Bankroll and Variance

[Bankerdude80]

Your trip bankroll is at $10K at the beginning of the year. Over the year, you plan for a two day Vegas excursion once per month strictly playing craps. 4 sessions each day. Your buy-in for each session is 5% of your total bankroll. As a DI, in each session you plan to toss the dice 3x before calling it quits with an average bet of $50. On other shooters, you limit your bets to an average of $15 per shooter (if that). You seek out low table minimums and empty tables. Each session has its own stop-loss that you set. Once your bankroll is gone it is gone--- no replenishment.

Is a regression strategy going to have the most potential for profit? Or, would a conservative strategy of limiting exposure and building on the hot rolls be the route to take? What would your approach be?

Your ultimate goal is to grow the bankroll over time. How much variance should one expect to confront with the ebb and flow of wins and losses over time? At what point and what $ amount would be the tell tale stop loss that says "STOP!-- you need to re-evaluate your betting and play"? Based on the parameters given, do you need a bigger bankroll?

[heavy]

Good question, and one that requires a little thought on the responder's part, which is probably why you haven't seen an answer yet. Let's break it down a bit. What you're talking about here is essentially a series of 96 sessions, each with a $1000 buy-in. That leaves you a $4000 cushion you can blow on happy ending massages over in China Town.

Personally, I'd set a 50% loss limit so if I blew throuogh $500 in any session I'd seriously consider exiting the game and adding the remaining $500 to the happy ending fund. I'd go with something like a 25% win objective. That is not a "stop win" amount. It's just a threshold. If I win $300 and the game starts to go the other direction then I won't give up that $250 win.

Now let's think about betting strategies. Believe it or not, long run players WANT a game with a lot of variance. They just need to sufficient bankroll to sustain the highs and lows. Because your biggest WINS come in games with high variance - but your biggest LOSSES come from the same game. Hence the need for loss limits.

I believe that (hands-down) the best way for long run players with sufficient bankroll to approach the game is (wait for it) Pass Line with max odds, Place the Six and Eight, Buy the Four and Ten if the vig is collected AFTER the win, and then make continuous Come Bets with odds, taking down the Place and Buy bets as the Come Bets travel.

Is that how I'd play? Probably not. But I think that's the "best" way to play from a mathematical standpoint.

Who else wants to chime in. Can I get a $204 across?

[London Shooter]

The $204 across here would be a great play to have a go with a genuine 10k bankroll.

It's a bit outside the scope of a 10% buyin and $50 risk that BD is talking about, but I have always thought MPs plan needed 10k and a comfortable feeling that a couple of PSOs weren't killing you so much that you would change your plan.

To that effect, I'd say sticking with the plan for up to a 5k draw down as the worst case would be a very interesting way to play. It's stood the test of time against verified rolls from DI get togethers, but you'd have to accept some days you might easily have a 3 roll session of -612, but would have to have the discipline to stick with it regardless to profit from the long hands.

If you play 204 across and do end up 5k in the hole, then time to re-asses. Chances are you certainly do not have any edge over the casino with your own rolls.

[mssthis1]

Personally, using those parameters, I would only play tables that I have a history on and preferably if I was the lone shooter. $10 or $15 pass line with odds and $135.00 across. After I recouped the initial layout I press. That's how I primarily played last year and grew my bankroll by several hundred percent until I decided to use it to buy farmland. There were times when I broke the rues last year and played strange tables or crowded tables. Most of the time on the crowded tables people showed up after I had been playing for a while. On those I started with a min pass line bet and $54.00 across.

I made 100% of my profit in 2015 on tables where I was alone or with one other shooter. I was net negative for the year on tables with 3 or more players.

So far in 2016 I've only had one opportunity to play alone and it was a push. In my local market the opportunity to play at an empty table rarely presents itself.

My first trip to Vegas this year is Easter weekend March 24/25/26 and again on the weekend of April 8/9/10. I'm looking forward to the opportunities.

