Thread Rating:

MathExtremist
MathExtremist
  • Threads: 88
  • Posts: 6526
Joined: Aug 31, 2010
April 15th, 2016 at 4:55:41 PM permalink
Quote: Steen

You seem to be stuck on the EV of individual passline bets rather than the EV of a players total action. It's a player's bottom line that he cares about at the end of his session - not the EV of his individual bets.

You measure bottom line in dollars, not in percentages. The bottom line is that if two players stand side-by-side at a craps table, one betting $10 pass with no odds and the other betting $10 pass with full odds, their expected losses are identical. My point is that it's misleading to suggest that one player expects to lose less because they take odds, but that's what the naive gambler would understand by looking at 1.41% vs. 0.37%. The odds bet is not a player-favorable wager and has no impact on the expected loss of the prerequisite pass bet.

Quote:

Part of the difficulty in cutting through the chaff of misunderstanding here is that players often try to compare flat bets alone to flats+odds without taking into account commensurate risk. For example, they might try to compare $10 flats to $10 flats + odds. Rather than believe the math which clearly shows the average percentage lost per dollar wagered to be lower with odds, they instead look at the actual dollars lost and (seeing that it's the same with or without odds) they come to the erroneous conclusion that adding the odds bets did absolutely nothing for them. But that's not a fair comparison.

Fair or not, it's the relevant comparison. It's all well and good to point out that the expected loss between someone betting $10 plus odds whenever they can vs. $50 with no odds is 5x smaller. That's trivially true, but it's also not the choice presented to a player by a dealer who, after a point is established on the player's line bet, asks "do you want odds behind that?" In that case, it's important for the player to understand that they get a chance to make another bet on the same outcome, but that the new bet does not actually decrease their expected loss. Contrast with a proper double-down in blackjack, which *does* decrease the expected loss compared to hitting.

The way we calculate the optimal house edge for strategy games like blackjack or VP is to consider each strategy choice and examine the expectation in units, conditioned upon having the cards given. In the case of 11 vs. dealer 7, doubling is better than hitting -- it leads to a greater gain in units (0.46 vs. 0.29) even if it requires extra wagering to achieve that gain. However, if you were to look at the gain divided by the total action, which is 0.46/2 vs. 0.29/1, you'd incorrectly conclude that hitting 11 vs. 7 was better than doubling. It's not. When you evaluate the unit gain correctly, doubling is the obvious correct choice because investing the extra unit yields an extra 0.17 average units returned. In craps, however, when presented with the opportunity to make an odds bet after a point is established, there is no difference in unit gain whether you take odds or not. The expectation is identical. The only thing that changes is variance. That's why craps is not a strategy game -- there is no point during the life of any craps wager where a player has a choice that can increase the gain on that wager. The only thing you can do is make poor choices by removing established don't pass bets. No other bets on the craps table have a decision point at all.

Quote:

It's absolutely not incorrect and not misleading. Whether a $10 player has read about the house edge or not, he knows he's risking a lot more when he pushes out that extra $50 for odds. What's misleading is when someone tries to get him to believe that the extra $50 doesn't count toward his expected outcome.

I know you don't believe that making a $50 odds bet has a non-zero expectation, so what do you mean by "doesn't count"?

Quote:

Businesses are interested in making money. They want to sell as much product as they can with the best margins they can. They want to MAXIMIZE profit. Craps players are interested product too - it's called action, and they want to get it for as little cost as possible - they want to MINIMIZE the casino's profit.

We both know that's not true for all craps players uniformly, or gamblers in general. You can argue that the cost of gambling should always be measured as a percentage of total action but clearly not everyone agrees, and for those gamblers who have different priorities, ignoring other measurements is foolish. Certainly there are many players that are somewhat risk-averse, as are people in other aspects of life. Insurance wouldn't exist if everyone wanted to maximize their expectation and didn't care about variance. The whole point of an insurance contract (bet) is to trade expectation for reduction in variance. That's exactly what hedging a line bet with an any-craps on the comeout does.

I think the real issue is that basing a pricing evaluation on the idea of total action isn't relevant to the behavior of actual gamblers. Measuring total action and playing based on some fixed goal for that action is simply not the way most gamblers play. APs trying to meet bonus/promo playthrough requirements do, but otherwise people play based on time or win/loss goals. I have never in my entire life seen someone tally up their total action at the craps table and say something like "okay, I've wagered a total of $1000, time to go" regardless of whether it was in the middle of a hand or ow long they'd been playing. People simply don't play like that. My goal is to measure the price of casino gambling based on how gamblers actually gamble, which is almost always based on time. To me, measuring expected loss over that time seems more relevant.

Suppose two gamblers play for an hour. Bob has an expected loss of $30 with a house edge of 3%, while Fred has an expected loss of $20 with a house edge of 4%. It can be argued that Bob got a better deal because he had a lower expectation as a percentage, but it can also be argued that Fred got a better deal because he had a lower expectation in dollars. Who's right?

If your answer isn't "it depends," you're wrong. :)
"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice." -- Girolamo Cardano, 1563
Steen
Steen
  • Threads: 0
  • Posts: 126
Joined: Apr 7, 2014
April 15th, 2016 at 8:13:25 PM permalink
Quote: MathExtremist

You measure bottom line in dollars, not in percentages. The bottom line is that if two players stand side-by-side at a craps table, one betting $10 pass with no odds and the other betting $10 pass with full odds, their expected losses are identical.



Nonsense. Percentages are vital. Knowing the bottom line in dollars is meaningless if you don't know what it took to produce it.

