Very serious question

[partypooper]

OK, so I have been hovering around the "nearly made it" crowd for several years now, some of my ideas showing a small profit, others a small loss, so basically overall about break even.

But I really do believe this time I have discovered an "edge" its only small BUT thats all you need. VERY consistent both in real time and retrospect.

My question is (not to gloat as I appreciate that it could change at any minute)

to those that have already discovered an edge, how do I play it? obviously I'm not going to disclose the details publicly, but once b4 i hit on a winning place plan, only to see it disintegrate when I increased the stakes.

this one I have played out in real time with $200 bets (Place only, BUT lots of bets) and it has remained at the acceptable profit margin, what are your thoughts?

PS the same method breaks even for the win after many thousands of bets.

[wesmip1]

partypooper,

Firstly nice work.

What you want to do before you start upping your bets on this is check it mathematically to confirm you have an edge. Have you done this ?

If not then I am sure you would have kept records as every professional punter does. :)

Just place here the number of selections, the number of winners and the sum of 1/price. Reading your post it seems your betting place so it would be 1/(place price).

From this I can give you my thoughts on if the system is ready for betting.

The reason most systems go bad when you bet them is yoy probably haven't checked them mathematically over a series of live bets.

If you don't want to post the info here you know my email.

[Shaun]

I will say if it makes a profit then that is fine, one thing with place bets is people try to use all ups and such, i agree you get bigger divis but do you miss out on profits each day.

I always say percentage of bank is best, how you work out when you use the percentage of bank is up to you, daily, weekly ,monthly or maybe when the bank has increased by percentage, the only thing i say is don't use it per bet as the swings could turn a winning system in to a losing one.

If you could average 2% profit on the bank each day you could double the bank in just over 1 month.

[partypooper]

Wes,..... will be in touch.

Shaun, I take your point about doubling the bank, but I've never thought along those lines, my mind is set on POT........ LONG,LONG term.

I got myself a long way in front using % of bank on one particular idea ( and pulled out whilst still way ahead) this has financed many other ideas both useless and one or two with merit.

The idea that I'm referring to here is showing consistently 2-3% POT with only 1 filter,.... and I'm talking about a very long trial in REAL time.

With 2 more filters, i.e. 8-14 runners only, and LS 1,2,3,4,

this almost doubles to 5-6% POT (betting place only) with more action than anyone wants really!

[Shaun]

Not saying you need to aim for this POT is fine i was just saying what it could take to double the bank, a lot of people would say 2% was crap money but fail to see what compounding can do.

[wesmip1]

partypooper,

Haven't seen an email from you so not sure if you sent one through and I didn't get it.

[partypooper]

Wes, no; got bogged down with other stuff, I probably won't get to it
til Sunday

[partypooper]

W4es, seem to have misplaced your ee male addy,

anyway,

those stats stand as of today

Total bets (for the place) 3080
Hits 2200 (71.43%)

Returns: 3170.40

POT 2.94%

Ave. Divi = $1.42c

The thing that strikes me this time is that it is so consistant with all the stats hardly wavering reagrdless of any particular day (or weeks) betting.

All these are REAL TIME BETS @$200 level stakes.

But I have retrospective stats going back a fair way and show almost identical results.

[wesmip1]

partypooper,

I can't do proper analysis unless you have the stats for all the prices. I can give you a best guess but its not really good as avgs are not good to work with.

I really need the sum of 1/odds to do this properly


But going quickly on the avg odds you are roughly expecting 2169 winners out of your 3080 bets. Your actual winners were 2200 which is only 31 more winners then expected.

The real figures might give a different story though. But on what I am seeing you are currently just in front and tht could be due to luck or an edge. In fact I did this quickly so calcs might be wrong but I calculate a 75% chance the result is due to luck.

Based on this I wold suggest the wait and see more results before betting. There is only a 25% chance this is due to an edge based on the figures you have provided. If you can give more figures I can check those for you.

The more history the less chance will have an effect.

[Dale]

Wes,

Not understanding the maths involved i was wondering if those findings are in any way objective?

I mean i dont dispute the formulas just the suggested outcomes proportioning a percentage to either chance or edge.

From your experience have you seen a 75% chance swing to a strong edge% or vice versa,does it really come down to the more history the more likely a percentage favoring edge?

[partypooper]

Mmm, interesting, I can provide the extra info though but it would be the win odds 1 minute b4 the off.

But alas I don't think it would do us any good as I am not taking prices but betting at best tote?

Actually no I can't , I've just realised that I've only got the prices of those that placed, I can do it but it will take some time.

