MJ: Setting a Target to Break Bank

[MJ]

Is there any disadvantage from the investor's perspective if a team sets a target to break bank once 50% of the starting bank has been won instead of waiting to double bank?

For example, if the investor was to put up 50k, the team breaks bank after 25k has been won. Doubling a bank requires a great deal of time, and setting a smaller target means both players and investors see a payday much faster. Moreover, the chances of hitting target also improve with this approach.

Beside a smaller ROI for the investor, what is the downside?

MJ

[Don Schlesinger]

> Beside a smaller ROI for the investor, what is the
> downside?

Why is it a smaller ROI? Such numbers have to be annualized to have much meaning. Does the investor not get the opportunity to "try again" with the next bank?

If you make $25K in, say, 50 days, doesn't that imply that, on average, you'd make $50K in 100 days? Annualized, the ROI is the same.

That isn't the problem. The problem is s.d. While e.v. is linear, s.d. isn't. So, because of the fluctuations, the probability of actually hitting the $25K mark in a specified number of days, is not the same probability of hitting $50K in double the number of days. This may sound like just what I said above, but it isn't. There's a difference between the mean and median number of days to hit these goals.

Have to think about it for a while.

Don

I'd have to play with the goal calculators that Norm created from the BJA3 formulas, but that will likely be the main difference.

Don

[MJ]

> Why is it a smaller ROI? Such numbers have to be
> annualized to have much meaning. Does the investor not
> get the opportunity to "try again" with the
> next bank?

If the investor puts up 60k and the payout is 50/50 player/investor, then if you play to a target of 60k within a year then the investor receives 30k/60k = 50% ROI once the bank is broken.

If the target is only say, 40k, now the investor only receives 20k/60k = 33% ROI once the target is reached. Of course, both of these examples assume no expenses are incurred for the life of the bank. So, the smaller the target, the smaller the ROI for the investor.

If at least a 15% ROI is not incurred after the first bank is broken, the investor would not reinvest.

BTW, what does it mean when the investors make a "margin call"? Is this the practice of pulling out their investment when things are going bad? There ought to be a some sort of penalty for this!

> That isn't the problem. The problem is s.d. While e.v.
> is linear, s.d. isn't. So, because of the
> fluctuations, the probability of actually hitting the
> $25K mark in a specified number of days, is not the
> same probability of hitting $50K in double the number
> of days.

I assume that the probability of hitting the 25k mark is greater. This is why I suggest setting a smaller target to break bank.

> This may sound like just what I said above,
> but it isn't. There's a difference between the mean
> and median number of days to hit these goals.

Is the median number of days going to be above or below the mean?

MJ

[Don Schlesinger]

> If the investor puts up 60k and the payout is 50/50
> player/investor, then if you play to a target of 60k
> within a year then the investor receives 30k/60k = 50%
> ROI once the bank is broken.

Which ought to take, on average, x amount of days.

> If the target is only say, 40k, now the investor only
> receives 20k/60k = 33% ROI once the target is reached.

Which ought to take 2/3(x) amount of days. ROI is not quoted in a void. To make sense, it has to be annualized. Would you like 33% return on your money in a month or a year? Does it make any difference to you? To most people, it would. (Of course, I'd be delighted with either these days! :-))

> Of course, both of these examples assume no expenses
> are incurred for the life of the bank. So, the smaller
> the target, the smaller the ROI for the investor.

See above. Not really true. If you allow me the opportunity to invest in serial banks, I might get my smaller return multiple times in one year.

> If at least a 15% ROI is not incurred after the first
> bank is broken, the investor would not reinvest.

What makes you break the bank, if the original goal isn't reached? Just time?

> BTW, what does it mean when the investors make a
> "margin call"? Is this the practice of
> pulling out their investment when things are going
> bad? There ought to be a some sort of penalty for
> this!

It isn't the investors who make a "margin call"; it's the players, who need more money!

> I assume that the probability of hitting the 25k mark
> is greater.

You can't hit a $50K mark without first hitting the $25K mark, now, can you? :-)

> This is why I suggest setting a smaller
> target to break a bank.