For example, let's say you have money invested in a few different accounts. Your broker/banker calls and says, "Good news, you earned $100."

You're ok with that? Don't want to know anything else? Just the bottom line? Wouldn't you like to know which account earned the money and compare it to the amount you had invested to learn your rate of return and then compare that to your other accounts???

What if you found out that $50 was from an account containing $1000 and $50 was from an account containing $100,000 and nothing was returned from two other accounts containing $10,000 each? Are you still happy with your "bottom line" or are you going to think about moving money to the account with the better return?

Efficiency is always measured as a percentage. When a percentage is not given it's because the baseline is known and you can figure the percentage yourself. For example, a banker might give a sample of possible returns based on a $10,000 investment.

You can't ignore the math. Additionally, the sim I showed demonstrated that a player taking odds loses far less both as a percentage AND in actual dollars. You can't keep snapping back to the assertion that the dollars lost are the same and ignore the amounts wagered.

Quote:

My point is that it's misleading to suggest that one player expects to lose less because they take odds, but that's what the naive gambler would understand by looking at 1.41% vs. 0.37%. The odds bet is not a player-favorable wager and has no impact on the expected loss of the prerequisite pass bet.



I guess I have to keep harping on this -- it doesn't impact the expected loss on the flat bet, it impacts the total dollars that the player gambles.

Quote:

Fair or not, it's the relevant comparison.



Not when one is analyzing the situation and trying to discover the truth. When you're standing next to a guy at the table and he asks you if he should take odds, I can understand that you might blow him off and just say it's not relevant. But when you're here in the forum, where people are discussing the nuances of the game, trying to analyze what's happening, and trying to figure out how to improve their play, then making fair comparisons is vital to answering these questions.

Quote:

It's all well and good to point out that the expected loss between someone betting $10 plus odds whenever they can vs. $50 with no odds is 5x smaller. That's trivially true,



Ahhh, well at least you admit it's true! :-) Now we just have to work on the "trivial" part!

Quote:

but it's also not the choice presented to a player by a dealer who, after a point is established on the player's line bet, asks "do you want odds behind that?" In that case, it's important for the player to understand that they get a chance to make another bet on the same outcome, but that the new bet does not actually decrease their expected loss.



Oh yes, I see ... it's much better to let that player wait until the next come-out roll and make another line bet, right?

Let's say a player has $100 in his rack which he intends to gamble. He makes a $10 line bet, rolls a 5 and asks you, "Should I take odds?" You might begin by asking what's his objective? Play time? Loss per hour? Free drinks? etc, etc. He says, "Naw, I don't care about all that stuff. I just want a fair shot. My buddy told me the house takes a cut out of all the action and I just want to minimize that cut. I want as close to a fair game as I can get." What's your answer? Wait for the next come-out roll my son? Bet a 3-way craps, grasshopper?

<added> The point here is that he has another $90 he's intent on gambling so forgoing the free odds in lieu of something else will cost him.

Quote:

We both know that's not true for all craps players uniformly, or gamblers in general. You can argue that the cost of gambling should always be measured as a percentage of total action but clearly not everyone agrees, ...



You're mischaracterizing what I'm saying. How each person values his gambling experiences is completely personal. Whether rational or not, it's for that person to decide. That doesn't mean he need be ignorant of the math and the expectation. Expectation is just a tool, one of many.

Quote:

I think the real issue is that basing a pricing evaluation on the idea of total action isn't relevant to the behavior of actual gamblers. Measuring total action and playing based on some fixed goal for that action is simply not the way most gamblers play. APs trying to meet bonus/promo playthrough requirements do, but otherwise people play based on time or win/loss goals. I have never in my entire life seen someone tally up their total action at the craps table and say something like "okay, I've wagered a total of $1000, time to go" regardless of whether it was in the middle of a hand or ow long they'd been playing. People simply don't play like that.



Hahaha. You've never seen anyone tally their action? But you said earlier that some people play certain amounts for certain time periods. You suggested that a video poker player might wager $1.25 per hand at 400 hands/hr. Well guess what? That represents $500 worth of action/hr. Players experience action through their betting, the number of bets they make and the amounts they risk. Whether they know the exact value of their action or not is immaterial -- they're playing for action.

By the way, have you conducted a world-wide poll to determine how all people play? Just because some may be ignorant of, or intentionally disregarding of the facts doesn't make odiousgambit's original suggestion wrong, specifically that a $50 passline player would lower his EV by betting $10 flat with $40 odds.


Quote:

My goal is to measure the price of casino gambling based on how gamblers actually gamble, which is almost always based on time. To me, measuring expected loss over that time seems more relevant.



Again, a noble goal, but not the issue at hand. The idea here was to CHANGE the way gambler's actually gamble by educating them as to the facts.

Steen
Last edited by: Steen on Apr 15, 2016
RS
RS
  • Threads: 62
  • Posts: 8626
Joined: Feb 11, 2014
April 15th, 2016 at 9:36:30 PM permalink
Steen, your sim showed nothing useful. It shows that if you make fewer wagers, you lose less money. You should run a sim where both players play for X hours or Y rolls or something like that, then look at results.

Ya may as well compare playing $100,000 coin in on 9/6 JOB with coining in $15,000 on 8/5 BP....then say 8/5 BP is better!
Steen
Steen
  • Threads: 0
  • Posts: 126
Joined: Apr 7, 2014
April 15th, 2016 at 9:51:42 PM permalink
Quote: RS

Steen, your sim showed nothing useful. It shows that if you make fewer wagers, you lose less money. You should run a sim where both players play for X hours or Y rolls or something like that, then look at results.

