[BChrisB]

I just don't understand it. I google this stuff all the time and yet so many people claim flat betting to be a preferred method.

Can some of you advocates of flat betting state to pros to this? I honestly see nothing but cons,

[That Foreign Guy]

It's much easier to understand.

It reduces the risk of going broke by over-estimating your edge (which even fractional Kelly suffers from).

Most players are losers so it reduces the risk of going broke by making them bet less (which is closer to Kelly optimal size of 0).

Most people with online sites are losers so they think it's the best method.

[Wrecktangle]

Ask yourself these questions:

Do you understand the time-value of money?
Do you keep deadly accurate records of your results?
Do you understand probability and statistics?
Do you have insight into how leagues change over the season?
And the killer, do you understand Bayesian Statistics?

If you don't, Kelly of any type will do more harm than good.

[Toples]

i use flat betting, but increase bets after 20% bankroll increase. Kind of flat-kelly betting.
Knowing i have an edge, i realize that my bad streak cant go on forever and if i lower stakes after bad run is over, i will win less...

[MadTiger]

The sites want you to lose with grace, or, if you are a blind squirrel who finds a nut, not to hurt them too bad (i.e. a 10-unit play).

[BChrisB]

Most people with online sites are losers so they think it's the best method.

This much I have noticed. Believe it or not but you can really judge a site by the sportsbooks it advertises for.

[BChrisB]

i use flat betting, but increase bets after 20% bankroll increase. Kind of flat-kelly betting.
Knowing i have an edge, i realize that my bad streak cant go on forever and if i lower stakes after bad run is over, i will win less...

If just seems that even with increases in "flat" bets. That it won't take long before your playing catch-up with the accumulated juice from your losing picks.

So your break % is something that will always go up and you will eventually be chasing for infinity.

Someone, please correct me if I'm wrong.

[solring]

...because if you can't calculate your edge or your winning percentage, you can't calculate the optimally correct bet size...

[BChrisB]

I'm going to find a coin flip program and adjust tails in the code reflect the 10% standard juice from a side bet and see what happens. I'm predicting in less than 200 flips a 50 unit bankroll will go broke at 1 unit bets.

I'll post my results

[splash]

i use flat betting, but increase bets after 20% bankroll increase. Kind of flat-kelly betting.
Knowing i have an edge, i realize that my bad streak cant go on forever and if i lower stakes after bad run is over, i will win less...

That is terrible. Not only does calling it a "flat-kelly" system not make any sense, but that means that after a bad streak, you're not lowering your bet size, which means you're probably overbetting in relation to your bankroll size, which sounds like a pretty good way to go broke.

[uva3021]

In waiting for BChris to post his results, I went ahead and created a coin-toss program. You can use it as well for a bet simulator. Its in a zip file so I can upload an .xlsm excel sheet.

Enable Macros in Excel

Update: 12/10/2010 4:22 pm EST

[LegitBet]

Very inciteful
I may just take out the good old abacus!

[BChrisB]

In waiting for BChris to post his results, I went ahead and created a coin-toss program. You can use it as well for a bet simulator. Its in a zip file so I can upload an .xlsm excel sheet.

Enable Macros in Excel

Awesome. I like this. I was looking for one in php. I want to use this one but I'm getting:

Cannot run the macro 'Coin_Toss.xlsm!Sim_CoinToss'. The macro might not be available in this workbook or all macros may be disabled.


Any clue why I'm getting this in excel. I like what you did and am definitely saving this one for future play.

EDIT: I got it working... I'll be tipping you some points tomorrow for this

[Peregrine Stoop]

the second you think about more than flat betting, you have to think about quantifying your edge. This leads to the dangerous idea that you could bet either side of a game depending on the price. Sites don't want too educated of a betting public.

[BChrisB]

FLIPS 200
LINE -110
TYPE Flat
EDGE 0.09
UNITS 0.5
EV 0.0632%
EG 0.0621%
OUTCOME GUESS
HEADS TAILS
TOTAL 29.5
WIN 0
WINS 90
RESULT -0.55
LOSSES 110


I got the excel macros running so here's what I have based off 200 flips -110 for losses and based off .5 Units (which would make it 1 unit on a 50 unit bankroll)

So you didn't go broke in 200 flips but you literally are on your way there.

