[The Philosopher]

Well, I can’t let you all beat up on my buddy Turk any more. So I’ll offer a defense of his position of scalping/middling out of an NAB-type book. I can offer only a very qualified and limited such defense, but if you read closely his posts and the posts of those on his side of this issue, I don’t know that they are claiming anything much stronger than that for the strategy.

Let’s look at some potential reasons in favor of this strategy:

1. Assume you have $1,000 in your account at a book in trouble, where there is some small but nonzero chance that you will eventually be paid what you have coming to you. For the sake of argument, let’s say you estimate that there is a 25% chance they’ll bounce back and pay their customers.

You shop lines and find a game that is even money on one team at this book, and even money on its opponent at another book.

If you stand pat, there is a 75% chance you will have $1,000 stolen from you. If you bet $1,000 at even money on each side of this game at these two books, there is a 50% chance you will have eliminated your balance at the shaky book while boosting your balance at the legit book by $1,000, and a 50% chance you will have doubled your balance at the shaky book while losing $1,000 at the legit book. Which is to say, you will have a 37.5% chance of having $2,000 stolen from you.

Now, what’s worse? A 75% chance of being cheated of $1,000, or a 37.5% chance of being cheated of $2,000? Would you trade the one for the other?

I contend that while in a mathematical sense they are of equal value—both equivalent to a loss of $750—their actual utility can differ from person to person and from situation to situation. Depending on your bank roll, your individual psychological make-up and risk aversion, etc., you may well prefer putting yourself at a smaller risk of losing a bigger amount, or a bigger risk of losing a smaller amount. And I don’t think such a preference is “wrong.” If the one has more utility for you than the other, then so be it. There is no logical requirement that the impact on your life or your psyche of losing $2,000 has to be precisely double that of losing $1,000.

Also, for most people I would think the opportunity for closure would have some nonzero weight. If you stand pat, you will be stuck in a position of uncertainty for whatever indeterminate time the book teeters. But if you make the play, you have a 50% chance of remaining in that position of uncertainty and a 50% chance of being able to walk free of the book and never worry about it again.

So even if your subjective weighing of the utilities renders a 75% chance of being cheated out of $1,000 and a 37.5% chance of being cheated out of $2,000 as being of exactly equal value, it might still be your best course of action to choose the latter due to the opportunity it provides of potentially settling your fate and ending this whole unpleasant uncertainty.

(Clearly the math will be different with other hypotheticals, but the principles will be the same. For example, let’s say your strategy is to bet $1,000, then if you lose, bet $2,000, then $4,000, then $8,000, and that you’ll stop if your account builds to $16,000. In that case, you would be exchanging a 75% chance of being cheated out of $1,000 for a 4.6875% chance of being cheated out of $16,000. There again the utilities of these two outcomes for you need not be exactly equal.)

2. A player should also take into account if different level balances or different betting histories might have some impact on his likelihood of being paid.

Above I was assuming a 75% chance of getting stiffed across the board. But it’s possible that this percentage will vary at different levels. Depends on what evidence happens to be available concerning a given book, but it's possible.

For instance, let’s say your book has shown a propensity to pay off customers owed only a small amount, but is stalling on even partial payments to those with a large balance. Maybe there’s a 75% chance you’ll never see your money if your balance stays at $1,000 or increases, but if you can get it down to $500 you figure there is a 50% chance they’ll be willing to give you that to have one less person harassing them, posting in the forums, etc.

Now that needs to be a part of your calculations. If you stand pat, you still have a 75% chance of being cheated out of $1,000. If you make a $500 wager, you have a 37.5% chance of being cheated out of $1,500, and a 25% chance of being cheated out of $500. Your choices are no longer mathematically equal, as the latter is equivalent to a loss of $687.50 rather than $750.

Not every situation is the same, and this very factor can cut precisely the opposite way given a different set of evidence instead. You may, for instance, be dealing with a book that has displayed a grudging willingness to pay off some $1,000 balances, but no higher. In that case, your risk may well increase if your balance grows, but not correspondingly decrease if you bring your balance down.