Ya may as well compare playing $100,000 coin in on 9/6 JOB with coining in $15,000 on 8/5 BP....then say 8/5 BP is better!



Then you need to look at it again. Both players are wagering the same total amount.

On the table of Bet Results, note that the Total $ of wagers for each player is basically 5 million (within a few hundred bucks of each other) and yet the net loss for Player1 is almost 5 times greater than Player2.

Player2 made fewer wagers but the average wager was almost $22 (versus $10 for Player1) and so the total amounts wagered are the same.

If you'd rather show both players at the table for the same number of rolls or resolutions that can easily be accomplished but then we'll have to adjust the amounts bet to commensurate the risk. As Sally mentioned, $50 flat would compare nicely with $10 flat and $60 odds or $30/$30 depending on your preference and odds allowed.

Steen
MathExtremist
MathExtremist
  • Threads: 88
  • Posts: 6526
Joined: Aug 31, 2010
April 15th, 2016 at 10:31:52 PM permalink
Quote: Steen

Nonsense. Percentages are vital. Knowing the bottom line in dollars is meaningless if you don't know what it took to produce it.... Efficiency is always measured as a percentage. When a percentage is not given it's because the baseline is known and you can figure the percentage yourself. For example, a banker might give a sample of possible returns based on a $10,000 investment.

That's true, but unfortunately it's entirely the opposite of what your point appears to be because on a proper investment, the sign is flipped. If I earn $1000 on a $100k investment my gain was 1%. If I earn the same $1000 on a $10k investment, my gain is a more respectable 10%. The gain on the odds bet is 0%, so the analogy here would be that you can invest $10k to earn $1k, or you can invest $100k and earn the same $1k. You're obviously ill-advised to invest the extra $90k for no expected gain relative to the initial $1k.

Quote:

Let's say a player has $100 in his rack which he intends to gamble. He makes a $10 line bet, rolls a 5 and asks you, "Should I take odds?" You might begin by asking what's his objective? Play time? Loss per hour? Free drinks? etc, etc. He says, "Naw, I don't care about all that stuff. I just want a fair shot. My buddy told me the house takes a cut out of all the action and I just want to minimize that cut. I want as close to a fair game as I can get." What's your answer? Wait for the next come-out roll my son? Bet a 3-way craps grasshopper?

Any non-AP who says they're not interested in play time is lying to you. No tourist gambler flies to Vegas to make one bet and leave (except for this guy). No local gambler goes to a casino once and then gives it up forever. You're focusing on a given amount of handle, which I contend is not something most gamblers look at and in any event is not what actually happens when gamblers actually keep playing (whether it's next month, next year, whatever). Focusing on an equivalent wagering handle would indicate that you would advocate that a line bet player should play for 20 minutes at full odds rather than an hour with less than full odds (or no odds at all). The expected loss is obviously smaller because played for 1/3 the time and therefore made 1/3 the bets. It should be self-evident that making 1/3 the bets has 1/3 the expected loss.

But why stop there? Even if you could find someone who was more focused on total handle than play time or win/loss limit, if that total handle goal is less than $500, why not advise them to make a single $5 line bet at Casino Royale and make a 100x odds bet behind it? Is that really your advice to someone who "wants to play craps?"

Here's what I'd say to our hapless underfunded gambler:
"The house's average win on each $10 line bet is 14 cents. The odds bet has no house edge, that's why they call it free, so making the odds bet doesn't change that 14 cents. You can make odds bets and have a chance at winning or losing a bigger amount, or you can stick with no odds and have smaller swings. With your $100 bankroll, you might only last 2 bets if you take full odds and lose the next two hands. If you take no odds, you'll last at least 10 bets. Are you willing to risk a quick bust-out for a shot at bigger wins, or do you want to play it safer even if it means it'll be harder to win a lot?"

I think it's inappropriate to advise him to always make a full odds bet when you don't know why he's playing, especially given his bankroll to bet ratio.

Quote:

Players experience action through their betting, the number of bets they make and the amounts they risk. Whether they know the exact value of their action or not is immaterial -- they're playing for action.
...
Again, a noble goal, but not the issue at hand. The idea here was to CHANGE the way gambler's actually gamble by educating them as to the facts.

They *are* playing for action, so suggesting they make fewer wagers is counterproductive. And what kind of change do you expect from education? Are you actually advocating for gamblers to play fewer hands and instead put more of their total action on the odds bets? Because otherwise, there's no way to change a pass bettor's behavior in a way that impacts the house win. Someone who plays $10 craps in Las Vegas can't bet any smaller than $10, and if they have the bankroll to make odds bets, choosing to make or not make them doesn't change their cost to play; it just changes the variance and the total action. Do you tell them to only bet every other comeout? Every other shooter? Once per hour? The reductio ad absurdum advice to a craps player who buys in for $100 on a $10 table and wants to minimize their expected loss is "color up your chips right now and leave." Any wagering activity is, from an investment standpoint, suboptimal. You wrote:
Quote:

Oh yes, I see ... it's much better to let that player wait until the next come-out roll and make another line bet, right?

What you're missing is that the player is going to make that next line bet anyway. Given that, advising them to make the odds bet because "the house edge is lower" is misleading. The cost isn't lower, all you're doing is inflating the denominator of total action and putting more money at risk for the same expected loss. If they want the extra action, great. But if they're underfunded, as your example above, that's pretty bad advice.