But, funnily enough this plan evolved out of another that I was working for some time, here I only bet if the place odds available were 1/4 the opening price 1 minute b4 the off (bet fair) this showed a 5% LOT @ BF (no commission) but just broke even if the same criteria was used but the bet placed @ Best Tote

Anyway thanks for the feedback, its pretty scary 3% POT, but over that no. of bets I thought I was on the right track (no PUN intended, hee hee!)

[The Ocho]

W4es, seem to have misplaced your ee male addy,

anyway,

those stats stand as of today

Total bets (for the place) 3080
Hits 2200 (71.43%)

Returns: 3170.40

POT 2.94%

Ave. Divi = $1.42c

The thing that strikes me this time is that it is so consistant with all the stats hardly wavering reagrdless of any particular day (or weeks) betting.

All these are REAL TIME BETS @$200 level stakes.

But I have retrospective stats going back a fair way and show almost identical results.

Hi Partypooper,

Let me get this straight, you are betting $200 at level stakes and have had 3,080 bets and have won $90.40 after all those bets. So one extra lost bet today would have you -$109.60.

I'm definitely no hexpert (believe you me), and I don't mean no disrespect, but that doesn't seem like a good way to be "investing" your $200 a pop.

You're almost better off going all up for a few wins in a row. My parlay place betting with my bot didn't exactly work out as I was betting on any and every fav but if you have a good strike rate you could parlay your winnings for 3, 4 or 5 races in a row then start over. I reckon that $90 could be a lot higher.

In my testing $5 level stakes worked out to be something like +$50 but 5 x all up (parlay bets) had the total around $500 (for $5 bets for crying out loud) :)

Maybe go over your results and see how many times you would of had a few winners in a row. And, if you are doing it right, you should only lose one set stake if you get a loser before the last bet in the series of say 5 bets. Just a suggestion.

[partypooper]

G'day Ocho, not quite that bad , a bit too long in the tooth for that performance.

the returns shown are to $1 stakes, for ease of understanding, I happen to bet $200 on each selection which means a total investment of $616,000 for a return of $634,080.

The stakes are irrelevant in the sense that its still only 2.94% profit on turnover.

I mentioned that my stakes are $200 on each just to emphasise that at that level it does not seem to affect the divi (as far as I know)

I appreciate that some professional punters would be betting 10 or even 20 times that, which I'm sure would alter the divvies and the profitability no doubt.

I was betting a % of bank quite aggressively a few years ago and observed a profitable plan go into decline as the dividends were affected. When I reverted back to $100 (my bets at that time) level stakes the plan again turned profitable, and is still going strong now (6 years+ on)

One further point, at $200 stakes and up to 60 bets a week, I have NEVER used 1c of my own money ever!! thats a bit cryptic but a few here will know what I mean.

[The Ocho]

Well, now you're talking :oops:. I mean, I was gonna say...lol.

Sorry about that chief!

You're obviously doing something right. Have you ever looked at your stats for going all up for a few bets in a row? The problem you would have then may be the same as when you were raising your stakes to the percentage of bank which means it may start affecting the divvie when your 5th bet is for $2000 or something.

I'm not sure what you mean by not using your own money (maybe using sign up bonuses or something?). I started off playing online poker with $12 in about 2006 and haven't put in any money since. I pulled some out a couple of years ago and put it into Betfair and, again, haven't put my own money in. Unfortunately it is starting to run low now though :(

Keep it up partypooper :)

[partypooper]

G'day Ocho, tried the lot mate every staking plan you can think of, the end result is level stakes is the ONLY way. Or put it another way, staking will NEVER turn a losing plan into a winning one (long term)

with a staking plan all you're actually doing is increasing stakes, so the POT stays the same over time.

Having said that the % of bank, is probably acceptable, but say, 1% never decreasing.

What I did was much more aggresive,, trying to makle hay out of the long winning runs, betting place only, but as I said it backfired.

[wesmip1]

Wes,

Not understanding the maths involved i was wondering if those findings are in any way objective?

I mean i dont dispute the formulas just the suggested outcomes proportioning a percentage to either chance or edge.

From your experience have you seen a 75% chance swing to a strong edge% or vice versa,does it really come down to the more history the more likely a percentage favoring edge?
Well they are as objective as you can get in this analysis. They are based on the odds obtained and odds are subjective. But as far the analysis it is usually a good guide on when to start betting. I only bet systems that have less then half a percent due to luck (and are not backfitted ... ie tested over real bets).

Its all to do with probability and some more advanced topics on it. I beleive using the analysis I do is the correct way to know when to bet a system.

The answer to your last thing is if you do have an edge the more bets you have the lower the % chance plays a role. So you would normally see a system sit at 100% chance to start with and slowly move away to be pure edge. As I said half a percent chance is the right time to start betting. before that you should collect more data.

Hope that helps. If you ever want a system analysed make sure you keep your stats for me (every odd on every horse win or lose).

[wesmip1]

One further point, at $200 stakes and up to 60 bets a week, I have NEVER used 1c of my own money ever!! thats a bit cryptic but a few here will know what I mean.
I'll assume arbitrage betting or bonuses.

There is also another way to get back a % of your money. It not well known and I don't want to divulge it here but it is possible to get back a percentage of what you bet regardless of if it wins or loses.

[partypooper]

thanks again Wes, I am devil with the cryptics arn't I?

Like Mark, the biggest kick I get is with arbing, even though the profit might be tiny. It started in the UK many years ago with a 3 horse race, I backed the 2nd fav, with a saver on the rank outsider, booger me if they didn't put a fortune on the 2nd fav, which duly shortened dramatically and I suddenly realised that I could now back what was the original fav as well and win no matter what, not true arbing in that case but ahhhhhh ! it's a great feeling!

Bonuses, well I never knock em back, but as the S/R is somewhere between 75-80% depending, I always hope the free bets are losers if you see what I mean?

And as for those "other %" sometimes I want to sing from the rooftops, BUT, its a door that could close very quickly, so mums the word!

[stugots]

wesmip1 would appreciate your analysis of the following set of figures - they represent the results of a set of self generated ratings with no filters applied & cover most races from maidens to the mc.

they are currently returning a pot of 1.56% and with the addition of a few simple filters show an acceptable profit & i have been betting live for a few months now.

I guess what i would find interesting is if your analysis puts the profit so far from underlying basis for the selections (my ratings) down to luck or not.


number of selections - 6130
number of winners - 1009
sum of 1/price (of all selections) - $962.41

[Dale]

Well they are as objective as you can get in this analysis. They are based on the odds obtained and odds are subjective. But as far the analysis it is usually a good guide on when to start betting. I only bet systems that have less then half a percent due to luck (and are not backfitted ... ie tested over real bets).

Its all to do with probability and some more advanced topics on it. I beleive using the analysis I do is the correct way to know when to bet a system.

The answer to your last thing is if you do have an edge the more bets you have the lower the % chance plays a role. So you would normally see a system sit at 100% chance to start with and slowly move away to be pure edge. As I said half a percent chance is the right time to start betting. before that you should collect more data.

Hope that helps. If you ever want a system analysed make sure you keep your stats for me (every odd on every horse win or lose).


My concern was someone taking a small sample of what later turned out to be a winning method would show it as a high chance probabilty,dont doubt the usefulness of it just the effect an early application of it could have on the confidence of the user.

Like everything,the more data the better.

[wesmip1]

wesmip1 would appreciate your analysis of the following set of figures - they represent the results of a set of self generated ratings with no filters applied & cover most races from maidens to the mc.

they are currently returning a pot of 1.56% and with the addition of a few simple filters show an acceptable profit & i have been betting live for a few months now.

I guess what i would find interesting is if your analysis puts the profit so far from underlying basis for the selections (my ratings) down to luck or not.


number of selections - 6130
number of winners - 1009
sum of 1/price (of all selections) - $962.41

Stugots,

I ran those numbers and came up with a 10.1906% chance that the result is due to luck. This is an almost 90% chance that the result is due to an edge. These numbers are only relevant if the numbers you have provided are on real bets and not on a back fitted system. Back fitting will always tend to give a lower result.

Do you mind if I ask if this is betfair or tote prices ? If they are betfair prices have you taken into account commission at your level (2%-5%).

Another 6000 bets and a similar 1/price and actual winners would reduce it to a 2% chance that luck is the reason you are showing a profit.

Another way to analyse these figures is to check the standard deviation of the results lumped into 100 bets. You have enough selections there. You would have 60 data points and this should produce a nice distriubution. This is very useful in analysing your risk. If its not showing a nice bell curve (or close to one) then I would be worried and not betting the selections.

For others who do not have 6000 bets I would suggest using 20 or 50 bets. This can easily be plotted and you should see a bell curve as a result. You will be able to see your edge pretty quickly doing this as the peak of the curve should be to the right of 0.

Hope that gives a little bit of insight on one way to anaalyse a system.

There are another few tests I would consider still but I need access to your data to do these and thats not possible. But they include a number of advanced tests.

I think the information above on plotting the distribution is pure gold. How many people actuall start doing that is probably 1% of the people who read it. If you are the 1% then i wish you luck, by doing this you will learn more about your system/rating, probability and how to be a winner then you will ever imagine.

[wesmip1]

My concern was someone taking a small sample of what later turned out to be a winning method would show it as a high chance probabilty,dont doubt the usefulness of it just the effect an early application of it could have on the confidence of the user.

Like everything,the more data the better.
Dale,

I see what your saying and fully agree. Its an analysis tool and it requires data to be useful. Without enough data it will always bias towards chance playing a larger role then it should. Also if you input biased data into the calculation you end up with it leaning towards having an edge when realistically its the biased data influencing the results.

As always bad data in means bad data out.

I would query anyone on why they are betting (professionally) without enough selections for a proper analysis though. This is a mistake 99% of punters make.

[stugots]

thanks for your advice wesmip1, re your points raised -

- the figures provided are not backfitted, just the results of my of ratings. I have backed about 60% of them as I have use the raw ratings to create a couple of systems.

- the prices are betfair & applying commission @ 5% on all selections changes the sum figure to 1002.59.

- will have a go at the bell curve & upload a spreadsheet

just a question - at what point would you consider taking a selection method live based around the luck/edge percentages?

cheers

[wesmip1]

just a question - at what point would you consider taking a selection method live based around the luck/edge percentages?


I usually wait for half a percent chance due to luck. This means that there is 1 in 200 chance that the system/rating is performing due to luck alone.

For instance yours was at 10% ... So there is still a 1 in 10 chance the whole thing could come crashing down in the next 6000 bets. Thats too high for me but might be acceptable to you.

[stugots]

attached a spreadsheet, had a go at the bell curve using the guide at - http://support.microsoft.com/kb/213930

interested in any thoughts

[stugots]

increased the random number sample & used the correct chart with this one

[wesmip1]

Looking at the chart not sure what you are plotting.

I will assume the blue line is the random sample (perfect bell curve).

What are the purple and green lines ?

[wesmip1]

I wasn't sure what you were plotting so a bit hard to work out. I normally plot the return as a percentage over X bets. So over 100 bets if it returns 110 then that is ploted as +0.1, and if it returned 90 that is plotted as -0.1

In th csae that you have done If I assume yours is the green line then it looks like a good curve but I can't make any other comments without knowing the type of data you plotted.

[stugots]

what i plotted was the sum of 1/price in 100 bet blocks

the attached chart is based on net returns on 100 bet block, seems now the peak is fractionally to the left of centre.

[stugots]

try to get it right one last time

[wesmip1]

I redid the chart as something I could understand but used your figures.


I like plotting the distribution as it makes it easy to see the issues. The way I plotted your data Cols G and H as a scatter plot with smooth lines connecting them .

Looking at it I am slightly worried in the distribution on the losing side. It looks like it needs a few more data points to smooth that out. The problem you have at the moment is the tail on the positive side is much higher then the negative side. I would question why this is the case and what I find is this occurs with backfitted systems (not saying yours is). You will at some point have the same losses on the negative side and these are seriously going to dent any of your profits.

At this stage with the 10% chance due to luck and the distribution I would put this into a 3 month watch period. I might even bet it with small stakes but I wouldn't be betting a large amount on the selections just yet.

This isn't probably what you wanted to hear but it might save you a few dollars in the long run. More data will definitely help though.

[stugots]

thanks for the observations, not surprised that at this point there is a likelihood of a fair variation on results achieved to date as those results are from a raw set of ratings.

as long as it can continue at around the breakeven mark, the addition of a few simple filters has to date proved profitable, as always time will tell

[wesmip1]

stugots,

At least you have some idea of what to look at now. If you ever want the selections with filters applied checked just let me know.

Stix,

I am looking into your stuff tonight and tommorrow so expect an email reply by Saturday.

[stugots]

I like plotting the distribution as it makes it easy to see the issues. The way I plotted your data Cols G and H as a scatter plot with smooth lines connecting them .

Looking at it I am slightly worried in the distribution on the losing side. It looks like it needs a few more data points to smooth that out. The problem you have at the moment is the tail on the positive side is much higher then the negative side. I would question why this is the case and what I find is this occurs with backfitted systems (not saying yours is). You will at some point have the same losses on the negative side and these are seriously going to dent any of your profits.

wesmip1 if i could pick your brain - do you put more emphasis/importance on standard deviation/distributions than the chi-test?

for example if a chi-test shows <1% due to luck but the bell chart is marginally biased to the negative side, which would you pay most attention to or do you consider them equally important?

[wesmip1]

stugots,

I would say equally important and if either one is showing something negative I am wary of the system for betting. You want everything to line up right if your going to bet a system (and even then it can fall over in a heap some times).

If I had to choose one I would choose the distribution as I find it gives a better overall picture and it not as easy distorted by backfitting. If you plot a back fitted system the distribution will usually show it.