Which is the best bankroll management?

[the Great Gatsby]

discuss

[Waterstpub87]

Flat bet. Listen to KVB. 1% of BR.

[TommieGunshot]

https://www.sportsbookreview.com/betting-calculators/kelly-calculator/

The proof can be found here: https://www.princeton.edu/~wbialek/r...s/kelly_56.pdf

[veriableodds]

Flat bet. Listen to KVB. 1% of BR.

Concur 1% is just fine and should be ok. The way I look at it is over exposed one may want to look at it a little different. say you are a value volume investor, well a shit load of games tend to come. Your list gives 11 totals, 11 sides, and a lot of wagers in one sport, or specific market like (spreads, totals, ect) you might want to reduce each wager to a reasonable risk. Flat wagering is the way to go also

[Waterstpub87]

I get the arguments for Kelly. It is one of those things like looks really clean on paper.

Not going broke is a really important part. If you have a 5 or even 6 figure bankroll, yes, it can be replaced, but its going to hurt. Your chance to go broke is significantly reduced betting a lower percentage.

Also, you have to consider the fact that you don't know your exact edge for that bet. A 1% edge on average can easily translate to a range of edges that are anywhere from -2%+ to 3%+, your edge is not uniform. Terrible situations can arise from thinking you are a 3 and being at -2 with kelly betting. It adds volatility, at a significant level.

For example, I have beaten college football totals at .55 points vs the close this season. I normally bet Sunday nights, for next Saturday. Many things can happen in between. The most of I have beaten a total by this season is 8.5 points. The most its moved against me is 6 points. Depending where I had it, with kelly criterion, this could end up with me betting 3x as much on the -6 vs the 8.5.

You cannot calculate your edge till the closing line on a major sport. You cannot calculate the individual edge on a minor sport/prop bet, as the close is not efficient. You can only calculate your average edge over time. If you are betting a non-closing line, ie an opener/the night before ect., there is information yet to be incorporated that you don't have. You cannot, therefore, know your edge at the moment.

Really, it comes down to preference. Are you the type who likes 10% Return, 20% Vol or 8 Return and 6% vol? Thats really the answer.

[StackinGreen]

Whats KVB or your best guess on how to calculate edge (he says law of large numbers)

[TommieGunshot]

I get the arguments for Kelly. It is one of those things like looks really clean on paper.

Then you have a much stronger ability to digest dense academic writing than most people. I find Kelly Criterion to be an eloquent formula to something that is extremely messy on paper.

Not going broke is a really important part. If you have a 5 or even 6 figure bankroll, yes, it can be replaced, but its going to hurt. Your chance to go broke is significantly reduced betting a lower percentage.

Half-Kelly solves that.

Also, you have to consider the fact that you don't know your exact edge for that bet. A 1% edge on average can easily translate to a range of edges that are anywhere from -2%+ to 3%+, your edge is not uniform. Terrible situations can arise from thinking you are a 3 and being at -2 with kelly betting. It adds volatility, at a significant level.

If the average of all bets is still +1%, and the range is known, Kelly can still be applied. And fractional kelly makes it so those terrible situations are about as likely as any other staking method, which more growth.

For example, I have beaten college football totals at .55 points vs the close this season. I normally bet Sunday nights, for next Saturday. Many things can happen in between. The most of I have beaten a total by this season is 8.5 points. The most its moved against me is 6 points. Depending where I had it, with kelly criterion, this could end up with me betting 3x as much on the -6 vs the 8.5.

This might be a case of doing a better job of estimating odds. Why are there so many questions about bet sizes and "bankroll management", but so few about finding good bets? How much of an advantage is an extra 0.5 points against college football totals?

You cannot calculate your edge till the closing line on a major sport. You cannot calculate the individual edge on a minor sport/prop bet, as the close is not efficient. You can only calculate your average edge over time. If you are betting a non-closing line, ie an opener/the night before ect., there is information yet to be incorporated that you don't have. You cannot, therefore, know your edge at the moment.

All the available information can be used to determine the edge at the time of the bet. The information might change and the value of the bet might change, and if the bettor ends up with no edge over time, then it goes back to needing to do a better job at estimating probabilities, not bet sizes.

[Waterstpub87]

I like your response Gunshot.

Yes, have to read a lot of math papers involving finance. Similar situations.

If your half kelly leads you to be 1% of your bankroll, sure. This would be a case of how you applied it. If you end up with an average over that your risk of going broke is still larger than flatbetting. You can call it quarter kelly, or whatever you'd like, your risk is higher betting more and different amounts.

I disagree. Like I said, you are adding higher degrees of variance.

I'm not concerned with finding better bets. I don't have questions about bet sizing. I know my preferred staking method. Unfortunately, that depends on what the closing total is. I've done very well at college football totals over the past few season.

This is false. For a particular game, you don't know the final edge that you will end up with. Injuries happen, and sometimes, you are just wrong. Fundamentals change as well.

Your argument is also circular. There are a lot of assumptions in there, like "I know my bet has an edge, and I know exactly what it is, and if I'm wrong then I need to make better bets". You don't know the second part ahead of time. Rather than averaging that uncertainty by flat betting, one leaves oneself open to compounding this uncertainty.

As I said up above, its personal preference. Highest growth rate isn't always the most important thing.

[TommieGunshot]

This is false. For a particular game, you don't know the final edge that you will end up with. Injuries happen, and sometimes, you are just wrong. Fundamentals change as well.

I know what my edge is, based on all available information, at the time of the bet. If I bet under 52.5 on Tuesday night, and at that time, all the information says there is a 54% chance of the game going under, then I have a 3% edge. Even if that 54% changes between Tuesday and Saturday, it was still a 3% edge at the time of the bet.