Do you actually want to change that typical behavior and recommend that craps players actually make fewer bets in order to minimize their losses? That seems facile. Telling someone who wants to play craps "don't play as much craps" is not the sort of advice they're looking for. Even if you add "but you can have the equivalent handle by making odds bets and your house edge will be much smaller." The reply will be "but I only get to play for 1/3 the time?"
"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice." -- Girolamo Cardano, 1563
Steen
Steen
  • Threads: 0
  • Posts: 126
Joined: Apr 7, 2014
April 16th, 2016 at 1:55:07 AM permalink
Quote: MathExtremist

That's true, but unfortunately it's entirely the opposite of what your point appears to be because on a proper investment, the sign is flipped. If I earn $1000 on a $100k investment my gain was 1%. If I earn the same $1000 on a $10k investment, my gain is a more respectable 10%. The gain on the odds bet is 0%, so the analogy here would be that you can invest $10k to earn $1k, or you can invest $100k and earn the same $1k. You're obviously ill-advised to invest the extra $90k for no expected gain relative to the initial $1k.



First, the main purpose of this little analogy was to show that percentage (which you apparently abhor) is a vital, useful metric when evaluating monetary performance.

Second, you ignored the questions regarding performance. Are you happy knowing how many dollars you've earned without knowing what portion of your money produced them?

Third, since you acknowledged that one account performed better than the other I must assume you'd recommend moving money from the 1% low yield account to the 10% high yield account.

Fourth, unfortunately you can't just switch the signs and say the same thing - nice try. The goal with an expected positive return is to maximize the amount earned by moving money from the low yield 1% account to the high yield 10% account. The goal with an expected negative return should be to minimize the amount lost by moving money from the -10% account to the less negative -1% account.

Similarly, the goal with negative expectation Passline bets should be to move as much money as possible from -1.41% flat portion to the 0% odds portion.

Quote:

Any non-AP who says they're not interested in play time is lying to you. No tourist gambler flies to Vegas to make one bet and leave (except for this guy). No local gambler goes to a casino once and then gives it up forever. You're focusing on a given amount of handle, which I contend is not something most gamblers look at and in any event is not what actually happens when gamblers actually keep playing (whether it's next month, next year, whatever). Focusing on an equivalent wagering handle would indicate that you would advocate that a line bet player should play for 20 minutes at full odds rather than an hour with less than full odds (or no odds at all). The expected loss is obviously smaller because played for 1/3 the time and therefore made 1/3 the bets. It should be self-evident that making 1/3 the bets has 1/3 the expected loss.



Telling me what you think gamblers usually do is irrelevant to this discussion. It may be a worthy consideration for a casino intent on maximizing profits by studying behavior but not for a player seeking to minimize those same profits by analyzing the expectation of his wagers.

The original assertion was that switching $50 flat to $10 flat/$40 odds lowers the EV. Telling me that locals won't do it is immaterial.

Quote:

But why stop there? Even if you could find someone who was more focused on total handle than play time or win/loss limit, if that total handle goal is less than $500, why not advise them to make a single $5 line bet at Casino Royale and make a 100x odds bet behind it? Is that really your advice to someone who "wants to play craps?"



My intent in this thread is not to administer gambling advise, but simply to inform. There's a calculated house edge on all bets, it's not the same on every bet, an analysis can reveal which bets have more or less edge, and IF you want to minimize that edge then there are some things you can do. I haven't the time to interview each and every gambler to recommend a course of action. Given the information, those who care can decide for themselves. Want to keep betting that $50 flat w/o odds even though it comes at a cost? Knock yourself out.

Quote:

I think it's inappropriate to advise him to always make a full odds bet when you don't know why he's playing, especially given his bankroll to bet ratio.



I never gave any such advise. I simply said that taking odds lowers one's expected loss per dollar wagered.

Quote:

Do you actually want to change that typical behavior and recommend that craps players actually make fewer bets in order to minimize their losses? That seems facile. Telling someone who wants to play craps "don't play as much craps" is not the sort of advice they're looking for. Even if you add "but you can have the equivalent handle by making odds bets and your house edge will be much smaller." The reply will be "but I only get to play for 1/3 the time?"



I never said "Don't play as much craps." I simply said one must make a fair comparison using equal action. It matters not to me whether people change their behavior after learning the facts. I'm simply commenting on good information which they can act on if they choose. Taking odds doesn't restrict you to betting just 1/3 of the time. There are many ways to accommodate one's desires.

Steen
odiousgambit
odiousgambit
  • Threads: 327
  • Posts: 9775
Joined: Nov 9, 2009
April 16th, 2016 at 3:10:48 AM permalink
I can say for sure that when I have met up with WoV people to play Craps, often many of us* will play the way M-Extremist says no one does: decide on a size of bet that is comfortable and calculate from that the least that can be put on the line and the amount that can be put on the odds and think of the combination as the bet, declining to make all other bets [besides identical Come bets] too. It kind of shows education can pay off.

*the ones that don't learned to play before they ever heard of the Wizard
the next time Dame Fortune toys with your heart, your soul and your wallet, raise your glass and praise her thus: “Thanks for nothing, you cold-hearted, evil, damnable, nefarious, low-life, malicious monster from Hell!”   She is, after all, stone deaf. ... Arnold Snyder
odiousgambit
odiousgambit
  • Threads: 327
  • Posts: 9775
Joined: Nov 9, 2009
April 16th, 2016 at 3:12:03 AM permalink
It's easy to see that when there are two people trying to win an argument, each hates to concede any point to the other guy. So I guess I shouldn't give the following concession, on the other hand ...

