KennyVictor]

You had me scratching my head for a minute there, ROO, we're going to need an acronym glossary soon.
I don't isolate races by field size when I bet so it never really occurs to me what the run of outs is on any particular field size. Maybe I'm having a longer ROO on those races but hopefully a 5 or 6 horse race will come up in the mean time and ease the pain.
If you run the strike rates through a ROO predictor (23%ish against 29%ish) that would give the likely run of outs on those races compared to the overall figure.
Funny, I just ran NSW figures again and I get a strike rate of 35.6% in 6 horse races (2,135 of them) and 25.4% in 12 horse races (3500 of them). Now there's one for the statistics guys. Why don't you get a strike rate twice as good with half as many horses. And how can we use this to our advantage? 14 Horse races 23.7% against 7 horse races 32.9% - same question.

KV

[crash]

Good SR averages Kenny. Unless you are backing fairly short priced runners, your making reasonable doh-ray-me $$$

I think the reason for your smaller than [mathematically] expected ROO prediction for larger races based on your SR in 6 and 12 horse fields, is simply due to the fact that larger fields are loaded with more rubbish horses.

What mainly counts is the 3 or 4 real chances [on average] that make up 80% winning chance in a 12 horse race, compared to say 2 to 3 real chances making up 80% of winning chance in a 6 horse race. The number of winning chances are not halved in a 6 horse race compared to a 12 horse race.

If all horses had the same chance in both field sizes, you would find a bigger ROO potential for a 12 horse field and would get a smaller SR. Considering the real winning chances in an average 12 horse field, your SR of 24.9% should actually be better. The reason it isn't is that more possible unforeseen events happen [having to run wider, bumping and less opportunities to come through from behind etc.] in larger races. So in fact the SR% in larger races look OK, but my seasoning and explanations above suggest they are not as good value as they look on paper and point too larger races actually being poorer value if you follow my drift[?].

9 to 12 runners is my prime field size pick for odds and confidence ratio. Above 14 runners is my cut off for perceived value. The SR's you have mention seem to support this logic I think, but in the scheme of things as far as your concerned, I'd just keep doing whatever you are doing.