[heavy]

The $204 across here would be a great play to have a go with a genuine 10k bankroll.

There are many ways to define "bankroll." For some people it's session bankroll, for others it's trip bankroll, for others it's the total amount they'll have in their gaming account at any time. Many of the advantage play crowd define it as your lifetime bankroll and speak in terms of the total amount you're willing to lose over the course of your lifetime without giving up the game.

Let's look at an example. Let's say a player takes $1000 with him to each session he plays. He is willing to lose up to $500 per session. He plays four sessions a month. That's $2000 a month he's willing to lose - times 12 months - makes $24000. So this guy has a $24000 annual bankroll. Now let's say he's 45 years old and plans to play for at least 30 years. That puts this fellows lifetime bankroll at $720,000. And many AP's start calculating things like average bet, etc. based on THAT number. Of course, many of these guys are playing a partial Kelly and that becomes and issue at craps because it's extremely difficult to calculate what your positive EV is - especially if you go to the tables and bet on every shooter that touches the dice. But for the sake of argument, let's say this player has a .4% positive EV. That's .004 times $720,000 for a full Kelly. Works out to a $2880 average bet based on the lifetime bankroll and the advantage over the game. So clearly a $1000 buy in isn't going to do the deal. So we were probably over zealous in determining what out lifetime bankroll was.

So go "real world" and say you're willing to lose $200 per session. Now you're at $800 a month times 12 months a year - times 30 years. That gets your lifetime bankroll down to $288,000. Still enough to buy a condo in Vegas or three rent houses in east Texas. And let's say we play a partial Kelly instead of a full Kelly and go with .002%. That gets you down to a $576 average bet. Still a whopper for a $1000 buy in. Anyway, my point in this was just to give you an example of how these things are calculated (I hope my math is working this morning).

Way back about ten years ago I did the whole Kelly calculation on what I felt my bankroll was at the time and came out with an average bet of somewhere around $80. That's actually fairly close to what I was betting at the time. However, I latched on to the "average" part of average bet and decided I could get there via regression betting. Basically $110 inside for a couple of rolls, then reduce to $44 inside for a couple of decisions, then press up and out slowly. The average bet would be around $80, which worked according to the Kelly. It played well on the tables as well as in my mind. A couple of hits at $110 inside kicked off $70. Regress to $44 inside guarantees a $26 win for the shooter. Sometimes it worked. Sometimes it did not. But nothing works 100% of the time.

[Bankerdude80]

But nothing works 100% of the time.

Yes, and this discussion definitely provides potential starting points and things to consider. Many good ideas in all the posts above. Like most things in life, the key is to be effective at handling the fluidity of the game. It is good to develop general guidelines/goals/rules. In our pursuit of the game, it is important that we be flexible and adapt to the changing variables (fatigue, table conditions, toss performance, etc) that impact our betting results. You've said it many times that discipline is key. I'd like to parrot that as well. We need to be able to make the right decisions in the moment. We also have to accept the fact that there are no "secret formulas", just sound principles backed up with math and sound thinking.

[mssthis1]

In our pursuit of the game, it is important that we be flexible and adapt to the changing variables (fatigue, table conditions, toss performance, etc) that impact our betting results. You've said it many times that discipline is key. I'd like to parrot that as well. We need to be able to make the right decisions in the moment. We also have to accept the fact that there are no "secret formulas", just sound principles backed up with math and sound thinking.

Bingo. I used to play using Americraps EO system on every player, no matter what. I found out my pockets weren't deep enough and my intestinal fortitude was not high enough, so I've honed table reading skills and now try to take what the table is giving me, especially when it comes to shooters other than myself.

[Americraps]

Hi Mssthis1,
That's RCs EO, not mine. Betting on randies worked for him, but he also stacked the table with DIs.

Hi Bankerdude, I would bet $30 6&8 and PL with max odds.