*Although I have taken sides, I have dropped out of the argument - it's an odd if interesting form of circular argument when a discussion isn't really circular but is two sides being repetitive without anyone giving in.

*I wouldn't mind hearing what Steen thinks of it [M-Extremist: I know what you'll say]

Although I bet like Steen and the Wizard do recommend, I also find I am willing to make a larger sized combined bet when good odds are offered. If 2x odds are offered, my bet will be 5 on the line plus 10 on the odds. If 10x is offered, I'll bet 5+50 on odds every time [with less Come betting though]. It's just a fact that I respond to what is offered by taking on more action. I look at it as increasing the variance, if I can do that, the kind of action I want changes. That's a fact.





the next time Dame Fortune toys with your heart, your soul and your wallet, raise your glass and praise her thus: “Thanks for nothing, you cold-hearted, evil, damnable, nefarious, low-life, malicious monster from Hell!”   She is, after all, stone deaf. ... Arnold Snyder
MathExtremist
MathExtremist
  • Threads: 88
  • Posts: 6526
Joined: Aug 31, 2010
April 16th, 2016 at 7:56:27 AM permalink
Quote: odiousgambit

I can say for sure that when I have met up with WoV people to play Craps, often many of us* will play the way M-Extremist says no one does: decide on a size of bet that is comfortable and calculate from that the least that can be put on the line and the amount that can be put on the odds and think of the combination as the bet, declining to make all other bets [besides identical Come bets] too. It kind of shows education can pay off.

Okay, well that's how I play too so obviously there's miscommunication happening. For me, if I want to have about $100 at risk each shooter, I'll play a $10 three point molly with odds rather than make a single $100 line bet, a bunch of place bets, or 10 $10 hop bets. Someone who is thinking about betting $50 on the line with no odds is obviously better-advised to bet $15+3/4/5x odds instead, or $10+5x. There was never any dispute with the point that betting more of your money on 0% wagers than -1.41% wagers costs less, both as a percentage and in dollar terms, than otherwise. How could there be?

If this is what everyone means by "total action" then I'm simply misinterpreting what you're saying. To me, tracking "total action" is wanting to have a total wagering handle of, say, $300 and then stopping when you've wagered through that amount. At a $10 craps table, you could bet pass with no odds for about an hour, or you could bet pass with full odds for about 20 minutes. I don't think the latter is necessarily better from a gambler's standpoint even though the expected loss is obviously lower. Playing time is important because gambling is (or should be) about entertainment. If you're only focused on the cost of that entertainment, you're not paying attention to the real reason you're there. It's like sitting in a moving and fuming about how the tickets were $18 each instead of actually watching the movie.

Now, Steen chided me for abhoring percentages (which is clearly untrue if you know what I do for a living), but percentages are the output, not the input. Specifically, the player's realized edge is the ratio (percentage) of expected unit loss divided by the total units wagered across all their bets. You can decrease that percentage both by decreasing the loss and *increasing* the total units wagered. The only way it's accurate to say they're "losing less" is if you decrease the numerator, the expected loss.

If you find a $50 bettor who's not taking odds, it's generally good advice to tell them to switch to $10+odds because it lowers the theoretical loss to the casino. But if you find a $10 bettor who's not taking odds, the advice to take odds is not necessarily good even though odds would shrink the house edge (expected loss / total handle). Telling that $10 bettor to bet more in odds does not decrease their theoretical loss at all, it only increases their total handle. The naive gambler thinks "if I have a smaller house edge, that means I lose less" but that's not true here. The only way you understand that is if you also pay attention to the inputs (expected loss and total handle) rather than focus solely on the output (the percentage edge).
"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice." -- Girolamo Cardano, 1563
mustangsally
mustangsally
  • Threads: 25
  • Posts: 2463
Joined: Mar 29, 2011
April 16th, 2016 at 10:26:22 AM permalink
Quote: MathExtremist

If you find a $50 bettor who's not taking odds,
it's generally good advice to tell them to
switch to
$10+odds
because it lowers the theoretical loss to the casino.

only if asked
so you tell (kiss and tell)

never offer
that kind of advice unless asked!
<<< >>>
here why
<<< >>>
let us look at the next 60 pass line bets resolved

a $50 no odds player
has only a 0.4924518 (simple math)
probability of showing a net loss after 60 such bets
that's it!
so IS favored to be even or ahead in 2 hours or so of play.

This is what they can tell (and feel) by playing and watching.
craps players do have feelings

nows
others he sees (and pats his back while smiling and drinking)
Craps - Pass - $10-6x odds - $50 avg bet = .5096133
Craps - Pass - $10-5X odds -$43.3 avg bet = .50909667
Craps - Pass - $10-345X odds -$37.8 avg bet = .50635396
Craps - Pass - $10-3X odds -$30 avg bet = .51393233
Craps - Pass - $10-2X odds = .51619346

all favored to show a loss after 60 such bets (even at the same table)

"Ah, come out roll time"
bets another $50 pass line bet

seeing is believing

remember back to the 1600s and a gambler that wagered...
he knew (and could tell) by playing
I Heart Vi Hart
MathExtremist
MathExtremist
  • Threads: 88
  • Posts: 6526
Joined: Aug 31, 2010
April 16th, 2016 at 10:30:42 AM permalink
Quote: mustangsally

only if asked
so you tell (kiss and tell)

never offer
that kind of advice unless asked!

