Let's discuss the craps "five count"

[Alan Mendelson]

Let's discuss the "five count" which, in theory, keeps your chips in the rail at the craps table until it appears that the shooter is not one of those guys or gals who is going to seven out too soon.

In another thread Frank Scoblete wrote about using the 5-count when random rollers/shooters have the dice:

Alan, play the 5-Count and you'll only be betting on 43 percent of them but you'll still have the excitement of the game. Give it a try.

Frank, would you mind summarizing the 5-count again and explaining the theory behind it?

I have another idea: how about a reverse five count that says bet on the shooters' first four rolls of the dice and then call your bets off?

[FrankScoblete]

Hi Alan:

I have extensive information on all aspects of the 5-Count in Casino Craps: Shoot to Win but here is a quick summary. The main purpose of the 5-Count is to simply reduce the amount of money you bet on random rollers --- and random rollers are the overwhelming majority of players.

According to mathematician Don Catlin, the 5-Count will eliminate 57 percent of the random rolls. You will therefore be betting on 43 percent of the rolls. Thus you are saving money. You are still at the table enjoying that thrill of anticipation.

Now your idea is not a good one because you will miss all extended rolls. With the 5-Count you will be on all extended rolls. Over the long run, no matter what betting system you use, the casino will win its edge on all of your bets. But it is better to lose the edge on 43 percent than on 100 percent of the random rolls.

If you are lucky enough to be at a table with a controlled shooter, Catlin's simulation of millions of shooters showed that the 5-Counter will be on dice controllers 11 percent more of the time.

Are there other ways to bet on random rollers? Sure. The best is no bets. But if you are a random roller then you will never play craps again. If you are a dice controller (influencer) you always bet on yourself.

How should you bet on random rollers? One Pass Line or Come bet with odds after the 5-Count is completed. I will bet quite small if I have to be on random rollers. I bet far much more on myself or a fellow dice controller such as Jerry "Stickman."

Due to a computer crash I lost your address. I'll send you a copy of Casino Craps: Shoot to Win and I Am a Card Counter. I reference you in the blackjack book.

[mustangsally]

How should you bet on random rollers? One Pass Line or Come bet with odds after the 5-Count is completed.

1) Why not bet the Don't Pass on the random roller if they are that bad compared to a DI?

and really. 2) why wait to bet? Is it really more fun not betting?
and 3) Why the odds?

4) Does betting the odds increase your chances of winning on a random roller?
if it does not, why bet them at all on any random roller like me?

5)Does betting the odds increase your fun of playing craps on a random roller?
This seems confusing.
I must be one of many that just does not "get it"

6) and are ALL DIs the same?
in other words,
Why would I bet on your rolls only if you do not roll numbers that I bet on?
Maybe your friend rolls the numbers I do bet on better than you.

or Alan rolls the numbers that I bet on more often than you.
Are you a better DI than Alan?
I think Alan is more handsome than you, I would bet on him for more fun.

Hi Alan!
My husband a few months ago said Hi to you at Rincon, I was not with him.
He was impressed with your voice as am I too.

Sally

[Alan Mendelson]

Frank, I never understood your 5-Count except that most players roll the dice four or five times before they 7-out.

But I have seen many players bet on a new shooter, and after three or four rolls take down their bets... and wait for the next shooter.

Regarding missing the "long rolls" well... just how many shooters have long rolls?

I was at Caesars this past weekend and I hit the fire bet for four numbers on two of my turns with the dice. But I must have had the dice about 30 times over the weekend.

Over the entire weekend, at the tables I played at, I was the only shooter who made four numbers on the fire bet, though one other shooter made 6 passes (five of them were 8). When most shooters had the dice we were lucky to see one pass, as most established their point and the seven came within 5 rolls.

So... with so few good shooters in the world, wouldn't it make more sense to "bet the norm" which would mean bet on the first few rolls and then take your bets down to wait for the inevitable 7?

