Value betting and the bad runs

blimey! and you wanted putter ons last week

Early season mate. **** happens and its a marathon not a sprint etc. Might be down a bit so far but ave price taken is 28.81% better than SP so don't think I'll be far away for too long. Anyway as i said with the prices I take can be long periods without winners. Remember on year there was a 57 loser run and another year with one winner in 73 but never had a losing year and profits on EPS over 20% over that period and 17% overall to stake over that period.

If I had some runners this year the prices taken would be over 30% better than SP! Anyway would still like to get in the black soon.

so long as you are:
not over-staking
are reducing stakes in line with bankroll

Then I would only panic if you stop beating SP.

If you can't handle the bankroll swings then reduce your stakes further for your sanity.

i think reducing stakes is what you need as u are uncomfortable with the idea of losing a significant portion of ur bank

http://www.youtube.com/watch?v=-QaHN-57Ajo

I think reducing your stakes is the only way to go - seriously.

Thoughts like this can have some horrible effects on your mental state regarding gambling so, you need a more manageable amount.

Heartbreaking stuff ... especially when those big overlays keep running 2nd, and the favourite backers keep getting the money.

RG- Have you not tried to puzzle out why you are on a "bad start" to jumps season betting?

Of course it is getting far more competitive in NH racing in Britain. More and more Irish trainers attacking races.....and also with more trainers, who are considered flat/Aw trainers,  having NH runners, making it harder to assess races etc.

The above factor just makes solving the puzzle more "interesting".....though harder!

Of course bad runs of results will happen with such a complicated sport. As long as your "edge/s" hold up, then hopefully you will turn it around as you have done in the past. GL

I have some stats on how applying a HIGH ODDS strategy could have worked out for some days analysed --

    DAILY
    TOTALS
    OF HIGH
    ODDS
    STRATEGY
   
   
    -42.2695 A LOSS FOR THE DAYS COVERED......
   
   
   
15.07.08    65.9085
16.07.08    -31
17.07.08    -46
18.07.08    -37.6785
19.07.08    -1.5685
21.07.08    32.8825
22.07.08    40.1585
23.07.08    -9.35
24.07.08    -23.95
06.08.08    -11.6385
07.08.08    -41
08.08.08    56.9315
09.08.08    -15.447
03.03.09    -8
04.03.09    -15
05.03.09    2.5515
06.03.09    -11
07.03.09    -25
08.03.09    -5
09.03.09    59.9
10.03.09    -38
11.03.09    -64
12.03.09    34
13.03.09    -11.8695
14.03.09    31.1035
16.03.09    -15
17.03.09    12.178
18.03.09    40.2155
19.03.09    -10
20.03.09    -10
21.03.09    -8
22.03.09    6.9265
23.03.09    -12
24.03.09    -7
25.03.09    -17
26.03.09    4.85
27.03.09    -15
28.03.09    60.1275
   
   
   
   
   
25.07.08    -0.501



Some horrific days there !
==============================================================


However, applying same strategy to other days the following results could have been attained....


    DAILY
    TOTALS
    OF HIGH
    ODDS
    STRATEGY
DATE   
   
   
   
   
    251.538 total profit for period checked below...
   
01.04.08    14.428
02.04.08    -4.496
03.04.08    -34
04.04.08    25.695
05.04.08    10.65
06.04.08    -20
07.04.08    -22
08.04.08    8.5355
10.04.08    29.8905
11.04.08    24.198
12.04.08    85.972
13.04.08    -16
14.04.08    70.0715
15.04.08    -28
16.04.08    -24.45
17.04.08    -55
18.04.08    71.248
19.04.08    16.671
20.04.08    28.453
21.04.08    -32.397
22.04.08    41.0845
23.04.08    -6.9825
26.04.08    -24.2125
27.04.08    48.17
28.04.08    7.65
29.04.08    108.677
30.04.08    -10.1065
01.05.08    10.6
02.05.08    -44
03.05.08    -17.95
04.05.08    48.9565
05.05.08    -62.812
06.05.08    -37.3
07.05.08    -24.06
09.05.08    84.1265
10.05.08    36.221
11.05.08    17.8095
12.05.08    58.4915
13.05.08    -26
14.05.08    -25.883
15.05.08    -15.6115
20.05.08    -49
31.05.08    -15.8



Again some horrific runs of results........just the nature of such a strategy though

lowering stakes or liabilities is not the way to go .

Even the strategies on LOW ODDS horses can give rise to losing days where big losses are made.