I understand.

> Is the median number of days going to be above or
> below the mean?

Below, I would imagine.

Don

[MJ]

> Which ought to take 2/3(x) amount of days. ROI is not
> quoted in a void. To make sense, it has to be
> annualized. Would you like 33% return on your money in
> a month or a year? Does it make any difference to you?
> To most people, it would. (Of course, I'd be delighted
> with either these days! :-))

I understand. You are taking into account the "time-factor" when annualizing.

But what about the 'probability factor'? Shouldn't that be considered as well? See below...

Suppose a team makes you an offer to up 60k, which can be doubled on average in 6 months. If you get 50% and do NOT reinvest profits on the next bank (but leave the principal), then that would be an annualized ROI of 100%. Sounds pretty sweet! But what if the probability of hitting target was only 60% for each bank? Now the probability of attaining that ROI of 100% is only 36%.

What if another team makes you an offer of putting up 40k, gives you 50% when the target of 20k is reached in what should take an average of 6 months. You will be given the chance to leave the principal in there for the next bank but cannot rollover profits. Again, ROI is 100% annualized. But let us say the probability of hitting target is 80% for each bank. Now the probability of attaining that ROI of 100% is 64%.

To determine which is the better investment, is it simply a matter of multiplying the 36% by 60k and comparing it to 64% x 40k?

In addition to the annualized ROI, you agree that the probability factor should be taken into account, right?

> What makes you break the bank, if the original goal
> isn't reached? Just time?

Correct.

> It isn't the investors who make a "margin
> call"; it's the players, who need more money!

Going back to the 60k bank, do you think it would make more sense for the investors to put up the capital in 1/3 installments on an as needed basis? This way, it could happen that they put up 20k initially and the bank goes upward so the remaining 40k is never needed. The investors could put that 40k in the bank earning interest rather than just sitting in the team bank not really being utilized.

It is also less anxiety for the investors putting up capital in installments rather than all at once.

MJ

[Don Schlesinger]

> I understand. You are taking into account the
> "time-factor" when annualizing.

Right.

> But what about the 'probability factor'? Shouldn't
> that be considered as well? See below...

> Suppose a team makes you an offer to up 60k, which can
> be doubled on average in 6 months. If you get 50% and
> do NOT reinvest profits on the next bank (but leave
> the principal), then that would be an annualized ROI
> of 100%. Sounds pretty sweet! But what if the
> probability of hitting target was only 60% for each
> bank? Now the probability of attaining that ROI of
> 100% is only 36%.

> What if another team makes you an offer of putting up
> 40k, gives you 50% when the target of 20k is reached
> in what should take an average of 6 months. You will
> be given the chance to leave the principal in there
> for the next bank but cannot rollover profits. Again,
> ROI is 100% annualized. But let us say the probability
> of hitting target is 80% for each bank. Now the
> probability of attaining that ROI of 100% is 64%.

> To determine which is the better investment, is it
> simply a matter of multiplying the 36% by 60k and
> comparing it to 64% x 40k?

You might want to take a look at BJA3, pp. 298-301. The question you are asking may be a bit more complicated than you realize.

> Going back to the 60k bank, do you think it would make
> more sense for the investors to put up the capital in
> 1/3 installments on an as needed basis?

Yes.

> This way, it
> could happen that they put up 20k initially and the
> bank goes upward so the remaining 40k is never needed.

Exactly. Greater ROI, because the capital required for success may not be as much as originally planned. The drawback is that, if you don't get the full $60K upfront, and the team starts out poorly, your investors may balk at putting up the rest of the capital when the "margin call" comes.

> The investors could put that 40k in the bank earning
> interest rather than just sitting in the team bank not
> really being utilized.

Absolutely. And, in the bank, these days, they could be earning, oh, at least 0.3% interest!! :-)

> It is also less anxiety for the investors putting up
> capital in installments rather than all at once.

Just make sure you get it in writing that the other installments will, in fact, be forthcoming. (In reality, there is, of course, no real way to force anyone to invest who decides not to.)