Let me give you another option which is something I have also seen done with success: After three or four rolls you REDUCE your bets. For example: there is a new shooter and you bet $270 across, and after three rolls you reduce your action to $54 across just in case this shooter continues... and then you can press accordingly. This particular method gives you the best of both worlds: you take advantage of the likelihood that a shooter will throw the dice four times and you collect, and then using your profits you stay at the table in case the "hand" continues.

Sally... be sure you and your husband say hello again. Would love to meet and chat.

[FrankScoblete]

I'm heading out. I'll get to these two posts tomorrow.

[mustangsally]

Frank, I never understood your 5-Count except that most players roll the dice four or five times before they 7-out.

Don Catlin also showed this table of probabilities.

 Roll     	Probability of Seven Out
1	0.00000000
2	0.11111111
3	0.22788066
4	0.33264746
5	0.42387109
6	0.50278913
7	0.57095589
8	0.62980865
9	0.68060930
10	0.72445344

how-long-is-a-roll-part-2-

Looks like 50% of all shooters 7out by and including that 6th roll.
That feels right to me from all the times I have played Craps.

Mr Catlin is such a nice man, after I talked to him in an email a few years ago to really learn how to do some gambling math.

Sally... be sure you and your husband say hello again. Would love to meet and chat.

I will if I ever go there. I mostly visit Morongo, Pechanga or San Manual.

I see some do not use the 5 count as some rolls do not count in the 5 count but use a first 7 roll count instead.

Intuitively, your bet big then reduce your bets feels like a better way to go than waiting around for someone to get 5 qualifying rolls, then bet on them just to see them 7out right there.
There can not be that much fun in playing Craps that way for most players.
I never play with more than $100 so I would have to watch you play.

I know how to run computer simulations also (and can do math too, but many can not follow much math past the multiplication tables)
and could do your idea of a few bets and then take all the bets down.
This would only work over many many sessions, not just one or two.
That style of play would lose less than just reducing your bets over time.

Are you still playing craps by making pass line and odds and Fire bet only?
or do you sometimes say "go for it" and bet all the numbers too?
I can not do that as I play scared and am not scared to admit it.

Sally

[Alan Mendelson]

Sally... When I was at Caesars, I was betting passline and fire bet only when I was "stuck" at a $25 table and I was not the shooter. I was not betting odds. As you know, Sally, betting "odds" does not help the dice hit the point.

My new way to play craps is fire bet and passline only. This is the best way to hit the rare "jackpot" of a long roll. As I wrote on Wizard's site, I was at a table a few months ago when two shooters both threw 5 fire point hands and my $10 fire bets paid $2500 each time... and I only had $25 on the passline. Now, in hindsight had I been betting $135 across (or $130) I could have made perhaps six or seven thousand on each shooter -- but look at the money I saved on the other shooters who didn't have good hands.

At the $10 tables (most tables at Caesars are now $10) I placed the inside numbers if I was up.

On myself... I bet across.

[mustangsally]

I was at a table a few months ago when two shooters both threw 5 fire point hands and my $10 fire bets paid $2500 each time... and I only had $25 on the passline.

Fun!
That bet adds so much excitement to the game as it progresses to becoming a winning bet.
I was at CP last month (my birthday month) and the 3rd shooter after I started playing hit 5 points. He 7out right after that without a chance of going for the 6th point.
I had $5 bet on him. He threw the dice fast. Never took the time to even hold them.
I could not believe what he did, but he did have a rhythm to his rolls. Repeat after repeat. I have the actual dice rolls in another computer too.
I remember he went out on the 28th roll and everyone made good money from him. Place, Come, Pass with odds.
The other shooters that followed him went out fast and so did most players
With that shooter I am a lifetime winner at Craps.
I am only 24 years young and now up about $700
I play more Roulette, Video poker and Baccarat.
My husband and his friends love craps so I wanted to learn about it.