TOTAL   
OF   
STRATGIES   
EXCLUDING    
COL G   
    DATE
   
   
   
   
349.2659509 PROFIT OVER DAYS BELOW.   
   
20.5639066    01.04.08
-10.0938267    02.04.08
38.6947717    03.04.08
13.669976    04.04.08
-10.6605929    05.04.08
-0.4212163    06.04.08
28.3488027    07.04.08
20.7134257    08.04.08
25.070692    10.04.08
0.5255682    11.04.08
21.7874368    12.04.08
6.2248177    13.04.08
19.4580971    14.04.08
0.7812552    15.04.08
10.1805586    16.04.08
3.795669    17.04.08
51.8246063    18.04.08
-0.9065743    19.04.08
24.3014318    20.04.08
-5.4649061    21.04.08
-0.6036892    22.04.08
10.1154503    23.04.08
-6.8791332    26.04.08
-8.6084789    27.04.08
36.0557858    28.04.08
-53.7191061    29.04.08
33.8839862    30.04.08
56.461575    01.05.08
15.6405726    02.05.08
7.6529559    03.05.08
42.0482245    04.05.08
8.648357    05.05.08
11.4390547    06.05.08
-28.1834445    07.05.08
-15.0606362    09.05.08
-23.9178807    10.05.08
16.3270924    11.05.08
-11.8281699    12.05.08
23.2658576    13.05.08
-57.0132742    14.05.08
12.0668163    15.05.08
14.7323415    20.05.08
8.3477949    31.05.08
=====================================================================


FAVS    
STRATEGY   
RE-RANK   
   
TOTAL   
OF   
STRATAGIES   
EXCLUDING    
COL U   
   
401.791783 PROFIT OVER DAYS BELOW.   
   
   
   
70.2273337    15.07.08
18.2781564    16.07.08
3.5448172    17.07.08
-9.7217215    18.07.08
3.3692797    19.07.08
-3.9039255    21.07.08
28.79880031    22.07.08
71.7490646    23.07.08
33.206587    24.07.08
14.5561515    06.08.08
-33.5084364    07.08.08
14.4439349    08.08.08
59.5699229    09.08.08
13.3870501    03.03.09
25.4832747    04.03.09
2.3072125    05.03.09
9.6740934    06.03.09
-9.9543787    07.03.09
16.2505    08.03.09
13.6628713    09.03.09
2.2408567    10.03.09
-1.9078952    11.03.09
2.5657612    12.03.09
14.8420186    13.03.09
-9.827527    14.03.09
19.9917168    16.03.09
-8.4975576    17.03.09
6.1066239    18.03.09
16.5578348    19.03.09
13.1527579    20.03.09
-10.3004167    21.03.09
-0.1635313    22.03.09
-3.2102681    23.03.09
0    24.03.09
0    25.03.09
0    26.03.09
0    27.03.09
0    28.03.09
===============================================================


Both the HIGH ODDS strategies and the LOW ODDS strategies are based on VALUE bets.......so bad losing days occur whether HIGH ODDS or LOW ODDS are attacked. Just appears to be the nature of such VALUE bets strategies.
   
   
   
   
   
18.8208209    25.07.08

Managing that money

May 3rd, 2011 - Posted by Andrew in Betting Theory
BettingExpert Blog editor. Always taking the alternative route to finding the value.


What's the best way to manage you betting bankroll? How much should you bet on any one bet? Is there a way to optimize value in the betting markets? Andrew takes us through a few of the options and tell us how he determines how much money to put down.

There are two essential characteristics to being a successful long term sports gambler. The first is the ability to recognise value in a given market, and the second is how to manage your investment bankroll. In other words, how to optimise that value and manage risk.

Many gamblers have the ability to recognise value in a particular sports gambling market, whether that be by 'gut' and 'feel' or some statistical model to asses the probability of an outcome. But once value has been identified, how much do you invest in that wager to make the most of that value while protecting your bankroll?

There are many and varied investment strategies that can and have been applied to sports gambling, a number borrowed from the world of finance. Any Google search however will surely confuse the uninitiated, as a carnival of mutant theories and strategies parade down the page. Regardless, by the end of this article it's hoped even the most inexperienced of sports gamblers will be familiar with the most popular money management strategies, will have a better understanding of the fundamentals behind money management and how to increase their chances of joining the small percentage of sports gamblers who make a long-term profit.

A slice of the pie
One strategy is to maintain a constant percentage of your bankroll with each bet. Its assumed that this method will protect any gambler from losing their entire bankroll as the amount to be bet diminishes as your bankroll diminishes. It's a little like Zeno's paradox. The arrow will never reach its target as it halves the distance at each interval. However, while unlikely, even betting 5% of a $1000 bankroll will leave you with a bankroll of less than $200 after 33 consecutive losses.

But more to the point, it assumes equal value for each bet and additionally it expects equal odds being offered for each bet. Would you really want to bet the same percentage on a $1.70 favourite as a $3.50 outsider without any recognition of their respective value? The answer should clearly be no. Further, even if you only bet on $1.70 favourites, is the probability of each $1.70 favourite winning identical in each instance? In other words, does each $1.70 favourite offer the same value? It could be the case, but its unlikely.