Don

[MJ]

If the investors are willing to put up 60k in 1/3 installments, do you think a target of 10k is reasonable or is this too small? Is there really any downside to setting a target which is essentially a small fraction of the "pseudo" BR? Again, this way players/investors will be compensated faster.

After that target is hit, pay outs are made and then the team does it all over again with the same target.

Now, one concern is keeping players/investors interested even when the BR is down. Suppose the BR goes south 20k early on. Investors are reluctant to advance additional funds and the players are discouraged b/c a payday is a long way off.

Do you think it would be unreasonable to set a target of -10k and then take 20% of the 10k win (2k) and split it 50/50 between players and investors? You might argue that this is just paying investors back with their own money because the BR is still in the red. But at least they are getting something in the short term. Same goes for when the BR goes from -10k to 0k.

Another idea if the BR is 20k down is to pay players a salary of 10% of EV as an advance for when the bank is broken. Assuming trip expenses are an additional 20% of EV, that is 30% of EV lost to expenses. Anyway, the salary would last until the BR was out of the red. The investors would be given 1k every time the bank moves 10k in the positive direction until the BR is 10k ahead, and then split 50-50 less the player salary advance. Both of these ideas will enhance ROR as we sustain the full brunt of losses but 'skim' some off the wins.

MJ

[Don Schlesinger]

> If the investors are willing to put up 60k in 1/3
> installments,

Again, they may be willing BEFORE the first 20K goes south. When you come back for the second 20K, they may have flown the coop. Get it in writing (and, even then, it isn't going to matter).

> do you think a target of 10k is
> reasonable or is this too small?

You don't mention the stakes you're playing for nor what your hourly e.v. or expenses might be.

> Is there really any
> downside to setting a target which is essentially a
> small fraction of the "pseudo" BR? Again,
> this way players/investors will be compensated faster.

Again, getting compensated "faster," but with a smaller amount, is not the idea. It's a bad deal for the investors. You split the win, but the investors bear the entire brunt of losses. So, you've now given away half of the 10K profits to the players, and now, you lose back all of the 10K and then some. Meanwhile, you haven't made a dime, but the players are sitting with half the profits. Now, you lose more. And, the investors lose back all of their profits, while the players have the cash and don't give a shit. Do you get the picture?

> After that target is hit, payouts are made and then
> the team does it all over again with the same target.

Sounds great, doesn't it? Just wait till you start losing. The whole thing falls apart. And, lose you will! I guarantee it.

> Now, one concern is keeping players/investors
> interested even when the BR is down. Suppose the BR
> goes south 20k early on. Investors are reluctant to
> advance additional funds and the players are
> discouraged b/c a payday is a long way off.

Welcome to the world of team play!

> Do you think it would be unreasonable

YES!!! (I already know the question!)

> to set a target of -10k

I assume you mean +10K.

> and then take 20% of the 10k win (2k) and
> split it 50/50 between players and investors?

No can do.

> You
> might argue that this is just paying investors back
> with their own money because the BR is still in the
> red. But at least they are getting something in the
> short term. Same goes for when the BR goes from -10k
> to 0K.

You won't get unanimity on these points. I'm not a fan of doing anything for the short run. If the players or investors require instant gratification, you're engaging in the wrong endeavor.

> Another idea if the BR is 20k down is to pay players a
> salary of 10% of EV as an advance for when the bank is
> broken.

Even less incentive for the players to "dig in." They get paid for losing!

> Assuming trip expenses are an additional 20%
> of EV, that is 30% of EV lost to expenses.

I never was a fan of paying players salary. They have to accept certain risks, too, as the investors do. One of those risks is, if you don't win, you aren't going to get paid. Consider it further incentive to want to do well. (Note that it is also a tremendous incentive to steal.)

> Anyway, the
> salary would last until the BR was out of the red. The
> investors would be given 1k every time the bank moves
> 10k in the positive direction until the BR is 10k
> ahead, and then split 50-50 less the player salary
> advance.

Many, many plans sound wonderful on paper, until you go for six months and you haven't won a dime.

