Systematic errors
Are you flooring your TC estimates? Do you know how to do the fractional math to determine TC in SD games? Are you using risk averse indices? It would seem to me you must have a systematic problem causing you to bet or play incorrectly.
Flooring TC estimates is always rounding toward negative infinity. 3.9 would be 3. and -2.1 would be -3. This makes sure your plays and bets are not made incorrect by rounding in the TC conversion. The mistake of not flooring your conversion would have you over betting. That would cost you money. If there are 23 cards played that would be (52 - 23)/52 = .558 decks left. Obviously you are not going to be able to TC = RC/.558 so you round so your estimation is less than the actual amount. 29/52 must be rounded to a larger fraction that you can divide with for when RC is positive and the other way when RC is negative. The obvious divisor is 3/5. It is slightly high so is better for a positive RC. The important thing is not which way you round your divisor but that your TC estimation error is always on the side of conservative bets. Usually flooring the real number TC estimate to an integer is sufficient. If it has an error in the wrong direction you are over betting and misplaying your hands that are close to the cusp of the decision which costs you money.
Dividing by a fraction is the same as multiplying by its inverse. X/(3/5) = X*5/3.
Indices for doubling and splitting are typically determined by the threshold Double if it increases your EV by 1 cent. You risk twice as much money to win 1 more penny. You are making variance shoot through the roof for not much gain. This results in less profit in the long run because you lost the option of taking another card if you need to hit hurting your per hand advantage. Before you doubled you would win more of the hands while winning or losing less money. When you double if you are likely to want another card due to a particular hand matchup you end up with twice as much money out there with a lousy hand.
My favorite hand to illustrate this is doubling A6 v 2. You double and catch a ten now you are stuck with twice the money out there hoping the dealer will bust with a 2 showing. You always hit soft 17 when not doubling because it is a bad total that is always better to take a safe card to. Only 4 out of 13 cards you can get to your double will change this position. In a positive count this ratio is smaller. The dealer will bust roughly 35% of the time with up card of 2. As the count goes up it is unfavorable to your hand being improved as the cards you need are being used up. The cards the dealer needs most 9,8,7 are those cards you have no information about with most counts. This affect makes the reward of a few pennies in EV to be less than the penalty from lost flexibility and losing double your bet. This ultimately cost you money in the long run and greatly increases variance. A much higher risk averse index is the most profitable. The increase in variance when not using risk averse indices make you much more likely to be down big in the short term!!!!
Are you flooring your TC estimates? Do you know how to do the fractional math to determine TC in SD games? Are you using risk averse indices? It would seem to me you must have a systematic problem causing you to bet or play incorrectly.
Flooring TC estimates is always rounding toward negative infinity. 3.9 would be 3. and -2.1 would be -3. This makes sure your plays and bets are not made incorrect by rounding in the TC conversion. The mistake of not flooring your conversion would have you over betting. That would cost you money. If there are 23 cards played that would be (52 - 23)/52 = .558 decks left. Obviously you are not going to be able to TC = RC/.558 so you round so your estimation is less than the actual amount. 29/52 must be rounded to a larger fraction that you can divide with for when RC is positive and the other way when RC is negative. The obvious divisor is 3/5. It is slightly high so is better for a positive RC. The important thing is not which way you round your divisor but that your TC estimation error is always on the side of conservative bets. Usually flooring the real number TC estimate to an integer is sufficient. If it has an error in the wrong direction you are over betting and misplaying your hands that are close to the cusp of the decision which costs you money.
Dividing by a fraction is the same as multiplying by its inverse. X/(3/5) = X*5/3.
Indices for doubling and splitting are typically determined by the threshold Double if it increases your EV by 1 cent. You risk twice as much money to win 1 more penny. You are making variance shoot through the roof for not much gain. This results in less profit in the long run because you lost the option of taking another card if you need to hit hurting your per hand advantage. Before you doubled you would win more of the hands while winning or losing less money. When you double if you are likely to want another card due to a particular hand matchup you end up with twice as much money out there with a lousy hand.
My favorite hand to illustrate this is doubling A6 v 2. You double and catch a ten now you are stuck with twice the money out there hoping the dealer will bust with a 2 showing. You always hit soft 17 when not doubling because it is a bad total that is always better to take a safe card to. Only 4 out of 13 cards you can get to your double will change this position. In a positive count this ratio is smaller. The dealer will bust roughly 35% of the time with up card of 2. As the count goes up it is unfavorable to your hand being improved as the cards you need are being used up. The cards the dealer needs most 9,8,7 are those cards you have no information about with most counts. This affect makes the reward of a few pennies in EV to be less than the penalty from lost flexibility and losing double your bet. This ultimately cost you money in the long run and greatly increases variance. A much higher risk averse index is the most profitable. The increase in variance when not using risk averse indices make you much more likely to be down big in the short term!!!!
