## Forums

Where sports bettors talk!

# Forums

 Behind the numbers, Math & Sports Betting Topics #1,2,3,4 and 5 Posted
Joined: 07/22/2009
Posts: 4979

Stdbsdbsmdscnks,

I think you got a rough response because you were pretty much saying that a system of betting that has been used for a long time and by many successful gamblers will not work.  It is certainly not a martingale system.

I don't understand why you keep going back to the 50% thing, why does everything have to be 50%?

You are completely right that sports betting is far from an exact science and it's not like flipping a coin.  I have said over and over again that unless you are a person who can make a very accurate model (or has access to something which predicts events to the exact percentage) I would advise not using kelly.

I read something you said about you have no idea what you implied win percentage is in any given situation because teams change, situations change, etc.  Again, you could be correct.  How do you know if you are a person who has a model which reacts to all the changes which occur in sports betting?  This can be done by back testing.  You can test you model against games that have already been played, and if it proves to be profitable over a very long period of time, you can dabble with Kelly.

You also don't have to use Kelly which exacting standards.  Even if you are using 1/4, 1/3, or half kelly, you don't have to bet twice as much on a 2% as a 1% play.  You could say, "If it gives me between 2-4% I will bet one unit, if I get between 4-6% I will bet 1.5 and so on.  Yes you will not be winning as much as you could because you're not betting as much on higher yield plays, but you are decreasing your variance and leaving more room for error.

Lastly, I heard Jester say something about you shouldn't use Kelly with large favorites.  You can use Kelly with any moneyline, favorite or dog.  You have to remember that Kelly always tells you what the actually wager you put down should be.

For example, If you have a dog of any size, you simply bet was it tells you.  If you want to play full kelly, and it says 2% at +200, you bet 2% to win 4%.

When you have a favorite, it's more complicated.  If is tells you 2% when you have a -180 favorite.  That means the BET AMOUNT should be 2%, not the win amount.  So 2%/1.8= 1.11.  You would bet 2 to win 1.11.

You want a handicapping tip?  Spend as much time getting a good number as you do researching games. Depositing/withdrawing/depositing/withdrawing is a bitch, but I guarantee you will lose less, or win more.  I can't guarantee that for any other piece of advice

"There is no such thing as luck, just random fluctuations of probability"

Joined: 10/19/2010
Posts: 749

Furthermore, and i am just free associating here, I don't believe in star system plays, they are for touts, not guys like me that use sports investing as a necessary second stream of income.

If you bet 1 unit on the majority of your games, it is crazy to bet 3 units on other plays unless you think those plays offer 3x the value of your normal play.  To me, that is virtually impossible because if you can find plays that offer 3x the edge of your normal bet, you probably don't have an edge on your 1 unit plays.

However, if you are running good and in the zone, I think you should step up your unit size when appropriate; it is a fine lie though.

I would be interested in hearing others feedback on this point.  Thanks.

"I mean really, I don't see why you people just can't watch the horses run around the track and not bet on them."

Joined: 07/22/2009
Posts: 4979

The Game

However, if you are running good and in the zone, I think you should step up your unit size when appropriate; it is a fine lie though.

Have to strongly disagree with this point.  I know how it feels when your running well or running poorly, but that is nothing but variance, and you have no idea when that variance is going to turn in any particular direction.  You aren't actually any better of a sports bettor when you happen to be running well, you just happen to be in a good patch of variance.  There is no way to determine how long that good patch will last.

You want a handicapping tip?  Spend as much time getting a good number as you do researching games. Depositing/withdrawing/depositing/withdrawing is a bitch, but I guarantee you will lose less, or win more.  I can't guarantee that for any other piece of advice

"There is no such thing as luck, just random fluctuations of probability"

Joined: 11/01/2011
Posts: 2306

It's getting better than real housewives of New Jersey. jesus we just trying to get more math in front of everyone and hopefully learn. Im drinking another goose and cran and going to bed!

Rule #1. Bankroll Management!

Rule #2. Discipline!

Joined: 08/13/2012
Posts: 310

I would have to say I think star-system plays are appropriate in some situations.  I use between 1-5 unit plays.  I understand the argument against this, but I find it useful and here's why...(Feel free to correct me if this is an inaccurate or misinformed way to do this, for I am always learning)

For example if I feel I have a 54% chance of winning a game, I don't think a 3 unit plays has 3x the chance of winning (so 162%)...I think I have a 1.7% advantage over the break even point.  If my game doesn't have a >53.8% (1.5% over break-even) I won't make the play.  So a 2 unit play would be twice as much as the 1.5% over break-even (3%)(55.3%)

So, my breakdown would be...

1 unit play:  53.8%>Expected play win %>55.3%

2 unit play:  55.3%>Expected play win %>56.8%

3 unit play:  56.8%>Expected play win %>58.3%

4 unit play:  58.3%>Expected play win %>59.8%

5 unit play:  59.8%>Expected play win %

Thoughts on this approach?

CFB 2012

5 units (18-5) +62.5 units
4 units (28-17) +37.2 units
3 units (28-25) +1.5 units
2 units (21-15) +9.0 units
1 unit  (7-7) -0.7 units

Total record (102-69)(60%) +109.5 units!!

Joined: 10/19/2010
Posts: 749

Lloyd Christmas

The Game

However, if you are running good and in the zone, I think you should step up your unit size when appropriate; it is a fine lie though.

Have to strongly disagree with this point.  I know how it feels when your running well or running poorly, but that is nothing but variance, and you have no idea when that variance is going to turn in any particular direction.  You aren't actually any better of a sports bettor when you happen to be running well, you just happen to be in a good patch of variance.  There is no way to determine how long that good patch will last.

I tried to couch my comments with the caveats "when appropriate" and "fine line" and I do not necessarily disagree with your comments above but we also won't manufacture 100% consent on this issue either.

Here is why:

1. I have a small roll and if I want to increase it I should be more aggressive than if I had a larger roll.  Aggressive is a relative term because I have an extremely conservative investment profile (approximately 90% plays are 1 unit).

2. If I am running good, just in the zone Lloyd, I will ocassionally step out because even if I lose my 1.5 or 2 unit play(s), I will still be ahead for the week and can dial back immeadately if things take a negative turn.

I understand your point about experinceincing positive variance and not being a better capper as a result but I also think are times when one is just on point; where the decisions you make match the outcomes.

I know you disagree and perhaps I am just fooling myself but overall, this strategy has resulted in a net postive for me.  Just as dialing back when I hit negative variance not only helps me sleep better but preserves my roll and allows me to fight another day.

"I mean really, I don't see why you people just can't watch the horses run around the track and not bet on them."

Joined: 10/19/2010
Posts: 749

m00coyne

I would have to say I think star-system plays are appropriate in some situations.  I use between 1-5 unit plays.  I understand the argument against this, but I find it useful and here's why...(Feel free to correct me if this is an inaccurate or misinformed way to do this, for I am always learning)

For example if I feel I have a 54% chance of winning a game, I don't think a 3 unit plays has 3x the chance of winning (so 162%)...I think I have a 1.7% advantage over the break even point.  If my game doesn't have a >53.8% (1.5% over break-even) I won't make the play.  So a 2 unit play would be twice as much as the 1.5% over break-even (3%)(55.3%)

So, my breakdown would be...

1 unit play:  53.8%>Expected play win %>55.3%

2 unit play:  55.3%>Expected play win %>56.8%

3 unit play:  56.8%>Expected play win %>58.3%

4 unit play:  58.3%>Expected play win %>59.8%

5 unit play:  59.8%>Expected play win %

Thoughts on this approach?

Thank you for the thought provoking post and I am certainly got the guy to offer math corrections:)

Here is what I can offer:

I am essentially a fundamental, situational and market trader.  I consider myself conversant in the "personell and scheme" matchup aspect of the sporting event and I am above average in comparison price shopping to land the best number.   I rely on my "risk quotient" so to speak when executing any trade and won't pull the trigger unless I percieve an advantage.   Since I don't run models, I have no definitive way to define my edge with any granular accuracy and even if I did, I am not sure it is possible to quantify ones edge to a degree where I would feel comfortable laying down 5 bills when my standard play is a dollar.

I don't have a patent on the truth and my earlier post was an approach that works for me and fits my risk profile. I simply can't take a 5 unit hit, it would exceed my risk parameter protocols.  Other people, may feel I am leaving money on the table but not stepping up when my edge is greater.   Thankfully, there is room for disagreement.

"I mean really, I don't see why you people just can't watch the horses run around the track and not bet on them."

Joined: 11/23/2009
Posts: 270

The Game

Furthermore, and i am just free associating here, I don't believe in star system plays, they are for touts, not guys like me that use sports investing as a necessary second stream of income.

If you bet 1 unit on the majority of your games, it is crazy to bet 3 units on other plays unless you think those plays offer 3x the value of your normal play.  To me, that is virtually impossible because if you can find plays that offer 3x the edge of your normal bet, you probably don't have an edge on your 1 unit plays.

I completely agree.  I wish i could identify which of my plays i expect to hit at 54% vs. 57% but i simply cannot and would be kidding myself if i tried.  .For me, and 95% of bettors, flat betting is the best approach IMO.

Joined: 10/11/2008
Posts: 133

Here is the answer to the question that I have been asking ... instead of implying or coming up with each owns theory to imply probabilty based on handicapping you would use this formula .... Now if you plug these %s in to the Kelly Critera when you place a bet, I still am curious as to the outcome... I will still be working on this problem and share the results.

So how do we convert the 'plus' and the 'minus' moneyline odds into their implied probabilities?

For 'minus' moneyline odds, it's the following calculation.

So let's take our example of a moneyline odds offer of -120.

Implied probability = (- (-120) / ((- (-120)) + 100)

Implied probability = 120 / (120 + 100)

Implied probability = 120 / 220

Implied probability = 0.545

Multiplied then by 100, we get the implied probability percentage of 54.5%.

Now the calculation for 'plus' moneyline odds, we make the following calculation:

So let's use our example of moneyline odds offer of +180

Implied probability = 100 / (180 + 100)

Implied probability = 100 / 280

Implied probability = 0.357

Multiplied then by 100, we get the implied probability percentage of 35.7%.

Joined: 10/11/2008
Posts: 133

At a glance , it looks like i have inverted the implied probablities, but when you plug these implied probabilites into the Kelly Criteron , I am not sure how this would come out , it seems then you would aways be betting on the Favorite and never a dog according to this equation, where I thought the opposite.

• 172 Replies
• 4 Subscribers
• Posted7 months ago