win s/r

[shifty]

what would be the longest losing streak for a system with a win s/r of 8.4% please?

[The Ocho]

what would be the longest losing streak for a system with a win s/r of 8.4% please?
Check this thread out shifty.

http://www.propun.com.au/racing_forums/showthread.php?t=23205

[Rinconpaul]

80

[UselessBettor]

Its easy to work out:

You basically have a 100-8.4 = 91.6% chance of a loss.

For 1 loss the chance is 91.6%.
For 2 losses the chance is 0.916x0.916 = 0.916**2 = 83.9056
For 3 losses the chance is 0.916x0.916x0.916= 0.916**3 = 76.86
...

For n losses the chance is 0.916**n

Where ** = to the power of (available on all good calculators)

You then determine what you think is a longest losing streak. Is it 1% chance or smaller ?

you then substitute a losing streak in n and you get your result:

For 80 it is 0.916**80 = 0.09% chance. which is less then 1%.

[PaulD01]

what would be the longest losing streak for a system with a win s/r of 8.4% please?

Hi shifty,

On a sample of 100,000 trials the LLS for a SR of 8.4% is:

101 (69.56% chance that this will occur)

137 (4.93% chance that this will occur)

155 (1.04% chance this will occur).

The issue is the LLS is only one piece of the puzzle. More importantly is the maximum draw down on your betting bank which at the end of the day is the key piece of information that most punters are blissfully unaware of. There are lots of methods/systems that appear to work when one looks at the end results. The problem is though that more than likely, the method/system has a draw down peak that exceeds the punters betting capital.

My suggestion is that you place more emphasis in learning what the risk factors are than focusing purely on strike rate. You might want to take at a recent post I made that deals with the relationship between strike rate and its impact on bankroll management and profits. I have posted the link below.

http://forums.ozmium.com.au/showthread.php?t=27589

[shifty]

The problem is though that more than likely, the method/system has a draw down peak that exceeds the punters betting capital.



http://forums.ozmium.com.au/showthread.php?t=27589thanks all, tested over 12 years and more than 4,500 selections and not reliant on a single longshot, although the odds range is $10 to $40, pot is 39.8%. the thing that will reduce the odds (and pot) is the bet I have on the eventual winner.

[woof43]

thanks all, tested over 12 years and more than 4,500 selections and not reliant on a single longshot, although the odds range is $10 to $40, pot is 39.8%. the thing that will reduce the odds (and pot) is the bet I have on the eventual winner.
If your meaning payoffs reducing, as another poster related if winners strike randomly, but over the long term you can ascertain a given s/r. how about you randomised the payoffs based on your collected data for the payoff on your selection. Remember there are basically 3 ways of calculating your avg payoff and the one most commonly used by most, is not the best.

[Michal]

Hi Shifty,

To be completely honest, with the other information you provided I have grave concern about the success of your method. Don't get me wrong, I wish you all the best, however most punters totally ignore what I call the realities of betting when they square up to betting a long-shot system like yours.

In order for you to have any chance of success your betting must be automated. You have 375 bets a year from your sample, thats a lot of wasted time waiting to put the bets on as I also suspect that they are price filtered ?

In your sample I can guarantee you there will be a bank draw-down of longer then 367 bets. Meaning that you will, for a whole year be less then your previous bank high and possibly longer. (Run through your profit/loss analysis of your bets in time order.) Can you sustain that demoralising run and still bet each days selection never knowing when that magic run of rebound happens?

The Reality of betting is almost a law: Your real life betting will never pan out like your test.

I'll give you a few examples from my own life.

Running an automated long-shot system with selections each day I experienced the following: I had a run of outs of about 50, which was nothing unexpected, my internet went down and in that hour I was on the phone with support I missed 6 races, 5 of which were won by one of my long-shot selections. I missed out on over 70 units profit. I have never had an 'hour of power' like that before or since. Just a random event and of course it was followed by a further long run of outs.
Another time I was betting a system that required 2-3 bets per race, and wasen't automated. It too had a horid run the few days before. BUT it was fathers day and my family wanted to take me out , so we did go, and had a wonderful time the whole day. When I got home from the 23 qualified races on that Sunday I would have won 19 of them with the lowest win div of about $4.5.