The Slice of the Pie strategy while offering a manner of managing your bankroll, fails to recognise value as a key part of a successful management strategy. Essentially, in the end, it will only manage the way you lose your money.

Inside out
This method is similar to the Slice of the Pie, except that it takes the further step of taking into account the odds being offered for a particular bet. So for example, you bet %5 of a $1000 bankroll as a standard unit, $50. If betting on $2 odds, the unit remains $50, but when betting on say a $11 outsider, that unit becomes $5. It can be calculated like this:

eg. ($1000 * 0.05) / (11 -1) = 5

This works fine when betting on outsiders but what if we wish to bet on short priced favourites? If betting 5% of any bankroll, betting on a $1.05 favourite would mean betting your entire bankroll. Now I would never recommend betting on anything remotely close to a $1.05 favourite let alone your entire bankroll, (unless the odds were on the sun rising tomorrow, but even then I would probably spend a nervous night watching the darkened night sky). Regardless, the limitations of the Inside Out strategy should be clear as it still doesn't take into account the recognised value of a particular betting proposition.

K is for Kelly
While it still has its critics, the Kelly Method has stood the test of time since its creation in the mid 1950's. Essentially this method takes into account both the probability of a given team or player winning and the value of the odds offered in relation to that probability - the overlay. This means that it suggests you bet more depending upon how great the overlay, but it also means you have to assess the probability of a given outcome with consistency. (Assessing probability is something we will discuss in future Blog posts).

The overlay is calculated simply as:

Overlay = (probability * odds) - 1

There have been many amendments to the Kelly method over the years, and although the fundamentals remain the same, each edition of the method offers its own insights.

The Full Kelly - While a proven method, the Full Kelly can make a wild ride of your betting experience and can suggest risky amounts be bet. It can easily recommend a bet of even %50 of your bankroll, which can reduce your bankroll to merely nothing in short time. On the other hand, a winning streak can send your banroll into orbit

It is calculated as:

Percentage of bankroll to bet = overlay / (odds - 1)

e.g With an overlay of 0.20 and odds of $2.40 and a bankroll of $1000, the Full Kelly would recommend a bet of $143, or 14.3% of the bankroll.

The Fractional Kelly - This is a simple and conservative amendment to the Full Kelly method whereby the gambler only bets a certain fraction of the recommended bet. It could be %50, known as the Half Kelly, 25% the Quarter Kelly or any percentage you feel comfortable with. Further, while the intuition may be that this will reduce your winnings according to the percentage you choose, it can be shown that a fractional Kelly method can return better results longterm than the Full Kelly method.

The Constant Kelly - The same as the Full Kelly method, but rather than recommending a percentage of a varying amount, it recommends a percentage of a constant. So for example, instead of suggesting 15% of a diminishing or improving bankroll, it recommends 15% of a constant amount.

Busy busy busy
One of the drawbacks of any Kelly method is the issue of wanting to bet on multiple events at the one time. Lets say there are 4 games you want to bet on, being played at the same time. And what if the recommended percentage of bankroll for Bet A is 25%, Bet B is 35% and Bet C 35% and Bet D 40%? This amounts to betting 135% of your bankroll, which is obviously impossible. One solution to this problem is to adjust the percentages proportionally so that 100% of the bankroll can be bet. i.e Bet A would proportionally become 19% and so on, 25% being 19% of the 135% recommended. The issue with this is that firstly, you're still betting %100 of your bankroll on 4 events which could all easily lose, and secondly, it means you're not giving the same value to say a 25% bet on a busy day (where in this example it becomes %19) as you would on a day when it might be your only bet.

One way to solve this is to use a fractional method so that no matter how many events you want to bet on in a single day, the total recommended is unlikely to eclipse %100 of your bank. This could well work, but as I know myself, it is easily possible to be betting on up to and over 20 events in a single day, meaning the %100 could still be eclipsed.

So what do you suggest?
Personally, I use a 10% fractional Kelly method. This allows me to protect my overall bankroll while I diversify and place many bets on sporting events being played daily.

So if my bankroll is $5000, and I am betting on a $2.60 team with an assessed winning probability of %50, the recommended bet amount would be calculated as:

(bankroll * chosen fraction) * (overlay/(odds - 1))

i.e ($5000 * 0.10) * (0.30 / 1.60) = $93.75

Try it out for yourself
While many bankroll management strategies are available to apply, I believe that the fractional Kelly method is best, as it takes into consideration the odds on offer, the probability assessed of a team or players winning and the resulting value identified in order to recommend a bet amount that will optimise that value without risking your bankroll and an early end to your career as a sports gambler.

lowering stakes or liabilities is not the way to go .

