Random Poster: Flooring, Truncating, Rounding..

[Brick]

optimal bets. This is where the ev is at. I'm constantly fine tuning my game no matter how small the ev gain. Sims only give me optimal bets for whole number tc's. I then use them as guidline to actually improve and correlate my betting strategy by .5 tc's. I also posted several years ago suggesting this to others but was met with negative feeback. Many believe there is no way(or no need)to improve ev and(or)optimal bets that simulators use.

> I think I even gave you numbers to document
> the gain in EV,obviously in vain :-).
> One thing is sure: the fact that so many
> players take interest in these decimal
> indices shows that people are not happy with
> the whole number solution.

> Francis Salmon

[Francis Salmon]

It's hard to come up with new or not so new ideas that challenge the prevailing opinion here.

Francis Salmon

[Norm Wattenberger]

> Many believe there is no way(or no
> need)to improve ev and(or)optimal bets that
> simulators use.

Simulators can do exactly what you are proposing. And they can show if this helps or hurts. If you are doing it by the seat of your pants, you are may be harming your RoR.

[Brick]

I find optimal bet spread and SCORE for a particular game. Then I will increase bets at .5 tc increments. For instance if an optimal bet of 4 units at 2tc and 8 units at 3tc is called for,I'll place a 6 unit bet at a 2.5tc. Not sure if you consider this to be by the seat of my pants,but it seems to be efficient.

[Norm Wattenberger]

If the optimal bet is 4 units at 2tc, this means it is 4 units anywhere in the range of 2 to 2.99 if the optimal bets were calculated using truncation or flooring. The optimal bets are not calculated at a "point." They are always calculated in a range. They certainly are not calculated at the bottom of that range. That wouldn't be optimal. If the advantage were linear, then 4 units would be ideal at 2.5; not 6 units. But, it isn't linear and the correct bet should be calculated if you wish to bet optimally.

> I find optimal bet spread and SCORE for a
> particular game. Then I will increase bets
> at .5 tc increments. For instance if an
> optimal bet of 4 units at 2tc and 8 units at
> 3tc is called for,I'll place a 6 unit bet at
> a 2.5tc. Not sure if you consider this to be
> by the seat of my pants,but it seems to be
> efficient.

Serious Blackjack Software

[Brick]

What if optimal bets are boxed and based on the exact tc average,in other words 2.5 to 3.5 = 3 tc?

[Norm Wattenberger]

> What if optimal bets are boxed and based on
> the exact tc average,in other words 2.5 to
> 3.5 = 3 tc?

There is no general rule when you are making assumptions. The correct numbers can be calculated. Look, no one at high stakes makes perfectly optimal bets. I think the point is that red chip players should play as close to optimal play as possible to build bankrolls. Black chip players should, at least, know what the Hell the perfect bet is even if they can't always put it forward. And, they should know the penalty for cover when they use it. Combinatorial Analysis is useless in these calculations.

[Sonny]

> What if optimal bets are boxed and based on
> the exact tc average,in other words 2.5 to
> 3.5 = 3 tc?

Rounding the TC calculations will not help since it still has the fundamental flaw of overbetting the TC ranges that are duplicated. As Norm said, the advantage at a TC of +3 includes all advantages from 3.0 to 3.99. If you want to raise your bet at 3.5 then you must first exclude all of the advantages from 3.5 to 3.99 in your calculation of the advantage at 3.0 otherwise you are counting those advantages twice and will be overbetting all integer-valued TCs.

What you are talking about when you say "fine tuning" is finding the advantage from 2 to 2.5, then 2.5 to 3, then 3 to 3.5 and adjusting your betting strategy that way. That seems completely reasonable, although I don't know that the gain would be very much. It would depend on the frequency of the "interim" TCs and the amount of increased action one could obtain using them. It would also depend greatly on the ability of the player to estimate these TCs properly, which, I agree, is well within a skilled player's capability.

However, I think you will find that, as Norm showed, the difference in advantage between 3.0 and 3.5 is not enought to safely raise your bet. Unless you are an astronomically high roller, an extra 0.006 units will probably not mean much to you.

-Sonny-

[Sonny]

> Not sure if you consider this to be
> by the seat of my pants, but it seems to be
> efficient.

The phrases "not sure" and "seems to be efficient" aren't exactly selling your point. I know that your conclusions seem logical and reasonable to you, but you should listen to Norm on this one. Logic and reasoning can get you only so far. You will need cold hard evidence to take you the rest of the way. Norm carries all of these things in his pocket.

He's not just some jackass like me. =)

-Sonny-