When you play for fun, I find you win more often. Maybe just makes you feel better than trying to win and can't

someone should, maybe me soon, compare the 5 count and it's 57% no bets to the 7 count and it's 57% no bets to your idea of just maybe 3 bets and stop. I have to read more on how those simulations were done. I think pass and come were the only bets being made.


The weekend is here!
Have fun!
Sally

[regnis]

Wierd fire bet at grand Victoria in Illinois. They used to not have a fire bet but if shooter made all six points he got $4,000. Now, they took out that freebie and have a fire bet. But the payoffs for 4 to 6 are low because they are also paying on 3.

I dont have the exact payoffs as I have not played there but will report back when I have all the info. But everyone is complaining to me that it's not worth it with the smaller payoffs. Even tho you get paid for 3, it's not enough to make it exciting. No chance for the big win.

[Alan Mendelson]

regnis I heard about this modified Fire Bet before but to be honest I would prefer a FB that paid for three different points. I've gone through entire weekends without a 4-point payoff, and only once was there a 6-point payoff, and 3 5-point payoffs.

I honestly think I have a profit making the Fire Bet as it stands now, but I think I would be a lot better off if there were a payoff on 3 points.

[mustangsally]

a fire bet. But the payoffs for 4 to 6 are low because they are also paying on 3.

I dont have the exact payoffs as I have not played there but will report back when I have all the info. But everyone is complaining to me that it's not worth it with the smaller payoffs. Even tho you get paid for 3, it's not enough to make it exciting. No chance for the big win.

They have this too Downtown Vegas at the Plaza and the Vegas Club
pays 7 for 1 for 3 (6 to 1)
and 300 for 1 on the 6
The WoO shows this pay table that looks like what I played on.
We hit many times the 3 points but never higher than that for the time I played there

POINTS PAYS	PROBABILITY	RETURN
0	-1	0.593939	-0.593939
1	-1	0.260750	-0.26075
2	-1	0.101275	-0.101275
3	6	0.033434	0.200605
4	29	0.008798	0.255147
5	149	0.001640	0.244350
6	299	0.000162	0.048568
Total		1	-0.207295

the highest return is at the 4 and 5 point win

compare that to the standard 1000 for 1 pay table

POINTS PAYS	PROBABILITY	RETURN
0	-1	0.593939	-0.593939
1	-1	0.260750	-0.26075
2	-1	0.101275	-0.101275
3	-1	0.033434	-0.033434
4	24	0.008798	0.211156
5	249	0.001640	0.408343
6	999	0.000162	0.162272
Total		1	-0.207628

the highest return is at the 5 point win

Sally

Alan, while I was out tonight my computer finished a few simulations for you.
I want to do some math on it before I post later to see the calculated values.
Your "bet for 4 rolls per shooter" and take down the bets looks to be about equal to the 5 count in loss per shooter.
(I compared to Don Catlin's sim result values but can easily calculate what they should be too once I grasp how exactly that 5 count works. Some rolls are actually not counted)

time for weekend play!
have fun!

[mustangsally]

Alan,
Your "bet for 4 rolls per shooter" and take down the bets looks to be about equal to the 5 count in loss per shooter.
(I compared to Don Catlin's sim result values but can easily calculate what they should be too once I grasp how exactly that 5 count works. Some rolls are actually not counted)

OK
The 5 count starts counting (starts to increment) as soon as the shooter rolls a point number.
so if your first rolls are 7,7,11,3,7,7,11,7
your count is still = 0 and you have not been qualified to bet on
next roll 6, count = 1
the next 3 rolls will count no matter what they are, if we make it that far
now we are at count = 4
for count to get to 5 that 5th roll must again be a point number