I'd agree until he hits a winner then I'd reconsider it.

Why is whether he hits a winner relevant to lowering stakes?

If I was betting to a level stake and I was on a 50+ losing streak, I wouldn't want to lower my stakes at this point as if you were betting at value you would expect a return soon. Unless I was financially ruined.

if you hit 50+ losers regularly - stake 1/500th of bank on each selection .

"If I was betting to a level stake and I was on a 50+ losing streak, I wouldn't want to lower my stakes at this point as if you were betting at value you would expect a return soon. Unless I was financially ruined."


I would expect a return exactly as often as if I was on a 50+ winning streak.

I would find betting on selections that throw up runs of 50+
and so would most other people. I bet only on prices between
1/3 and 3/1, as I would rather be looking at a minimum of 25%
strike rate than that of a 2% strike rate.

..runs of losers

I think I'm still calm about it all and as bet sizes are determined by modified Kelly of current bank thats not a concern.

The irish horses are a royal pain in the backside especially as they are just way better handicapped than our lot. I get concerned as historically October is a great month for me and typically I need to be way in front for I have a bad spell through February before March/April being very good again.

More concern to me at this stage are the overnight vultures who are ruining the prices in the morning.

Why are Irish horses better handicapped than English?

Racingguru, I don't think you have anything to worry about, sounds like you know exactly what you're doing and what to expect. I've dropped 50% of the bank before too, but all the time knew that I was still getting value, but it is always stressful to drop so much - human nature.

Back then I reduced stakes but of course that just made the recovery longer, so now I keep level stakes no matter what. That said, there has to be a point where reducing stakes would be prudent I still believe, so I increased my bank so that hopefully I'll never suffer a 50% drop again, but some would rightly see that as not utilising capital I know. GL with the recovery :)

Investor - just my opinion but think with the fields in ireland for chases (which is what I'm referring to) being much bigger a really good run say 5th of 16 can go unraised whereas in UK its more likely to be a 2nd of 5 and raised 3 for nothing much of a performance.

Just look at the amateurs race at Chelt on friday. Would have been 1st 3rd,4th or 2nd,3rd and 4th depending on whether you thought Romanesco was staying. What do Uk horses ever do when they do over to Ireland for a handicap. Was Barnbrook Again the last winner?

Cheers racingguru.

Trevh, I don't get it. Sounds to me like what you were doing before makes more sense. Why not use some form of fractional Kelly adjusted for Betting exchange specifics.

If you go on a crazy losing run and lose 50% of bank, of course you have to reduce stakes. If you think you can still bet to the same level staking after dropping 50%, it just means you were massively under staking before the loss.

I know what you mean Investor, but ateotd I'm happy with what I make even though I know I under stake as a percentage of my bank, but that's where my comfort zone is. I'm still ambitious and increase stakes every 3 months and can afford to have a bigger bank than I actually need, so am very happy :)

If a person is backing heads in a coin toss at 5/4 and endures a heavy losing run it's pretty easy to absorb a losing run using a reasonable
staking strategy (kelly being the optimum one). In the long run you are bound to win as your getting factual value.
Value in betting on Horse racing is subjective though and the punter has to work out during a losing run whether he's encountering
bad probability variance or his judgement has become poor. It's not that easy to tell the difference as the mind can play tricks.
The orginal poster states that he is "massively beating the odds on everything I bet" (assume this means taking a price better than
SP). This doesn't automatically mean that value is being obtained as the SP isn't a set bench mark as to the horses true price.
A person constantly backing horses at a significant price bigger than SP is showing good skill at spotting trading opportunities
but there is no guarantee that he's good at spotting win odds value and he may well lose in the long run.

Oct 28, 2012 -- 10:59PM, tinkler wrote:

If a person is backing heads in a coin toss at 5/4 and endures a heavy losing run it's pretty easy to absorb a losing run using a reasonablestaking strategy (kelly being the optimum one). In the long run you are bound to win as your getting factual value. Value in betting on Horse racing is subjective though and the punter has to work out during a losing run whether he's encounteringbad probability variance or his judgement has become poor. It's not that easy to tell the difference as the mind can play tricks. The orginal poster states that he is "massively beating the odds on everything I bet" (assume this means taking a price better than SP). This doesn't automatically mean that value is being obtained as the SP isn't a set bench mark as to the horses true price.A person constantly backing horses at a significant price bigger than SP is showing good skill at spotting trading opportunitiesbut there is no guarantee that he's good at spotting win odds value and he may well lose in the long run.

very good point.

quick update - few winners and a nice fc so back in the black but nowhere near the % profit I get normally in Oct. These overnight vultures marking bookies cards are a real pain but such is life. Just having to adapt to the changing climate on Early price betting.