I’ve been posting about regression in the last few days and it seems to me that I always go back to a 6 & 8 regression as offering more bang for the buck, but I was wondering:
Assuming an SRR of 7, what are the probabilities of hitting a 6 or an 8 on any given hand before hitting a 7? Combined, there are more ways to make 6 and 8 than 7.
Thanks.
Alamo
Replies:
Posted by: Skinny on November 6, 2016, 10:20 pm
Assuming an SRR of 7, the probability of hitting a 6 or an 8 on any given hand before hitting a 7 would be 50%.
Assuming an SRR of 6, the probability of hitting a 6 or an 8 on any given hand before hitting a 7 would be 45%.
I would like you to understand how I do this calculation because it is relevant to the question you did not ask. Since the only piece of information you have given me is SRR all I am able to do is equally distribute the calculation across all the numbers.
In other words, with an SRR of 7, one can expect to throw 5.14 sevens in 36 rolls of the dice instead of 6 sevens for a random roller. That means there are an additional .86 rolls of the dice in 36 rolls to distribute among the other numbers. By doing that proportionally I increase the number of sixes and eights from 5 to 5.14 each.
This is a crude way to doing the calculation but it is the best given the data we have with which to work. It would be better if you had statistics to show how many sixes or eights you can expect to actually roll over a significant amount of rolls. By comparing that to the number of sevens you roll over that time would give you a better way of determining the real probability.
Posted by: Chuckman on November 7, 2016, 5:29 am
If that is the case, with a SRR of 7 the chance of hitting a (6 or 8) before a 7 would be 2/3 or 66.7%
With a SRR of 6 the chance of hitting a (6 or 8) before a 7 would be 5/8 or 62.5%
Posted by: Skinny on November 7, 2016, 7:38 am
"Chuckman" wrote: I think what AlamoTx is asking is what is the chance of hitting a (6 or 8) as opposed to what is the chance of hitting a 6 or hitting an 8.
If that is the case, with a SRR of 7 the chance of hitting a (6 or 8) before a 7 would be 2/3 or 66.7%
With a SRR of 6 the chance of hitting a (6 or 8) before a 7 would be 5/8 or 62.5%
Agreed.
Posted by: AlamoTx on November 7, 2016, 1:12 pm
That 66.6% is about what I’m seeing in practice, maybe a little higher. Still, I am finding that even with a 70% likelihood, placing the 6 and 8 for $60 each and regressing on the first hit very often nets you a small win if you can repeat a few $30 6/8’s before a 7 out. However, that one time when you have a $60 6 & 8 and a $10 PL bet with single odds and you point 7, the hole is a big one to dig out of. I’ve actually had the board get wiped out 2 and 3 hands in a row in practice!
Unfortunately, it looks like I am still going to be a non-believer when it comes to regression betting over the long term. It is easier and less risky to build from small to large than to start large, get small and then re-build. In other words if you have a 70% chance of hitting a 6 and 8, starting with $80 on the layout as a worst case proposition, then hitting a 6 or an 8 immediately reduces your 7 risk to $45. This, versus risking $120 for the same single hit to get you back to even. This is just an opinion so it is worth what it’s worth.
And none of it works without some LUCKY bounces. Watch a football game sometime!
Alamo