[StackinGreen]

What includes "all the information "?

[Waterstpub87]

I know what my edge is, based on all available information, at the time of the bet. If I bet under 52.5 on Tuesday night, and at that time, all the information says there is a 54% chance of the game going under, then I have a 3% edge. Even if that 54% changes between Tuesday and Saturday, it was still a 3% edge at the time of the bet.

No, You don't. The true odds of the game begin at kickoff. They include the final rosters, game conditions, ect. You are buying a future probability that is reflected at game time. You precieve your edge will be 3% at the close. It is unknown when you bet it. You believe it, you don't know it.

If your belief, long term, across many bets, leads you to profit then you tend to be more right. However, it will consist of many times you were wrong. There is a variance in your degree of rightness.

[TommieGunshot]

No, You don't. The true odds of the game begin at kickoff. They include the final rosters, game conditions, ect. You are buying a future probability that is reflected at game time. You precieve your edge will be 3% at the close. It is unknown when you bet it. You believe it, you don't know it.

If your belief, long term, across many bets, leads you to profit then you tend to be more right. However, it will consist of many times you were wrong. There is a variance in your degree of rightness.

Are you really suggesting that there is no way to know before Saturday morning if Alabama or New Mexico State is more likely to win this week?

[Waterstpub87]

Are you really suggesting that there is no way to know before Saturday morning if Alabama or New Mexico State is more likely to win this week?

This is intentional misreading. I'm not going to go down this road.

If your belief is that the edges you calculate are exact and eternal, no matter what happens, then bet away with your kelly. This is faith, this not rooted in anything in the real world. You can't argue with faith.

But, for the sake of argument, pretend that not everyone is a super sharp. In that case, betting more is worse. Kelly causes you to bet more. Therefore flat betting is better, again for the normal folks.

[Vyasports]

This is intentional misreading. I'm not going to go down this road.

If your belief is that the edges you calculate are exact and eternal, no matter what happens, then bet away with your kelly. This is faith, this not rooted in anything in the real world. You can't argue with faith.

But, for the sake of argument, pretend that not everyone is a super sharp. In that case, betting more is worse. Kelly causes you to bet more. Therefore flat betting is better, again for the normal folks.

Good point. You've done your job here, if people disagree then let's just

[roanildinho]

kelly is good for going broke

[oilcountry99]

winning

[TommieGunshot]

If your belief is that the edges you calculate are exact and eternal, no matter what happens, then bet away with your kelly. This is faith, this not rooted in anything in the real world. You can't argue with faith.

Things obviously do change. But the way they change will always be just as likely to add to the advantage as it is to take away from the advantage. And the overall average edge will approach whatever was calculated at the time of the bet. If that wasn't true then the edge was never there to begin with and Kelly would again be the best way to bet, because of exactly what you said here:

But, for the sake of argument, pretend that not everyone is a super sharp. In that case, betting more is worse. Kelly causes you to bet more. Therefore flat betting is better, again for the normal folks.

If there is no edge, Kelly would say to bet zero. How is it possible that is betting more than "flat betting"?

[Waterstpub87]

You don't seem to understand the concept that your edge that calculate is likely not your actual edge.

1. if it is your average edge, there is variance to your edges. This can be caused changes in condition between your bet and gametime, as well variables you may missed that effect the actual odds of the game. If you beat the close 100% of time, then ok, this isn't true for you.

2. If its historical, things may change, and your past edge is no longer your current edge

Both of these are sources of risk that add variance to your edge.

Flatbetting manages your risk, by averaging your edge over time. Kelly adds variance by increasing/decreasing your bet. Max growth is not the only factor. Not going broke is important too. Unless your average bet is less than 1%, you will be betting more.

If you have zero edge, most people, is it better to bet more or less? Therefore, the system that has most people betting less is the better one.

[TommieGunshot]

You don't seem to understand the concept that your edge that calculate is likely not your actual edge

If it is close, but not exact, that is good enough for me to earn profit. And by betting more when my estimated edge is higher, I have more profit. If the edges that I calculate are not close at all, could you explain any errors I am making?

If you have zero edge, most people, is it better to bet more or less? Therefore, the system that has most people betting less is the better one.

And with zero edge, Kelly says to bet zero. If betting less really is better, cannot get any better than that.

[OldBill]

wise guys and sharps use 2% rule max is 5% which we can do today locally minmum bet is $5

get lucky win 3 ML dog parlays $5 gets you hundreds like past week 14.5 dog 2.5 dog and 7.5 dog 6 point dog and 10 point dog all won the game s/u 14.5 dog would pay +850 i think 10 dog + +740

[Dancho032]

im using 1 to 5 % for single bets ,depending of the value ,the bigger the value the bigger the stake !

[TheLock]

Bankroll management

The 2 words that 85% of all bettors can’t even spell let alone adhere to

[the Great Gatsby]

bump

[OldBill]

as proven over 50 years or more by the wise guys sharps that wager for a living 2% of B R on any game

PERIOD ............... this equates to $2 per $100 so if you have say $50,000 cash to gamble with YOU can bet $1000 per game in nfl on sunday with a top out of 4 games flat bet NO P lays or other gambling bets straight flat bets only

sure some times i go perfect ats on 10 games or more but to bet all of them would be nuts there is always that one game that just jumps off the page at you so take the total you were going to bet on say 5 games and bet just that one

so many goofy upsets every week in NFL worng team favored

[Jimmy Beamish]

double down after a losing bet

[juicername]

double down after a losing bet

This. No one loses two in a row, and certainly not three!!