But the point is, in some cases—not all—it can skew the expected return in favor of the strategy of trying to scalp/middle out of the shaky book.

Similarly, your chances of being paid might be affected by your willingness to give the book action. For evidence of this, you could consult with other people who have and haven’t been paid by this book since they became a problem, and see if there is any correlation where the book was more amenable to payout requests from folks who were “sporting” enough to continue to play into them. It also could be worthwhile to look at past cases like the Dunes, a book that staggered along in slow pay mode for months or a year or more before finally going under. During all that time that partial payments were dribbling out, was there any correlation between those who continued to give them action and those who were sent at least some of the money they were owed?

If so, and if you’re in a similar situation, then again the percentages will not remain constant and you need to adjust your calculations.

3. This is not so much an additional factor, but I want to address a factor that seemingly cuts the other way.

As some posters have pointed out, often one of the reasons that a book, including NAB, gets in trouble is precisely that its lines are weak enough to be on the wrong side of scalps more often than chance. In that case, attempting to scalp/middle your way out would actually put you in a slightly worse position mathematically, and you would have to take that into account in deciding whether to use the strategy. So, in my earlier example, instead of exchanging a 75% chance of being cheated out of $1,000 for a 37.5% chance of being cheated out of $2,000, it might be more like a 40% chance of being cheated out of $2,000.

I think this objection can be eliminated if we more carefully specify the strategy (and this is consistent with how Turk and others have described what they have in mind).

The strategy would not be to take whatever scalps or middles happen to be available, but to take only a certain subset of them. That is to say, a bettor would bet precisely what he would have bet at his legit books even if he had no account at all with the shaky book. Then and only then, he would bet the other side of those plays at the shaky book where doing so provides a scalp or middle.

Now, that subset of scalp/middle opportunities will be null if lines at the book in question are so consistently weak that as a handicapper you take this factor to be so overwhelming that you literally will never make a play at another book that doesn’t either fade them or is neutral. But except in this unrealistically extreme case, there should still be at least some occasions where you like the same side of a game that the shaky book does, and thus this strategy can come into play after all.

4. I am ignoring the “evidence” of “But I’ve been successful with this strategy, therefore I know it works” (or, for that matter, "I tried this strategy and just dug myself into a bigger hole") because it carries only infinitesimal weight. If starting tomorrow, everyone makes a note of which plays they made with the phone in their left hand, I have no doubt that some players will discover that they won at a rate higher than chance on those plays. This basically tells me nothing about whether it is a good strategy to hold the phone in one’s left hand when wagering.

5. In conclusion, I disagree with those critics who frantically warn against this strategy and label it as irresponsible, “the worst advice you could give,” etc. If one is following the version of the strategy where you only bet the opposite sides of plays that would have been warranted on their own merits at the legit books anyway, then one of two things. Either a) None of the factors mentioned above happens to be present, and so you are engaged in a pointless exercise that neither helps nor harms you, except for a trivial loss of time making the extra plays. (It is the “mind games” Ronbets mentions, or the equivalent of making meaningless plays on paper at a nonexistent book as several described.) Or b) One or more of the factors mentioned above is present, and so you are to some small degree better off pursuing this strategy than standing pat. To contend instead that this is some recklessly harmful strategy smacks of superstition, like “Don’t mention the pitcher has a no-hitter until the game is over,” or “Don’t record your bet as a winner before its over.” I don’t think the gambling gods will punish your insolence for trying to scalp/middle out of a book. If used correctly in the way I’ve described, the strategy will either be neutral or very slightly to your advantage.

Perhaps think of it as a kind of “rebate on losses” promotion at your other books. If you win your plays at your other books, then your account at the shaky book is eliminated and that’s that. Nothing to rebate, because you won your “real” bets. But if you lose your plays at your other books, then you have at least some chance of recouping that loss, as long as there is a nonzero chance the shaky book will pay you after all.

So, it’s not “Get 25% of any losses rebated to you,” but more like “Get a 25% chance of having 100% of your losses rebated to you, and a 75% chance of having none of your losses rebated to you.” If you’re just making the same plays at your legit books that you would have made anyway, how would a promotion like this hurt you?