Talk about character building! Those were obviously major ones, but your life is filled with many small events of similar nature. Many outside of your control; Cyber-horse (very reliable) for instance has now been down for 5 days including the weekend, that is a lot of form and races missed by anyone that uses their form. Regrettable, but that is life. Its not perfect and sterile like your test.

You would think that the punting gods were against us. But that is not the case, the Realities of betting simply states that you will miss bets and most likely winning bets at the most inopportune time if you have any life other then punting 24/7. Everyone has them! While each of these misses might not of themselves be the end of the system profit wise, they tend to be rather decimating to the mental stability of the punter. (I still look up to the gods and ask, 5 in one hour? REALLY? Are you kidding me? :))

This is why these methods are up for grabs, professional punters will not touch them because of the volatility and small future prospects due to liquidity , meaning you will never have a shot at making a living out of it. PaulD01 talks about that.

Ok that's enough of the negatives. While the pros don't touch them doesn't mean that you can't do it, make it, enjoy the experience and earn a nice little side profit or have it as just one part of your strategy. You can! Especially if you are aware of and are prepared for the pitfalls along the way and are supper consistent with your application.

All the best;

[darkydog2002]

Well said Michal.

[SpeedyBen]

Paul
Your point about maximum draw down being the main item is very important when working out bank size.
Unrelated to this topic, do you find that you have tracks where your ratings perform well below your norm?

[PaulD01]

Paul
Your point about maximum draw down being the main item is very important when working out bank size.
Unrelated to this topic, do you find that you have tracks where your ratings perform well below your norm?

Hi Speedy,

Our ratings perform consistently across all tracks and geographical areas. From a betting perspective, I have no issue betting on any track if the right opportunity presents itself and I have the capacity to bet as I would like.

[aussielongboat]

michal,
I can relate to your post.
6PM on NRL grand final day - October 6 2013
HK race 5
Working on some software issues i get distracted with something else and don't fully finish of my selections. - i had some problem where i had to run my program twice to get the HK selections.
anyhow because of the distraction i didn't do that.
"Lucky Fortune" is my selection and i was scheduled to have 420 on it.
it wins at avg dividend of about $90.

IMHO that is serious coin

the punting gods are so cruel some times.

the minute you take your eye off the ball they will swoop.

so i agree with what you say - the actual execution is not as easy as it looks on paper and in retrospect.

[Barny]

I have a system that shows similar results, and Michal, your post is interesting, but it doesn't reflect what's happened to my longshot system. I've never had the run of outs you suggest, no where near it actually. My research shows that for instance, even money pops, or 50 / 50 chances can put you out of business quicker than a system with a lower strike rate because you're punting the stats "knowing" that consistency wont allow a run of outs that will destroy your bank, but if you're punting these shorties, then your outlay in terms of % of your bank is much, much higher, so the stress after a few runs of outs is infinitely greater. In short, a run of outs on the shorties and you're gone. A run of outs of say 5 or 6 can leave you hanging by your goolies. How do you come back from that? Yet the stats tell you that the best LOT are the shorties, increasing until you get to 50's plus then the chances of a winner are like 500 / 1. I use my longshot system as one of four systems I punt concurrently and it's as consistent as any system I've come across. shifty, even tho' you're pleased I'm not around (I'll disappear again shortly), if you have a system that's stood the test of time, and has some logic behind it, then go for it.

[Michal]

Hi Aussielongboat,

Thanks for that; Yes, you will remember that date for a while I suspect, its hard to find winners let alone deal with all the problems and issues and setbacks like that especially if one is not in front is hard to deal with, and for many might just be the beginning of the end. The only way to deal with the punting gods is never give them a chance to pounce, easier said then done!!!

Hi Barny,
All systems are different in their dynamic and performance, however no one is above mathematical laws. I learned that the hard way, many times. It simply is much easier to live with them, accept them rather then ignore and hope they miss you. Everyone has to find their own level of risk vs reward ratio and work withing their comfort zone. You obviously found yours, well done!

Kind regards

[aussielongboat]

what would be the longest losing streak for a system with a win s/r of 8.4% please?

this is the excel formula to calculate it

=ROUND(LN(D3)/-LN((1-E3)),0)
where D3 holds the sample size and E3 is the probability.

so if you had 1000 bets your longest run of outs (ROOTS) could be as high as 79
for 5000= 97 and 10000= 105.
even in your first 500 you could expect ROOTS of 71 and in the first 100 - 52 ROOTS

so if your bank wasn't big enough you could be ROOTED by the ROOTS

[Rinconpaul]

.. do you find that you have tracks where your ratings perform well below your norm?

IMO, you make a good point here Speedy. I record the performance of the Open 1st & 2nd favourites for every race, as it's inherit in the system I use. I believe that tracks used rarely for TAB races like Clare, Hamilton, Warren...etc, have a poor performance with regard oncourse bookmakers Open favourites. But in saying that you can get Eagle Farm say on the 13th of Nov and not one 1st or 2nd Open favourite won!

Design of the course has got a lot to do with it I think. I can't remember which track it was from yesterday, Port Macquarie or Townsville, but the nature of the track design was that after rounding the home turn they were bunched well and then the straight was long and wide which gave the backmarkers a chance to find the lead?

There's definitely something to be said for the track design and local knowledge at indistinct tracks.

[PaulD01]

Hi RP and Speedy,

RP, you have hit the nail on the head re track design and track racing characteristics. In my opinion anyway, that isn't so much shown in the performance of the ratings for a given track but rather from picking the right horse in a race or letting a race go because the horses that you fancy are disadvantaged. Track specific calculations are taken into account by our ratings software as well in overcoming what you describe on a track performance level.

Speedy, one of the many reasons that most ratings have flat-spots at some tracks, can be explained by the form and calculation methods they use. We have really not written much on this forum regarding the data correction processes that we undertake on a daily basis, but it's fair to say that these routines comprise approximately 35% of all our programming code. This gives you some idea about the enormity of the task required to turn the so called "official" form into something that is accurate and for the most part free from the plethora of errors that otherwise go unnoticed. Below is a link to our Axis educational articles should you be interested that explains in part what we do with our data. It is explained in the article titled R2W Data:

http://www.ratings2win.com.au/learn-to-bet-like-a-pro/axis-educational-articles

We source amongst other things accurate times for every Australian, NZ and HK TAB race run and apply that to our main database. This is a huge and expensive task. But for anyone using the times and margins for that matter, track conditions and even distances that are wrong to calculate ratings then they can't be surprised that many tracks offer seemingly poorer performance. The more significant the errors the bigger the decline in a ratings ability to be correct.

I'm not saying that you can't win with "official" form, of course you can! All I am saying, is that our corrections produce a superior level of performance across all tracks long-term. Having the luxury of not have to second guess the many (good or bad) performances of individual horses, wondering if they are real or if the time was out by a second or more, goes a long way towards maintaining a clear mind so that I can then undertake more detailed form study which in turn allows me to execute my bets with confidence. Oh and the better rating performance doesn't hurt either!

[SpeedyBen]

Paul
I will read the article. In another life when I used to delve into ratings I never found one that didn't have weaknesses either by course or distance. This wasn't always a problem if I had enough data to know their weaknesses and ignore the ratings in the approp cases. Congrats on overcoming the problem. As a programmer for many years I'm not at all surprised at the amount of code reqd to sort this aspect out.

[gunny72]

I have posted most of this before (I don't know how to provide a link):

Often I see people ask how many losers can be expected before a win (a worst case scenario).

Let C=the prob of at least 1 win from n races, called the certainty, expressed as a decimal.
Let q=the prob of a LOSS in a race, expressed as a decimal.

Now to be 100% certain you need an infinite number of races so this theory expects you to decide on an acceptable certainty usually 90% or 95%.

Thus C=1-prob of all losing
that is C=1-q^n

If you solve this the number of expected outs is

n=log(1-C)/log q

You can use log or ln on your calculator but it has to be the same throughout the calculation.

For example, with the previous place percent example (22% win rate) and chosing C=95%,

n=log(1-0.95)/log0.78
n=log0.05/log0.78
n=12.06 so you can be 95% sure of at least one winner in about 12 races.

To be 99% sure it takes a run of about 18 or 19 races.

The above is based on the binomial theorem of probability which does have some caveats.
gunny