> Both of these ideas will enhance ROR as we
> sustain the full brunt of losses but 'skim' some off
> the wins.

Not a good idea.

Don

[MJ]

> Again, they may be willing BEFORE the first 20K goes
> south. When you come back for the second 20K, they may
> have flown the coop. Get it in writing (and, even
> then, it isn't going to matter).

Yeah, well I can't really blame them. If somebody lost 20k of my money, I would be reluctant to give them more money. (I'm just trying to see it from the investors' perspective).

> You don't mention the stakes you're playing for nor
> what your hourly e.v. or expenses might be.

Well, I figure the WR/Hr = $100. The expenses per trip would be around $300-$400. Each trip would entail approximately 20 hours of play over a weekend.

> Again, getting compensated "faster," but
> with a smaller amount, is not the idea. It's a bad
> deal for the investors. You split the win, but the
> investors bear the entire brunt of losses. So, you've
> now given away half of the 10K profits to the players,
> and now, you lose back all of the 10K and then some.
> Meanwhile, you haven't made a dime, but the players
> are sitting with half the profits. Now, you lose more.
> And, the investors lose back all of their profits,
> while the players have the cash and don't give a shit.
> Do you get the picture?

I understand your point. You are concerned the investors might lose back their profits if the next bank goes south. Okay, so maybe 10k is too small of a target for a 60k BR. How about setting a target of 20k? Even if a target of 60k is set, the investors could still lose back their 30k profit on the subsequent bank if things go badly. Of course, the larger the target, the less chance of the investors losing it all back on the next bank.

> Sounds great, doesn't it? Just wait till you start
> losing. The whole thing falls apart. And, lose you
> will! I guarantee it.

This is why I feel incentives need to be properly lined up so as to keep players and investors in line, even when the bank is down. What motivation does a player have to keep playing if a bank is 20k in the red and they need to hit a 20k target before breaking bank? This 40k swing would require an average of 400 hours of play given the aforementioned hourly WR.

> You won't get unanimity on these points. I'm not a fan
> of doing anything for the short run. If the players or
> investors require instant gratification, you're
> engaging in the wrong endeavor.

It is just about motivating the team and finding the right balance in so far as player/investor compensation.

> Even less incentive for the players to "dig
> in." They get paid for losing!

Would you still be opposed to paying a salary if the salary is determined using that 3 page article in BJA3? I believe Griffin corroborated your solution to the question posed by the reader.

> I never was a fan of paying players salary.

What are your thoughts regarding paying investors dividends prior to reaching target? Suppose after every 100 hours of play, the investors receive a 10% cut (assuming there is a win) as an advance once the bank is broken. I am not sure if this arrangement would be fair to the players, though.

MJ

[Don Schlesinger]

> Yeah, well I can't really blame them. If somebody lost
> 20k of my money, I would be reluctant to give them
> more money. (I'm just trying to see it from the
> investors' perspective).

Well, then, it would be better to look for investors who understand that this is exactly what's going to happen, and that, if they don't like it, they shouldn't invest. (Ask them if 100% of their stock-market investments go up 100% of the time. Ask them how they did in 2008.)

> Well, I figure the WR/Hr = $100.

Kinda small for a joint bank, no?

> The expenses per trip
> would be around $300-$400. Each trip would entail
> approximately 20 hours of play over a weekend.

This is going to take a looooong time. With, maybe, $1,500 per hour s.d. (if you're lucky), weekend s.d. will be $6,700, compared to e.v. of $2,000. Add expenses, and you're going to lose 44% of the weekends. The natives are sure to get restless.

> I understand your point. You are concerned the
> investors might lose back their profits if the next
> bank goes south.

Not "concerned" -- certain of it!

> Okay, so maybe 10k is too small of a
> target for a 60k BR. How about setting a target of
> 20k? Even if a target of 60k is set, the investors
> could still lose back their 30k profit on the
> subsequent bank if things go badly. Of course, the
> larger the target, the less chance of the investors
> losing it all back on the next bank.