I think I got it

Now Don Catlin's massive simulation, that does not show interestingly any calculations - strange from a math guy - shows
"The 5-Counter’s average loss per shooter was 0.20396344"
5count proof

also the claims of the Captian's (and Frank) system of play is, among other things
"It would position you to be in the game when the other 50 percent of the shooters rolled, among whom would be four types of shooters:
1) shooters who would break your heart by sevening out shortly after you put your money at risk (these formed the majority of post 5-Count shooters)
2) the shooters who through sheer luck were having good rolls
3) shooters who had developed a "rhythmic roll" that was an unconscious controlled dice shot; this through trial and error in the casinos
4) shooters who were conscious rhythmic rollers or, as we now call them, dice controllers who had purposely learned how to flip the game into the positive zone when they rolled the dice"

#2 being normal variance and
#3 proof of the often heard line "There's a sucker born every minute"

Back to your 4 count.
The sim data I now have I need to get in a presentable order but I finished the math on it
average loss per shooter = $0.203389346
LESS than the Don Catlin sim results.

(The bets were $6 on Place 6 and Place 8, always working as those were the parameters Don used in his first sim. Bets taken down after 4 rolls and wait for next shooter)

Of course any simulation of any size will have an error of some size. That is why I am puzzled on why Mr. Catlin did not show the actual loss value per shooter calculated. Maybe he did but it was omitted from the article.


Alan, you are talented in more ways than one. (your great voice)

My math calculations.
One can skip this part if math causes any type of problems.

10/36 chance of winning $7
6/36 chance of losing $12
20/36 chance of a push (no win or loss)
I sum the products of each and get -$0.055555556 per roll

that = the expected value (or loss)
-$0.055555556 / $12 (avg bet per roll) = -0.00462963 = house edge per roll
that = 1/216 for those that see that
and we bet the first 4 rolls for each shooter b4 taking our bets down
But every shooter does not survive to 3 or 4 rolls.
We can use Don's calculations here, I do have my own that matches exactly with Mr Catlin

8/9 or about 0.888888889 = the probability of any shooter making a 3rd roll
about 0.772119342 = the probability of any shooter making a 4th roll
every shooter makes the 1st and the 2nd roll.
so the average number of rolls per shooter that we bet on = 1+1+ 0.888888889+0.772119342 = 3.66100823

3.66100823 * -$0.055555556 (-2/36) = about -$0.203389346
a figure less than Don Catlin's sim result - yea
and very close to my sim results - yahoo!

more later next week
Sally

[FrankScoblete]

Mustang Sally has presented fun material. However, we are not betting on sequences of numbers such 1-2-3-4-5-6-7-8 consecutive rolls because each roll is a unique event and past rolls have no bearing on the roll about to take place now. If you wanted to bet sequences you would have to place a bet at the beginning of the shooter's run that says, "He will make six numbers in a row that are not the seven." (Or however numbers you wanted.) If you are not betting a sequence but betting individually the casino edge is always there --- the seven will show up 16.67 percent of the time. The 5-Count is just reducing the total number of random rolls you bet on by 57 percent.

The 5-Count is not attempting to say that after it is reached the game starts to favor the player. A random game is a random game. Yes, as Catlin showed if a controlled shooter is at the table the 5-Count does allow a 5-Counter to be on that shooter more than the "bet all" player. (Bet all means the player who bets on every shooter.) But a six, seven, eight count would probably do the exact same thing (give or take). The purpose plain and simple of the 5-Count is to reduce exposure to random shooters, lower one's expected losses. Since the player merely looks like other players who qualify shooters by some means and does get to be on enough rolls to make it look as if he is playing it is rare that a 5-Counter has any problems with the casino.

Stewart Ethier, perhaps an even more famous mathematician than Catlin, went over the 5-Count for my book Casino Craps: Shoot to Win and I have his results in the book.

If there are shooters who are rhythmic rollers who never attempted to be such then there are such rollers. If there aren't such rollers it doesn't matter. Such has no influence on the 5-Count. Anyone who makes it through, you bet on (one bet, Come or Pass with odds is now my recommendation). Why take odds? Because it is a normal way to play and since the odds are a wash between casino and player as long as you have a decent bankroll you lose nothing on the odds over time. Or don't take odds. Your choice.

