How much to wager?

[motorhead]

Hello all

The question is.........."how much to wager"?

Or maybe it should be, "how much is too much to wager"?

Here is the scenario

You have a bankroll of $1000

You also have a method that has a win rate of 45% with an r.o.i of +20 cents

How would you bet this method?

Thanks

[Dave Schwartz]

The conventional wisdom (which I disagree with) would be that you bet "Kelly."

Kelly is:

advantage divided by odds

Thus, we are missing a piece of information.

[1st time lasix]

Two percent of $1000 is $50. Unlikely to "tap out" at that level over a full day of simulcast opportunities where you pick and chooce your plays. So with that in mind....I would suggest you consider an "average" of around fifty bucks per race seeking overlays in the exotics that will pay at least 3-1 or more. That is just an average so you could wager about $30 on some plays and $70 on others where you have a stronger conviction. That fifty could be used in win pool...exacta pool....trifecta pool or in multi-race exotics depending on your handicapping opinion and ticket structuring capabilities. Nothing says you can't spread to all of the pools you feel offer overlay value. I prefer a "bet small to win big" philosophy...using small denominations in trifecta/superfecta or pick four wagers to avoid the IRS as much as possible on a bigger "signer" score.

[Dave Schwartz]

Check your math... 2% of $1000 is $20.

[njcurveball]

Hello all

The question is.........."how much to wager"?

Or maybe it should be, "how much is too much to wager"?

Here is the scenario

You have a bankroll of $1000

You also have a method that has a win rate of 45% with an r.o.i of +20 cents

How would you bet this method?

Thanks


I know that everyone looking for a complicated answer will jump on this. But I would simply flat bet at 5% of bankroll and update that amount weekly.

$50 a race for the first week. If you double the bankroll then $100 the next week.

Casinos do not fear losing when someone is betting the same amount because they have an edge.

They fear people like Card Counters in Blackjack who vary their bets greatly.

With such a high win percentage, it doesn't sound "prudent" to try to adjust on a race by race basis. IMO.

Especially with Kelly where your bigger bets will wind up on horse with lower win percentages as the odds go up.

Of course, you did not say how many plays per day/week you get.

[1st time lasix]

Check your math... 2% of $1000 is $20. Yes excuse me...I meant five percent....still thinking of adding my two cents! ha!

[chickenhead]

if my math is right, your average winning odds are ~1.66.

so quarter kelly around 3%.

[kenwoodallpromos]

Beg, borrow, steal or mortgage and increase your bankroll as much as possible if you have a sure thing that has a win rate of 45%, just do not bet less than even odds!LOL!!
Seriously, I concur that wherever between 2% and 5% that you feel comfortable is fine.

[JustMissed]

The conventional wisdom (which I disagree with) would be that you bet "Kelly."

Kelly is:

advantage divided by odds

Thus, we are missing a piece of information.

Dave, I know you are better at math than that.

(Win%*AvgOdds)-Loss%=Edge or ROI

He gave you the win% and edge--just solve for the unknown.

JM :)

[Overlay]

Assuming you meant that your 45% winners are returning a cumulative average of $2.20 for each $2.00 that you wager, I'd say the optimum wager size should be 6.923% of bankroll [(.45-.40909...)/(1-.40909...)].

[highnote]

Hello all

The question is.........."how much to wager"?

Or maybe it should be, "how much is too much to wager"?

Here is the scenario

You have a bankroll of $1000

You also have a method that has a win rate of 45% with an r.o.i of +20 cents

How would you bet this method?

Thanks


To quote Dr. Z, "Bet only when you have an edge and never overbet. Overbetting leads to Gambler's Ruin."

What you need to do is calculate your Mathematical Expectation and then apply the Kelly Criterion.

I do this by looking at a series of 100 bets. In your case, if you bet $1 over the course of 100 wagers then 55 times you will lose $1 and 45 times you will win $1.67. At the end of the 100 wager series you will have $20 profit.


Edge = (-1 x 55/100) + (1.67 x 45/100) = 0.2015

Optimal Kelly Bet = 0.2015 / Odds

So your optimal bet depends on 1.) Your edge, 2.) The odds and 3.) your bankroll.