Your last sentence is the most reasonable one. The rest is just your wishing out loud that I'll change my mind! :-) (By the way, where have you been for the three weeks since I last posted? At this rate, this thread is going to last until Christmas!)

> This is why I feel incentives need to be properly
> lined up so as to keep players and investors in line,
> even when the bank is down. What motivation does a
> player have to keep playing if a bank is 20k in the
> red and they need to hit a 20k target before breaking
> bank? This 40k swing would require an average of 400
> hours of play given the aforementioned hourly WR.

As I said before, welcome to team play. You want a get-rich-quick scheme, play the lottery. Players like to whine a lot. Try to find non-whining players. (Not so easy to do. "Non-whining player" is almost an oxymoron.)

> It is just about motivating the team and finding the
> right balance in so far as player/investor
> compensation.

Team player/managers love to structure deals that are wonderful for the players and terrible for the investors -- to the point of practically stealing. A famous author/expert was the master at it. I take a slightly different point of view. I try not to victimize the investors. This famous author/expert was the Bernie Madoff of blackjack teams.

> Would you still be opposed to paying a salary if the
> salary is determined using that 3 page article in
> BJA3? I believe Griffin corroborated your solution to
> the question posed by the reader.

The solution and the math are fine. But, can you show me where I discussed salaries? I don't seem to remember that part. I remember formulas for breaking a bank earlier than the hours needed to double, in theory. It is highly doubtful that anything I write will contain salaries for players. You, of course, are free to structure a team in any manner that the players agree to.

I have had famous arguments with teams and highly respected managers who paid their winning players more than their losing players. I show, in BJA3 ("Random Walk") that this approach is, mathematically, an absurdity, but, hey, if all the players go for it, who am I to say no? If a team were to agree that all the losers should get paid more than the winners, and everyone is in favor, hey, God bless 'em! It's a free country -- or, at least, it used to be.

> What are your thoughts regarding paying investors
> dividends prior to reaching target? Suppose after
> every 100 hours of play, the investors receive a 10%
> cut (assuming there is a win) as an advance once the
> bank is broken. I am not sure if this arrangement
> would be fair to the players, though.

The players will tell you to give the same dividends to them. Ultimately, you have to realize the following: if you can get a group of investors and players to unanimously agree to a method, you are better than most. The best you can do is to think about it all very carefully, and hope to get the majority to go along with the logic and the plan. And, the more math you have to lend credence to your argument, the more likely the majority is to go along with you.

Of course, if your goal is to rip off investors who really don't understand the math and to whom you misrepresent the realities of your e.v. and s.d., simply with a mind to steal their money, then that's fairly easy, as well. I'm assuming that this isn't what you want to do -- at least voluntarily!

Don

[MJ]

> Well, then, it would be better to look for investors
> who understand that this is exactly what's going to
> happen, and that, if they don't like it, they
> shouldn't invest. (Ask them if 100% of their
> stock-market investments go up 100% of the time. Ask
> them how they did in 2008.)

You make it sound like there is 100% chance of losing that first 20k installment. Don't some banks start off positive and then never look back?

> Kinda small for a joint bank, no?

How much do you think one should earn per hour with a 60k BR?

Perhaps I am going about this all wrong. I initially devised a betting ramp that used a spread of 2x1 unit to 2x20 units. I capped it at 20 units because that is what a 60k bank would permit and maintain about a 5% lifetime ROR. But should I really be concerned with simple ROR when there are practical constraints in place?

In other words, rather then worry about simple ROR wouldn't it be more practical to just use the double barrier calculator to assess ROR? The life of the bank is 400 hours (short, I know) and the goal will probably be about 40k. So, say I recompute my ramp with a larger unit such that the simple ROR calculator now brings back 25% ROR but the DB calculator reports a 5% ROR. Is this a smart way to play it? A 25% ROR is almost double kelly, but then again I am not playing for eternity anyway.

> This is going to take a looooong time. With, maybe,
> $1,500 per hour s.d. (if you're lucky), weekend s.d.
> will be $6,700, compared to e.v. of $2,000. Add
> expenses, and you're going to lose 44% of the
> weekends. The natives are sure to get restless.

Good insight. Well, I figured a minimum of 50 SCORE. SD/Hr = $2k and SD/ Trip would be $9k. But you made your point. I will have to recompute my ramp for a larger SCORE. Just curious, what % of weekends do you think a team should theoretically lose after expenses?

> The solution and the math are fine. But, can you show
> me where I discussed salaries? I don't seem to
> remember that part. I remember formulas for breaking a
> bank earlier than the hours needed to double, in
> theory.

I stand corrected.

> The players will tell you to give the same dividends
> to them.

Alright, then how about after every 100 hours, players and investors each receive a 5% cut of the win as an advance on when the bank is broken, provided the bank is ahead overall. If the bank is down and there is a win for any 100 hour period of play, then no dividends. The problem is I am not sure how such arrangements would impact ROR.

> Of course, if your goal is to rip off investors who
> really don't understand the math and to whom you
> misrepresent the realities of your e.v. and s.d.,
> simply with a mind to steal their money, then that's
> fairly easy, as well. I'm assuming that this isn't
> what you want to do -- at least voluntarily!

I am not out to rip anybody off.

MJ

[Don Schlesinger]

> You make it sound like there is 100% chance of losing
> that first 20k installment. Don't some banks start off
> positive and then never look back?

None that I've been involved with. :-)

> How much do you think one should earn per hour with a
> 60k BR?

Depends on so many factors that it's impossible to answer. I'm sure you know all the variables.

> Perhaps I am going about this all wrong. I initially
> devised a betting ramp that used a spread of 2x1 unit
> to 2x20 units. I capped it at 20 units because that is
> what a 60k bank would permit and maintain about a 5%
> lifetime ROR. But should I really be concerned with
> simple ROR when there are practical constraints in
> place?

The spread is fine -- even a bit ambitious. 5% ROR might be OK for some, but high for others, if, by "bankroll," you really meant a non-replenishable amount that represented all the money in the world that you have to devote to BJ.

> In other words, rather then worry about simple ROR
> wouldn't it be more practical to just use the double
> barrier calculator to assess ROR? The life of the bank
> is 400 hours (short, I know) and the goal will
> probably be about 40k. So, say I recompute my ramp
> with a larger unit such that the simple ROR calculator
> now brings back 25% ROR but the DB calculator reports
> a 5% ROR. Is this a smart way to play it?

A bit risky, to me. Maybe not to others.

> A 25% ROR is almost double kelly,

Kelly ROR is, theoretically zero. You mean starting with full Kelly and never resizing.

>but then again I am not playing for eternity anyway.

So, what happens after 400 hours, when you have, say, broken even? Aren't you going to just keep playing? So, do you see why simple ROR may be important after all?

> Good insight. Well, I figured a minimum of 50 SCORE.
> SD/Hr = $2k and SD/ Trip would be $9k. But you made
> your point. I will have to recompute my ramp for a
> larger SCORE. Just curious, what % of weekends do you
> think a team should theoretically lose after expenses?

If you mean 20 hours, just use s.d., e.v., and do the math.

> Alright, then how about after every 100 hours, players
> and investors each receive a 5% cut of the win as an
> advance on when the bank is broken, provided the bank
> is ahead overall. If the bank is down and there is a
> win for any 100 hour period of play, then no
> dividends. The problem is I am not sure how such
> arrangements would impact ROR.

What happens, after receiving the advance, if you lose it all back?

Don

[MJ]

> So, what happens after 400 hours, when you have, say,
> broken even?

At some point you have to call it a day. When that point comes, then liquidation of bank occurs and investors are reimbursed in direct proportion to how much they invested.

So my idea of using the DB calculator could work provided I am conservative in estimating the constraints (larger goal and time than may be necessary).

Speaking of the DB calculator, what do you think about reducing WR by a certain amount to take into account hourly expenses, camo, salary, playing errors, etc and then using that adjusted WR in the DB calculator to assess ROR? Could this provide a reliable estimate?

MJ