Poll

8 votes (50%)
3 votes (18.75%)
5 votes (31.25%)

16 members have voted

7craps
7craps
Joined: Jan 23, 2010
  • Threads: 18
  • Posts: 1841
September 15th, 2010 at 10:47:40 AM permalink
Quote: MathExtremist

No, you misunderstand. I don't think your betting system is worth the effort to test, and I doubt 7craps thinks so either.



Make a poll from the question:
Is there "ANY" value in testing betting systems?
winsome johnny (not Win some johnny)
SOOPOO
SOOPOO
Joined: Aug 8, 2010
  • Threads: 101
  • Posts: 5770
September 15th, 2010 at 11:01:23 AM permalink
If the purpose of the 'test' is to try and prove that the system can win in the long run, then, no, there is no value in testing, since no system can beat the negative EV game, and you do not need to 'test' to prove that. However, there are other variables that 'testing' may give you. Average time to ruin, overall house edge compared to other systems, how long you can play from a given bankroll, how much in comps you could generate, etc. Example being
1. play pass line with single odds every come out.
2. play the 'yo' on every bet.
You could easily analyze those 2 different 'systems' for volatility, -EV, etc.
MathExtremist
MathExtremist
Joined: Aug 31, 2010
  • Threads: 88
  • Posts: 6526
September 15th, 2010 at 11:16:34 AM permalink
Quote: 7craps

Make a poll from the question:
Is there "ANY" value in testing betting systems?

If you agree that there is a difference between "betting systems" and "systems of play" read on and join in.

Actually , I do place some value on testing systems, mostly systems of play, not betting systems. But, MathExtremist is correct to a point. After testing many betting systems, Im worn out. They all end up doing the same things in the short and long runs.



I was specifically referring to the notion of testing to see whether betting system X can beat the house edge. We all know it can't. I believe that evaluating a method of play can be very useful. In many cases, it can adjust the win/loss frequencies (and amounts) to better hit a player's sweet spot. If you're a player that likes infrequent large awards like a slot machine pays, there are ways to bet at craps to achieve that distribution. If you instead like winning small awards frequently at the cost of an infrequent large loss, there are ways to do that too (e.g. Martingale). There are also ways to "work around" some of the annoying or upsetting events that happen at a dice table, such as the point-seven scenario. One system I play does this very well (it breaks even).

Quote:

Craps experts Always say "take full odds". It lowers the house edge.
Well, that is a true statement, but compared to what?



First, truth isn't compared to anything. It is true, without qualification, that taking full odds lowers the house edge as a percentage of your total wager. It does *not* lower the house edge as a percentage of the wager on the passline. If you make $5 line bets, your expected loss is $0.07 regardless of how much in odds you take. An expected loss of $0.07 on $5 is simply a higher percentage than the same expected loss of $0.07 on $30. The moral of the story: make sure you know what the numbers mean.
"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
nope27
nope27
Joined: Sep 5, 2010
  • Threads: 5
  • Posts: 126
September 15th, 2010 at 11:35:35 AM permalink
Quote: SOOPOO

However, there are other variables that 'testing' may give you. Average time to ruin, overall house edge compared to other systems, how long you can play from a given bankroll, how much in comps you could generate, etc. Example being
1. play pass line with single odds every come out.
2. play the 'yo' on every bet.
You could easily analyze those 2 different 'systems' for volatility, -EV, etc.


I think SOOPOO shows the best answer.
There are many variables that can be answered by testing. Some results more valuable than others.
rdw4potus
rdw4potus
Joined: Mar 11, 2010
  • Threads: 80
  • Posts: 6920
September 15th, 2010 at 11:42:54 AM permalink
Quote: SOOPOO

If the purpose of the 'test' is to try and prove that the system can win in the long run, then, no, there is no value in testing, since no system can beat the negative EV game, and you do not need to 'test' to prove that. However, there are other variables that 'testing' may give you. Average time to ruin, overall house edge compared to other systems, how long you can play from a given bankroll, how much in comps you could generate, etc. Example being
1. play pass line with single odds every come out.
2. play the 'yo' on every bet.
You could easily analyze those 2 different 'systems' for volatility, -EV, etc.



I agree. No system can beat a negative EV game in the long run. But they'll almost always have differentiating factors between them. Length of play, variance, probability of a winning session, comp value, and entertainment value would be the most interesting to me.
"So as the clock ticked and the day passed, opportunity met preparation, and luck happened." - Maurice Clarett
mrjjj
mrjjj
Joined: Sep 4, 2010
  • Threads: 75
  • Posts: 1579
September 18th, 2010 at 12:12:02 AM permalink
I think testing (in terms of roulette) is a guideline, nothing more. I dont need to test something for 800,000 RNG spins to put a label of 'good' on it. Yes, I do test, occasionally 30 hours a week. I sometimes test for a reason that gets talked about very little. I test (at home) so I cover all the 'what if' situations. The worse thing that can happen, you are playing your method and all of a sudden...."Oh oh, what do I do now? I never thought of that". I cover all aspects AT HOME instead of getting caught off guard during live play. Ken

  • Jump to: