Advice > Money Management
Honey - The First and Last Word on Money Management
Part Two - Lemme Tell You 'Bout Machine Gun Kelly...
There is hope.
The remarkable and brilliant Huey Mahl, a true handicapping
genius who passed on in 1996, wrote in his 1988 broadsheet,
Universal Money Management Plan, "Of all the money manipulation
schemes I've investigated over the years, I have found none
which could hold up to the vicissitudes of long-term play
as well as the properly controlled Proportional Bet
(emphasis is his), and not only optimize the return, but
protect your investment base from ruin. The mathematical
proof was developed by John Kelly, Jr., a Bell Labs engineer,
back in 1956. You've probably heard his treatise referred
to as the Kelly Criterion."
In Kelly's system
of betting, you always bet a fixed fraction of your bankroll.
The faction you choose is proportional to that which will
maximize its earning potential, or growth rate. In this
model, you bet 5% of your bankroll, with a starting total
of $2000.00. Your first bet of $100 wins, so you now have
$2100, so your second bet is $105. If that wins, your third
bet becomes $110.25. If, on the other hand, your first bet
loses, you would lose the $100, leaving you with $1900 operating
capital, and the second bet becomes $95. Nothing fancy here,
no astounding scores, but no shocking losses either. The
Kelly Criterion is built on the concept that the more you
have to bet, the more you'll risk, but you'll remain proportional
to a fixed percentage that will minimize your risk of ruin
and maximize your opportunities to win.
In his fine book
on casino and sports betting odds, Can You Win? Mike
Orkin writes that, "Statistician Leo Breiman showed that
the Kelly system is optimal for two reasons. First, it will
do better in the long run than any substantially different
strategy. Second, the expected number of bets necessary
to reach a specified goal with the Kelly System is lower
than with any other strategy."
The finest part
of this system is that it disciplines you to bet the same
percentage, thereby minimizing your chances of going broke.
And your bankroll has a decent opportunity of growing at
a modest but desirable rate. Betting 5.5% of your bankroll
on each bet, with a series of 100 bets, at a 55% win rate,
your bankroll would increase by 15%. Returning to our starting
bank of $2000, that would mean a gain of $300. Obviously,
betting at a higher percentage, or increasing your win rate
will significantly raise your growth rate. For example,
with a win rate of 65%, betting 26% of your bankroll, in
1000 bets, $2000 will grow to more than $50,000.
Is there a downside
to the Kelly Criterion? Of course (but you guessed that.)
This model works only with bets made offering a favorable
win expectation. Applied to situations where the house has
a built-in edge, Kelly goes right down the tubes. Huey Mahl
set the formula for this model as W % minus L%/ P (to $1)
where W%= the percentage of winners; L%= the percentage
of losers, and the denominator P (to $1) is the average
amount won per dollar.
The way Huey
figured it, you want to use the Kelly model in gambling
situations where the handicapper has distinct and favorable
odds. It is easy to see how this formula will eventually
wear down the bankroll with diminishing results in unfavorable
conditions. Using this system, you will never hit the big
one. But, the chances of a complete disaster are minimized
Those Low-Down Gamblin' Ways
term bleak enough to make anyone shiver. Send a real player
into the depths of despair. In fact this term is applied
to money management when, a player is involved with games
having a negative expectation, the smaller the bet in relation
to the size of the bankroll, the greater the chance of going
broke. An example might be if a gambler has a bankroll of
$1000 and he's looking to double his money; he decides to
divide the bankroll into units of 200 at $5 each. His chance
of going broke is 94.4%. I'd take the over he's going to
lose it all. If, on the other had he takes the whole shebang
and make one large bet, he has an almost even chance to
double his money.
holds true in situations where there is a positive expectation.
In this situation, the smaller the bet in relation to bankroll,
the better the chance to increase the money.
probability laws are based on large samples occurring over
a long duration. This is a reminder of the good old "due"
factor. When a team, or hand, or situation wins over a short
period of time, these unfortunate souls start betting against
it, because they think there is a short-term inevitable
opposite about to present itself. Therefore, rather than
minimizing a losing streak by stopping or lowering their
bets, they increase, expecting the opposite, based on the
"law of averages." Concurrently, instead of maximizing a
winning streak, they back down, fearing the worst. Mathematicians
call this "The Doctrine of The Maturity of Chances." Without
meaning to appear redundant: Each event of chance is
completely independent of everything preceding it or following
"When you make
a bet at less than the correct odds, which you always do
in any organized gambling operation, you are paying the
operator a percentage charge for the privilege of making
the bet. Your chance of winning has what mathematicians
call a "minus expectation." When you use a system, you make
a series of bets, each of which has a minus expectation.
There is no way of adding minuses to get a plus, or adding
losses to show a profit. "
"Add to this
the fact that all gambling operators, including race and
sports bookies, limit the size of the player's wagers so
it is impossible to double up bets indefinitely. This
and the house percentage make all gambling systems worthless."
This from long-time
casino expert John Scarne. He goes on to note that the only
exception is in games where skill and chance are involved
and this allows the player to make bets with a positive
expectation. If you have a power rating that indicates
a team should be the favorite, and the offered betting line
makes its opponent the favorite, you have value, referred
to as an overlay, and thus a positive expectation. In blackjack,
if you are a card counter, and the count goes plus, you
have a positive expectation. In poker, if you are aware
of your opponents' tendencies, you can use that to an advantage
given the right situations, for a positive expectation.
presented on this site is copyright to it's original author. Please
do not reproduce any content without permission.
If you have questions or comments about the site, please Contact