[BChrisB]

Total 40.35 wins 97 losses 103

[u21c3f6]

If you are wagering for recreation, then wager whatever amount is "fun" for you. If one is trying to make a profit, before you try to figure out what % of your bankroll to wager, you must have a selection method that produces an edge. No edge, no wager. If you can define your edge, then Kelly tells you that the max % of your bankroll that you should wager is your edge/odds.

The coin flip example has no edge and therefore the correct amount to wager is $0. on each flip which would mean that you still have your original bankroll.

Joe.

[Wrecktangle]

The century system is essentially dividing your BR by 100 and betting those 1/100 units until you win/lose 20% upon which you redivide to establish an new bet size.

It is extremely conservative and I recommend it because I doubt there are 1% of the betting public who can really calculate their win%. I fall back to it when a league gets chaotic, or I simply quit betting.

[Arilou]

Flat betting is recommended primarily because it is emotionally easier and easier in general. You don't have to think about bet size all the time. You don't have wild swings. You know where you are at and what you want. Most importantly, you don't get into situations where you lose one game, win three and end up behind. Kelly can be very emotionally draining. The other reason is the worry that when you think you have the biggest edges, that's when you are likely to be "trapped."

Let's assume you're betting to make money. The idea that you throw up your hands and say you know nothing if you can't calculate your exact edge is simply incorrect. In fact, you know at least one thing: You think you do have a substantial edge, or you wouldn't be betting. Betting 1% implies at most a 2% expected edge or so, so you'll double your bankroll approximately never. Why do you even bother?

[Grind-It-Out]

Flat betting is best for three types of people.
1) People who have a negative edge - and this is most people.
2) People who can't quantify their edge - and this is most people.
3) People who don't think rationally - and this is most people.

I hear crap all the time like "I went 3-1 but I lost money cuz I didn't flat bet. F*** kelly." Now, someone who would make a statement like that likely isn't intelligent enough to be able to quantify an edge anyway, but even if they could they'd be screwed.

[CHUBNUT]

Flat betting is best for three types of people.
1) People who have a negative edge - and this is most people.
2) People who can't quantify their edge - and this is most people.
3) People who don't think rationally - and this is most people.

I hear crap all the time like "I went 3-1 but I lost money cuz I didn't flat bet. F*** kelly." Now, someone who would make a statement like that likely isn't intelligent enough to be able to quantify an edge anyway, but even if they could they'd be screwed.

Its you that's delusional flatbet and raise stakes according to the circumstances not some mathematical jest from some egghead half a century ago. I love you people who quote garbage from people who never had a cent on a game. "people who cant quantify their edge" as if you can.

[uva3021]

What I need to do is run the same simulation side by side, I'll do that at length

For now, here is my latest Kelly coin-toss sim

TOTAL -551.35 WINS 96 LOSSES 104

[uva3021]

wait there is a minor (major?) glitch, i had units won = 1 regardless of bet size but units loss = bet size and vigorish, lol

so the kelly simulations were skewed

here is the update

[CrimsonQueen]

Kelly is too aggressive.
If you can pick at > 52.38% win rate, then you are guaranteed money flat betting. If you pick at 60% but lose a couple of the wrong bets, you could end up down a lot if you win 90% of your 1 unit plays and lose your one 75 unit play.
If you can accurately assess your edge, then you can do Kelly, but I still think even knowing my edge, I'd much prefer flat betting.
Also, most people bet incorrectly when they adjust their bet size: If you bet 1 Unit, thinking you have a 5% edge, and 1 Unit=2% of your bankroll, then when you're betting 10 units on a game, you must think you have an X% edge, where X = IMPOSSIBLY TOO HIGH! You're betting 20% of your bankroll on one game...

[CHUBNUT]

Farcical. what next.

[uva3021]

Did one side by side now so one can make a comparison

[BChrisB]

This debate started in my head a few years ago and had nothing to do with sports betting. Actually it was blackjack. The Colorado casinos at the time had a $5 dollar limit. So at the blackjack table your min and max bet would be 5 dollars.

My argument on this w/ blackjack is if I was laying the same bet over and over I was just surrendering myself to the house edge. It's mathematically simple, betting the same amount time after time ensured the that standard deviation could never be compromised in the long run.

I just don't see how this is any different with sports betting if your always betting the same unit and the house always has a mathematical edge (juice.)

My conclusion is that no finite bankroll could ever last in the test of time with flat betting when there is ever a house edge.

[Inspirited]

Yeah, thanks for stating the obvious. If you don't have an edge you lose. .

[BChrisB]

Yeah, thanks for stating the obvious. If you don't have an edge you lose. .

Was I stating the obvious?

[TT22]

Was I stating the obvious?

Yes.

You need to BEAT the juice. Then you have EDGE. If you can't do that, then you should bet ZERO $.

[BChrisB]

Yes.

You need to BEAT the juice. Then you have EDGE. If you can't do that, then you should bet ZERO $.

Gee!! Well thank you for stating the obvious. Clueless?

[Wrecktangle]

Flat betting is recommended primarily because it is emotionally easier and easier in general. You don't have to think about bet size all the time. You don't have wild swings. You know where you are at and what you want. Most importantly, you don't get into situations where you lose one game, win three and end up behind. Kelly can be very emotionally draining. The other reason is the worry that when you think you have the biggest edges, that's when you are likely to be "trapped."

Let's assume you're betting to make money. The idea that you throw up your hands and say you know nothing if you can't calculate your exact edge is simply incorrect. In fact, you know at least one thing: You think you do have a substantial edge, or you wouldn't be betting. Betting 1% implies at most a 2% expected edge or so, so you'll double your bankroll approximately never. Why do you even bother?

98%+ of the public are long term negative EV punters, that's why. Everyone seems to be in denial about it.

Let me repeat that: 98% OF YOU GUYS WILL BE LONG TERM LOSERS IN THE SPORTS MARKET. There is a town in the middle of the desert named Las Vegas that has no other reason for existing except to let folks bet. It was almost entirely built based on folks losing money betting, at least partially on sports.

I advocate flat betting to make sure you are still enjoying the season AT THE END OF THE SEASON.

If you are not, then by all means go Kelly. It's your money. Burn it at any rate you want.

[skrtelfan]

98%+ of the public are long term negative EV punters, that's why. Everyone seems to be in denial about it.

Let me repeat that: 98% OF YOU GUYS WILL BE LONG TERM LOSERS IN THE SPORTS MARKET. There is a town in the middle of the desert named Las Vegas that has no other reason for existing except to let folks bet. It was almost entirely built based on folks losing money betting, at least partially on sports.

I advocate flat betting to make sure you are still enjoying the season AT THE END OF THE SEASON.

If you are not, then by all means go Kelly. It's your money. Burn it at any rate you want.

In that case, why not advocate betting the book's minimum, if one must bet at all? The optimal bet size if one is a losing bettor is $0, but if one must bet, the second most optimal bet size is the minimum.

[Nookx]

flat betting 0 = the ultimate winning loser?

[Inspirited]

This debate started in my head a few years ago and had nothing to do with sports betting. Actually it was blackjack. The Colorado casinos at the time had a $5 dollar limit. So at the blackjack table your min and max bet would be 5 dollars.

My argument on this w/ blackjack is if I was laying the same bet over and over I was just surrendering myself to the house edge. It's mathematically simple, betting the same amount time after time ensured the that standard deviation could never be compromised in the long run.

I just don't see how this is any different with sports betting if your always betting the same unit and the house always has a mathematical edge (juice.)

My conclusion is that no finite bankroll could ever last in the test of time with flat betting when there is ever a house edge.

If I understand the whole point behind counting cards in BJ it's so that you get your money in when the probabilities are in your favor. So, if the count was not in your favor, ideally you would not bet at all because the house would have an edge over you, but if the count is in your favor you should have an edge over the house. It's not the betting system that allows you to win in BJ, it's when you bet. I've never played live BJ, so I don't know how you handle all this at the table.

When flat betting in sports, you are not betting every event. You are only betting on events where you believe you have an edge, where the "count" is in your favor, where you have +EV opportunities. If you do truely have the edge you think you have, you should end up a winner over the long term.

Thus, I do not think you have proved that flat betting is a losing strategy in general.