I think I answered most of the ideas here.

[FrankScoblete]

Forgot one thing. Don't waste your money on the Fire Bet. It has between a 20-25 percent house edge --- worse than the worst slot machine.

My two new books are I Am a Card Counter and Confessions of a Wayward Catholic

[Alan Mendelson]

Frank betting odds will not make you win. Odds allow you to win more if the dice go your way.

Sally thanks for the numbers but in plain English what does it mean?

[mustangsally]

Stewart Ethier, perhaps an even more famous mathematician than Catlin, went over the 5-Count for my book Casino Craps: Shoot to Win and I have his results in the book.

My math I presented was from Alan's suggestion of betting the first 4 rolls of a shooter then stop. all bets are off until next shooter.
If a bet lost it was replaced during those 1st 4 rolls
and my math (simulations also) shows a LOWER loss - calculated - per shooter than your 5 count.

I really want to be shown if I made a mistake by anyone.
BruceZ would be my first choice. How famous is he?

I have never simulated the 5 count but plan on it.
Also using a Markov chain should get about as close to the actual probabilities of those that pass the 5 count and those that do not pass than what a simulation can.

I am certain Catlin and Ethier both could easily do that without any simulations.
added: I see SN Ethier in his Doctrine of Chances book page 516 has these two questions after a description of the 5 count
(a) Assuming an unskilled shooter, what is the probability that he will
achieve the five-count?
(b) If one bets only on shooters after they have achieved the five-count,
what proportion of rolls does one bet on, assuming all shooters are unskilled?
answers from his answer pdf
15.18 (a) 0.513627.
(b) 0.434930
what is missing here is the expected loss per shooter.
But that can also be calculated.

I am guessing his answers for that is in one of your books.
Should be not a problem but a fun challenge duplicating his results.
(I am actually almost done)

I have seen both Mr Catlin and Mr Ethier make errors in articles and or books.
My Uncle even felt bad finding an error in Ethier's Doctrine of Chances book. he showed it to me before it came up in a thread over at Wizard of Vegas forum.
Every math guy/girl makes errors. Most are caught very quickly, but some do slip thru the cracks.

But for you to say Ethier is perhaps more famous than Catlin?
and that means what to whom? Why does that even matter?

What should matter is that 2 different math guys arrived at the same answer. Period, Sir.
But I do not know this to be true.

Is Ethier or Catlin perhaps more famous than William Feller?
Does that make one of them better at math than the other?

I really wonder how they both would answer that question.

I have emailed Mr Catlin before and he is such a nice man.
Ethier is a Professor and I know my Uncle has sent a few emails to him in the past and he answered them. To me that is wonderful.

Sally

[mustangsally]

Sally thanks for the numbers but in plain English what does it mean?

Basically I took an idea you posted about betting on a shooter for just the first 4 rolls.
Place 6 and Place 8 for 4 rolls, replaced if lost.
After 4 rolls the bets are taken down. Start over with next shooter.

That matched Mr Catlin's first 5 count simulation parameters

Your idea has a LOWER loss per shooter than the 5 count
from my calculations. IF I made no errors.

Great job!

more coming next week!
Sally

dinner and dancing

[FrankScoblete]

Sorry Mustang Sally. You're right. It doesn't matter the fame of the mathematician. I stand corrected and apologize. When it comes to math I am never wrong...Uh, I'll apologize for that statement in advance.

[Alan Mendelson]

Thanks for the math work Sally. I will call this strategy the "Mendelson Chicken-Out" Four rolls and run.

[FrankScoblete]

I'm telling you Alan, my psychic powers tell me that you will lose the house edge per bet as those who wait for the 5-Count to bet. Watch and see.

Scobedini