I agree. I never offer advice at the craps table, even if someone's betting on the big 6/8 with enough money to make the equivalent place bet.

Quote:

a $50 no odds player
has only a 0.4924518 (simple math)
probability of showing a net loss after 60 such bets
that's it!
so IS favored to be even or ahead in 2 hours or so of play.
...
Craps - Pass - $10-6x odds - $50 avg bet = .5096133
Craps - Pass - $10-5X odds -$43.3 avg bet = .50909667
Craps - Pass - $10-345X odds -$37.8 avg bet = .50635396
Craps - Pass - $10-3X odds -$30 avg bet = .51393233
Craps - Pass - $10-2X odds = .51619346

all favored to show a loss after 60 such bets (even at the same table)

"Pretty sneaky, sis!"

What are the odds of the same players being "even or ahead" after 59 bets?
"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice." -- Girolamo Cardano, 1563
mustangsally
mustangsally
  • Threads: 25
  • Posts: 2463
Joined: Mar 29, 2011
April 16th, 2016 at 10:50:25 AM permalink
Quote: MathExtremist

"Pretty sneaky, sis!"

was not meant to be that way
it is just what happens
and with variance many of these no odds players (i have seen many) have non-losing sessions many times in a row
they know what works for them (and may have burned and crashed taking the odds)

Quote: MathExtremist

What are the odds of the same players being "even or ahead" after 59 bets?

even or ahead does not contain odd (so it is not odd at all)
so I have never been there (hehe)

but I had a program open for the $10 with 345X odds = 50.625857%
to show a losing session after 59 bets

time to get ready to fly out from Ontario
go Angels!
start another win streak tonight!!
I Heart Vi Hart
MathExtremist
MathExtremist
  • Threads: 88
  • Posts: 6526
Joined: Aug 31, 2010
April 16th, 2016 at 11:24:20 AM permalink
One more thought on why it's more relevant to examine dollars than percentages:

Suppose you're standing outside Madame Tussaud's on the Strip. It's a Friday night and you want to play craps for a few hours. You can either go left, to the Venetian or go right, to Casino Royale. Casino Royale has a $5 table with 100x odds. The Venetian has $25 tables with 3/4/5x odds, except for the $5/2x table by the sports book. Let's assume you don't want to play there.

Say you think your bankroll can support action of about $100/hand. You could bet $25 +3/4/5x odds at the Venetian, or you could bet $5 + 20x odds at Casino Royale. By percentages, the Venetian game has a house edge of 0.37% while the CR game has an edge of 0.1%. CR is more than 3x better. And by units, it's even starker -- you're betting 5x as much at the Venetian and therefore your expected loss is 5x as much.

But how much exactly? If you go by the 1/2 bet per hour rule of thumb, it costs you $25 to play two hours at the Venetian and only $5 to play at Casino Royale. That means the theoretical cost difference is $20. To me, it's an easy call to play at the Venetian: I'm happy to accept that additional $20 theoretical cost in order to spend my time around well-heeled gamblers drinking Scotch rather than college kids drinking yards of beer. The point is that you need to examine the cost of your betting activities in theoretical dollars, not just percentages, in order to evaluate your wagering activity in the larger context of your overall vacation. If I spend $150/night on a hotel and $400 on plane tickets to get to Vegas, the extra $20 is just a rounding error. I spend more on parking at my home airport. And if I'm getting Macallan 12 at the Venetian and I can only get Johnnie Walker Red at the CR, then that $20 difference disappears very quickly...
Last edited by: MathExtremist on Apr 16, 2016
"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice." -- Girolamo Cardano, 1563
Steen
Steen
  • Threads: 0
  • Posts: 126
Joined: Apr 7, 2014
April 16th, 2016 at 3:16:48 PM permalink
Quote: odiousgambit

It's easy to see that when there are two people trying to win an argument, each hates to concede any point to the other guy. So I guess I shouldn't give the following concession, on the other hand ...

*Although I have taken sides, I have dropped out of the argument - it's an odd if interesting form of circular argument when a discussion isn't really circular but is two sides being repetitive without anyone giving in.

*I wouldn't mind hearing what Steen thinks of it [M-Extremist: I know what you'll say]



Far be it for me to give the wrong impression here. I have the utmost respect for Stacy. No doubt his mathematical skills far exceed mine. But the math in this situation is relatively simple and the debate a matter of logic and perspective. My hope was that a discussion would root out our misunderstanding.

The things he's said are worthy and something a gambler should consider. I hope I've made it clear that EV is not the only thing that a gambler should consider but instead an important element.

While maintaining the same amount of action by reallocating a portion of one's line bets to one's odds bets is not bad advise, it's also not bad advise to increase the size of one's wager by merely adding to the odds bet. Increasing the size of one's wager in relation to one's bankroll improves the chances to reach a monetary goal while (in the case of odds) lowering the average loss per dollar wagered.

Hence, the taking of odds would necessarily come at the expense of making other house advantaged bets (and at the expense of reducing one's play time). Taken to the extreme, the best chance a gambler has to double his bankroll is to bet his entire stake in one wager. But how many gamblers are willing or content to do that? The thrill of the game and desire for protracted action are therefore overriding interests (that come at a cost).