If your edge is 20%, the odds are 3-1 and your bankroll is $1000 then the optimal Kelly bet is 0.2 / 3 x 1000 = $67

To answer your second question -- how much is too much to bet?

The technically correct answer is $67.01 Anything more than $67 will slow the rate of growth of your bankroll over the long run.

It's important to note that your average edge is 20%, but your edge on a given bet may be more or less than 20%. It's important to know your edge on a given bet. It's likely that your edge on a 3-1 shot is different from your edge on a 20-1 shot.

[thespaah]

Hello all

The question is.........."how much to wager"?

Or maybe it should be, "how much is too much to wager"?

Here is the scenario

You have a bankroll of $1000

You also have a method that has a win rate of 45% with an r.o.i of +20 cents

How would you bet this method?

Thanksdepends on how many of today's races you intend on betting. Is the entire $1,000 money budgeted for that day?
10 race cards you should not exceed more than 10 or 15% of your bankroll on any one race. That is of course you plan on wagering each race
Here's what I do.. I have my budget for the day. If I drain it ,I'm out. No double dipping. So I plan accordingly. Simple math.

[Dave Schwartz]

Dave, I know you are better at math than that.

(Win%*AvgOdds)-Loss%=Edge or ROI

He gave you the win% and edge--just solve for the unknown.

LOL - I think you missed my point... he has the win pct and ROI on the entire sample. Unfortunately, he must make one bet at a time.



Dave

[Overlay]

I understand Dave's point, but I thought that, strictly speaking, Kelly was based on the disparity between an individual horse's actual odds in a specific race and the long-term, overall fair odds of a method's selections (which was a concept that I had trouble incorporating into my wagering scheme). I think I've seen basing wager size on the disparity between actual odds and fair odds in each race (which to me is more feasible) referred to as "hyper-Kelly".

[JustMissed]

LOL - I think you missed my point... he has the win pct and ROI on the entire sample. Unfortunately, he must make one bet at a time.



Dave

That's right. Kelly changes which each individual bet. Obivously the lower the odds the greater the bet and vice versus.

Good luck to him nailing down the off odds.

JM :)

[GaryG]

Bet only when you have a horse that you REALLY like that is offering an acceptable price.....then bet as much as you can afford without getting sweaty palms or reaching for the pill bottle.

[Overlay]

Good luck to him nailing down the off odds.

In the thread on "In a perfect world, racetracks would...", I was thinking about those same difficulties when I suggested that tracks offer betting at fixed odds, where the payoff at the time you make your bet would be the payoff you'd receive. (I remember the late Tom Ainslie commenting at one time on how that would be possible.)

[GaryG]

In the thread on "In a perfect world, racetracks would...", I was thinking about those same difficulties when I suggested that tracks offer betting at fixed odds, where the payoff at the time you make your bet would be the payoff you'd receive. (I remember the late Tom Ainslie commenting at one time on how that would be possible.)I know I am dating myself but you could do that at the old Caliente track in Tijuana. They had both mutuels and books so you could actually "shop around for a price" with the books. The ticket they gave you had the odds on it.

[Overlay]

Looking at the positive side, even with changing odds and after-the-start odds drops, for those who are willing to evaluate and consider the winning chances of multiple horses in a race, the pari-mutuel system still does allow them to shop for wagering value in almost any race, based on how their assessment of the winning probability for each horse compares to the public's estimate.

[Robert Fischer]

You leave room for error and give a low estimate what the odds will be at off

For example If a horse is 3rd choice @ 7/2 with 2minutes left in a decent pool size you can safely assume that he is going to be over 2.6-1.
You can calculate your wager @2.6-1 with 2 minutes remaining and have it ready. With 1 minute you have time to give a quick glance at your odds recalculate if necessary and punch it in.
On the internet I may have a wagering site , a tote sight , and a spreadsheet windowed on my screen , all visible at the same time.

This can also be done with exacta probables. If you use multiple horses than you want start with more time and make it a continuing process up until about a minute or two left.

I don't find that I lose a lot of edge at off betting major tracks with this process.