I think I've said what I needed, in enough different ways that it surely must be clear by now. Take it or leave it and I'll give it a rest :-)

Steen
Last edited by: Steen on Apr 16, 2016
bodyforlife
bodyforlife
  • Threads: 33
  • Posts: 180
Joined: Feb 25, 2013
April 17th, 2016 at 7:27:15 AM permalink
Bravo Steen!
mustangsally
mustangsally
  • Threads: 25
  • Posts: 2463
Joined: Mar 29, 2011
April 17th, 2016 at 12:06:50 PM permalink
Quote: Steen

Taken to the extreme, the best chance a gambler has to double his bankroll is to bet his entire stake in one wager.

that is called Bold Play by the math folks
as many due know now

and even a few math folks
have shown that there is a better method than bet all one time,
slightly better and takes more trials too

have never done the math on it
SN Ethier was on that team iirc
Quote: Steen

But how many gamblers are willing or content to do that?

Oh, just one wager!
a few (out of billions) I would think

<<< >>>
FYI
I put together a pass line betting system (calculated it too)
that can double a 20 unit bankroll exactly (say $100 buy-in with $5 flat bets) using 345x odds that doubles at
0.4921 68218
(pass line bet it all on one wager at 0.4929 29293)

average number of bets = 17.1
this adds to the fun factor, yes, yes

I do not think (m)any would really tell the difference, but could be wrong here

the system
Bankroll <6 units then bet it all on pass line
Bankroll >33 units then bet what it takes on the pass line (no odds) to hit the 40 unit mark
Bankroll 6 to 33 units then
1 unit flat and 345x odds

I think I tried this one time for fun
and was too
Sally
I Heart Vi Hart
MrGoldenSun
MrGoldenSun
  • Threads: 3
  • Posts: 252
Joined: Apr 1, 2016
April 18th, 2016 at 6:42:48 AM permalink
Quote: MathExtremist

Quote: odiousgambit

I can say for sure that when I have met up with WoV people to play Craps, often many of us* will play the way M-Extremist says no one does: decide on a size of bet that is comfortable and calculate from that the least that can be put on the line and the amount that can be put on the odds and think of the combination as the bet, declining to make all other bets [besides identical Come bets] too. It kind of shows education can pay off.

Okay, well that's how I play too



Adding my playing style to the list, I bet the table minimum on the line and then take whatever odds I feel like taking at the time, usually 1x. $5 at 3x4x5x is pretty much always sufficiently big for me.
rushdl
rushdl
  • Threads: 9
  • Posts: 177
Joined: Jan 15, 2016
April 18th, 2016 at 7:08:05 AM permalink
To the original "Q", the odds remain the same whether on come-out or not so leave your original odds bet alone and work them until this shooter SO's.
mustangsally
mustangsally
  • Threads: 25
  • Posts: 2463
Joined: Mar 29, 2011
April 18th, 2016 at 8:45:40 AM permalink
Quote: MrGoldenSun

Adding my playing style to the list, I bet the table minimum on the line and then take whatever odds I feel like taking at the time, usually 1x.

and over 60 such bets in one session (happens a lot)
that gives you a 52.334% chance to show a net loss for that session (all thanks to the shooters of course where one might be you)
compared to only a 49.245% chance when taking no odds

this is the stuff that many craps players know just from playing the game
they can "feel" that 3%+ difference
and not doing any math on it either
Quote: MrGoldenSun

$5 at 3x4x5x is pretty much always sufficiently big for me.

many find that making one line bet at a time is very boring unless lots of $$$ is bet
I can agree with that at times...
and only becomes fun (or up to super fun) the more money won during the session play.

and
trying to double a starting bankroll of $100 to at least $200 with $5 and 345x odds has a success I calculate at 0.466564401

thank you for sharing
<<< >>>
numbers are fun
math is fun
https://www.mathsisfun.com/numbers/
I Heart Vi Hart
mustangsally
mustangsally
  • Threads: 25
  • Posts: 2463
Joined: Mar 29, 2011
April 18th, 2016 at 9:06:46 AM permalink
Quote: rushdl

To the original "Q", the odds remain the same whether on come-out or not so leave your original odds bet alone and work them until this shooter SO's.

so
the house edge is reduced a little bit when the come odds are off on the cor

so the chance to double a starting bankroll is reduced a little bit when the come odds are off on the cor

so the chance to play longer in a session of time is increased a little bit when the come odds are off on the cor

so the chance to feel "crummy in the tummy" is increased a little bit when the come odds are off on the cor
and one of the come bet does win

so the chance to feel "crummy in the tummy" is increased a little bit when the come odds are off on the cor
and a 7 winner is thrown

so the chance to feel "crummy in the tummy" is increased a LOT when the come odds are 'ON' on the cor
and a 7 winner is thrown
<<< >>>
so
that fact still remains (and many craps players do know) that when a 7 does roll,
and it always will,
it does more often on a come out roll
compared to any other roll not being the come out roll.

it is because how the game is set up to play
the event is a come out roll
this is easily seen in actual roll data (and simulation data too)
1 million rolls
167,055 7s rolled (1 standard deviation for those in-the-know)
event	% of 7s	actual
cor7 29.72% 49656
cor+1 19.79% 33054
cor+2 13.99% 23376
cor+3 10.21% 17060
cor+4 7.33% 12237
cor+5 5.36% 8954
cor+6 3.78% 6315
7 2.71% 4525
8 1.98% 3300
9 1.39% 2326
10 1.00% 1673
11 0.73% 1225
12 0.54% 908
13 0.41% 679
14 0.28% 472
15 0.20% 340
16 0.15% 252
17 0.11% 176
18 0.10% 166
19 0.06% 97
20 0.05% 83
21 0.03% 47
22 0.03% 45
23 0.01% 23
24 0.02% 26
25 0.01% 13
26 0.00% 8
27 0.00% 5
28 0.00% 8
29 0.00% 4
30 0.00% 2
no math required for this
math has day off

Alan Krigman says at
here

"Or, do you believe the myth that seven is more likely during come-outs than other rolls?"

one can be the judge
Last edited by: mustangsally on Apr 18, 2016
I Heart Vi Hart
TwoFeathersATL
TwoFeathersATL
  • Threads: 37
  • Posts: 3616
Joined: May 22, 2013
April 18th, 2016 at 9:53:31 AM permalink
Quote: mustangsally

<snip>
so
that fact still remains (and many craps players do know) that when a 7 does roll,
and it always will,
it does more often on a come out roll
compared to any other roll not being the come out roll.

it is because how the game is set up to play
the event is a come out roll
this is easily seen in actual roll data (and simulation data too)
1 million rolls

More often on a come out roll?
You're teasing me aren't you?
Youuuuuu MIGHT be a 'rascal' if.......(nevermind ;-)...2F
rushdl
rushdl
  • Threads: 9
  • Posts: 177
Joined: Jan 15, 2016
April 18th, 2016 at 10:10:17 AM permalink
No I don't subscribe to that theory since the dice are not sentient beings and do not know what a come-out roll is.
MrGoldenSun
MrGoldenSun
  • Threads: 3
  • Posts: 252
Joined: Apr 1, 2016
April 18th, 2016 at 12:33:36 PM permalink
Quote: mustangsally

and over 60 such bets in one session (happens a lot)
that gives you a 52.334% chance to show a net loss for that session (all thanks to the shooters of course where one might be you)
compared to only a 49.245% chance when taking no odds



That makes sense that the chance of a session loss increases, given that the odds have exactly zero house edge but you win any given odds bet under half the time.

My odds bets depend on how much variance I feel like having at that moment. There's no algorithm behind it. I don't mind a lower-variance game, so I often play a craps style that would be pretty boring for some people, but it's exciting for me. Playing $5 on the line allows me to profit $30 on my odds bets, which is an amount with which I'm happy.
mustangsally
mustangsally
  • Threads: 25
  • Posts: 2463
Joined: Mar 29, 2011
April 18th, 2016 at 4:25:31 PM permalink
Quote: TwoFeathersATL

More often on a come out roll?
You're teasing me aren't you?

no
"when a 7 does roll, and it always will,
it does more often on a come out roll
compared to any other roll not being the come out roll."

It is a simple concept because the probability of a 7 rolling on any roll is constant.

It is by far the reason come odds are off on that come out roll
I guess I could do the math on it too, later
<<< >>>
same concept as "wait time" for an event
(simple ones like in craps)

'SEE A HORN, BET A HORN'
comes to mind

and craps experts all over laugh at the saying
but that saying IS true. 100% as it goes

given the last roll was a Horn #

the wait time starts
the 1st roll has a 1 in 6 probability of being another Horn # (16.7%)

when the 1st roll is not a Horn (prob = 5/6)
but it is on the 2nd roll after a Horn
we have the sequence of 5/6 * 1/6 = 5/36 *100 = 13.9%
less chance it happens on the 2nd roll than the first
this is the distribution beginning and it continues

both NBA and MLB is starting up
dinner party time
Sally

so in closing
players that have their come odds off on the come out roll
are missing the roll that the 7 rolls on the most (the distribution starting on the come out roll)
it is that simple
I Heart Vi Hart
rushdl
rushdl
  • Threads: 9
  • Posts: 177
Joined: Jan 15, 2016
April 18th, 2016 at 4:53:20 PM permalink
I think, that, 100% of the rolls end on 7.
bodyforlife
bodyforlife
  • Threads: 33
  • Posts: 180
Joined: Feb 25, 2013
April 19th, 2016 at 12:56:05 AM permalink
Quote: rushdl

I think, that, 100% of the rolls end on 7.



What if you hit your point and quit?
GWAE
GWAE
  • Threads: 93
  • Posts: 9854
Joined: Sep 20, 2013
April 19th, 2016 at 5:06:42 AM permalink
Quote: bodyforlife

What if you hit your point and quit?



your session may be over but the roll keeps going.

What happens if you are the only one at the table and you just quit? Do they clear the table and start over when the next person comes or does the next person continue the roll?
Expect the worst and you will never be disappointed. I AM NOT PART OF GWAE RADIO SHOW
GWAE
GWAE
  • Threads: 93
  • Posts: 9854
Joined: Sep 20, 2013
April 24th, 2016 at 7:53:37 PM permalink
Quote: GWAE

Second session in a row where my working come bets got murdered. I had 3 come bets with $20 odds lose on a 7 come out 8 times. I did win a few 4s and 10's though on come out. I was down a bunch from the working come bets but overall only lost $60 in 90 minutes because of 1 shooter repeating a 10 7 times in 1 roll. Would have had a huge win if I wasn't working my odds.

The really frustrating part is I would have a 5,6,9 working. They would throw a 7 and then a 9. That happened all but 1 time.



The box man must have felt bad for me and rated me really high. For next month I have 5 $100 match play coupons. Have never been given match play for more than $25 at this casino. Now gotta figure out what to do with them since my line bets are always $5.

I guess doey don't and a few on 12. What would thst even be, I guess 100+match/150 don't and 8 on 12. Or maybe since I have 5 just risk the 100 on the line.
Expect the worst and you will never be disappointed. I AM NOT PART OF GWAE RADIO SHOW
  • Jump to: