Most casino players bet more when they feel lucky and bet less when things aren’t going their way. The same players may pull back after a losing streak and push forward after a win. However, effective bet sizing adjustments only work when built on a foundation of perfect strategy. This article breaks down the relationship between bet sizing and perfect strategy and why one without the other can be meaningless.
What “Perfect Strategy” Means
A perfect strategy means playing every decision in a way that minimizes the house edge. Perfect strategy in blackjack is called basic strategy. It’s a complete set of rules that tells you when to hit, stand, split, or double down based on your cards and the dealer’s upcard. Basic strategy reduces the house edge to approximately 0.5% when followed without deviation.
A perfect strategy in video poker means selecting the optimal hold for every possible five-card deal. In baccarat, it means always betting on the banker. Each game has its own version of perfect play, and the house edge figures below reflect what happens when that strategy is applied correctly.
| Blackjack | Basic strategy | 0.46% – 0.5% |
| Video Poker (Jacks or Better) | Optimal hold strategy | 0.46% |
| Baccarat | Banker bet always | 1.06% |
| Craps | Pass line + odds bet | 0.18% – 0.37% |
| Blackjack | No strategy (guessing) | 2% – 4% |
| Roulette | No optimal strategy exists | 2.7% – 5.26% |
There is a significant gap between playing correctly and playing by feel. A blackjack player guessing their way through hands can face a house edge up to eight times higher than one using basic strategy. Bet sizing adjustments mean almost nothing if you start from that a disadvantaged position.
Why Bet Sizing Adjustments Require a Strong Foundation
Bet sizing adjustments help you capitalize on favorable situations and limit damage in unfavorable ones. But that they only works if you can accurately identify when a situation is favorable, which requires a perfect strategy as your baseline. A player who doesn’t use basic strategy will make suboptimal decisions during hands, inflating the house edge beyond what any clever bet sizing can compensate for.
The Kelly Criterion
The Kelly Criterion is the most respected mathematical framework for bet sizing in situations where you have a genuine edge. It was developed by scientist John L. Kelly Jr. in 1956 and has been used by professional gamblers and investors ever since. The formula is straightforward:
Kelly % = (Edge / Odds)
Where “Edge” is your advantage expressed as a decimal, and “Odds” is the net amount you win per unit wagered.
For example, if a card counter identifies a situation where they have a 1.5% edge over the house, and the bet pays even money (1:1 odds), the Kelly formula says:
Kelly % = 0.015 / 1 = 1.5% of bankroll per bet
This is the optimal bet size for this edge. You will not maximize your advantage of you bet too little. But you risk ruin faster than the edge can reward you if you bet too much.
The Kelly Criterion assumes your edge figure is accurate. If you are playing an imperfect strategy, your true edge is lower. Kelly-based sizing will lead you to overbetting a situation that was never favorable.
Card Counting and Bet Spreading
Card counting in blackjack only functions correctly when basic strategy is mastered first. The player gains an edge as the deck composition shifts toward high-value cards (tens and aces). The counter raises their bet. The house gains an edge as the deck shifts toward low cards. The counter lowers their bet. This is called bet spreading. A common spread used by recreational counters is 1:8, meaning they bet one unit at the minimum and eight units at the maximum advantage.
| -1 or lower | House has edge | $10 (table minimum) |
| 0 | Neutral | $10 |
| +1 | ~0.5% player edge | $20 |
| +2 | ~1.0% player edge | $40 |
| +3 | ~1.5% player edge | $60 |
| +4 | ~2.0% player edge | $80 |
| +5 or higher | ~2.5%+ player edge | $80+ |
This system generates a mathematical edge only because the bet sizing is tied to an accurate assessment of the situation. A counter who deviates from basic strategy during a high-count hand can easily give back more than the count advantage provides.
Flat Betting vs. Variable Betting
Many players debate whether it’s better to flat bet (same amount every hand) or vary bet sizes based on some system. The answer depends on whether you have a genuine edge.
- Without an edge (no perfect strategy). Varying your bet size doesn’t change the expected outcome. Every hand has the same house edge against you. You are just deciding how much of that edge applies to each hand.
- With an edge (perfect strategy + card counting). Varying your bet size based on the count is superior. You put more money at risk when the odds are in your favor and less when they are not.
| Flat betting | No strategy | 500 | $25 | -$250 (2% edge) |
| Variable betting | No strategy | 500 | $25 | -$250 (same result) |
| Flat betting | Basic strategy | 500 | $25 | -$57.50 (0.46% edge) |
| Variable betting | Basic strategy + count | 500 | $25 avg | +$150 to +$300 (estimated) |
Variable betting without strategy produces identical expected results to flat betting without strategy. The bet variation itself does nothing. Variable sizing can change the expected outcome only when combined with perfect strategy.
The Martingale Myth
The Martingale system is one of the most popular bet sizing strategies in existence. The idea is to double your bet after every loss, so that when you eventually win. Then, you can recover all losses and make a small profit. It sounds reasonable. But here’s why it falls apart mathematically.
The Martingale does not require a perfect strategy because it’s not trying to exploit an edge. Rather, it is trying to overcome a house edge through bet escalation. It fails for two key reasons.
- Table limits cap your progression. Most tables have a maximum bet. You hit the ceiling and can no longer double after just a handful of consecutive losses.
- The math of ruin is brutal. The chance of losing 8 hands in a row at blackjack with basic strategy is approximately 0.4%. But you may encounter this scenario at least once over 500 hands. When you do, the required bet to continue the Martingale may exceed the table maximum.
| 1 | $20 | $10 |
| 2 | $40 | $30 |
| 3 | $80 | $70 |
| 4 | $160 | $150 |
| 5 | $320 | $310 |
| 6 | $640 | $630 |
| 7 | $1,280 | $1,270 |
| 8 | $2,560 | $2,550 |
A standard $500 table maximum stops this progression cold at around hand 6. The Martingale is not a bet sizing strategy built on perfect play. Instead, it is a bet sizing strategy built on hope, which the numbers do not support.
Positive Progression Systems
Positive progression systems work in the opposite direction to Martingale. It allows you to increase bets after wins and return to base after losses. These are less dangerous because they don’t require escalating bets to recover losses. But they still don’t create an edge.
The only meaningful application of a positive progression is when combined with a game where perfect strategy lowers the house edge close to zero. In this scenario, you structure your session to make the most of winning streaks and limit exposure during losing ones.
| Martingale | After losses | Very high | No | No |
| Paroli | After wins | Low | No | Partially |
| 1-3-2-6 | After wins | Low to medium | No | Partially |
| Kelly Criterion | Based on edge | Calibrated | No | Yes |
| Count-based spreading | Based on count | Calibrated | No | Yes |
The pattern is consistent. Every system that works performs better with a perfect strategy underneath it. The systems that claim independence from strategy don’t hold up under mathematical scrutiny.
Doubling Down and Splitting
A perfect strategy doesn’t just affect how much you bet before a hand starts. It also determines when you increase your bet mid-hand through doubling down and splitting pairs. These are bet sizing decisions that only make mathematical sense when applied correctly.
Doubling down means doubling your bet in exchange for exactly one more card. Basic strategy identifies the specific moments where this is mathematically profitable, typically when you hold a total of 9, 10, or 11 and the dealer shows a weak upcard.
Splitting pairs means dividing a pair into two separate hands, each with its own full bet. This can turn one unfavorable hand into two profitable ones when done correctly. The expected value difference between correct and incorrect doubling/splitting decisions is significant.
| Player 11 vs Dealer 6 | Double down | +54% | Hit only | +35% |
| Player A-A vs Dealer 6 | Split | +63% | Hit only | +18% |
| Player 8-8 vs Dealer 10 | Split | -48% | Stand | -54% |
| Player 10-10 vs Dealer 5 | Stand | +68% | Split | +56% |
Every incorrect mid-hand bet sizing decision chips away at your expected value. Over hundreds of hands, these small differences compound into significant losses. And since doubling and splitting involve larger amounts of money on the table, getting them wrong is especially costly.
Conclusion
Systems, progressions, and gut-feel adjustments can attract attention because they feel like they are doing something. They may seem to work sometimes. But the math can be unwavering.
Bet sizing adjustments that hold up only function correctly when built on a foundation of perfect strategy. Thankfully, this perfect strategy is learnable. Basic strategy cards are available at most casinos and can be used openly at the table. Optimal video poker holds can be practiced online. Craps odds bets are simple to understand. The players who combine available tools with thoughtful, mathematically grounded bet sizing can give themselves the best possible chance at the table.
The Kelly Criterion
The Kelly Criterion is the most respected mathematical framework for bet sizing in situations where you have a genuine edge. It was developed by scientist John L. Kelly Jr. in 1956 and has been used by professional gamblers and investors ever since. The formula is straightforward:
Kelly % = (Edge / Odds)
Where “Edge” is your advantage expressed as a decimal, and “Odds” is the net amount you win per unit wagered.
For example, if a card counter identifies a situation where they have a 1.5% edge over the house and the bet pays even money (1:1 odds), the Kelly formula says:
Kelly % = 0.015 / 1 = 1.5% of bankroll per bet
This is the optimal bet size for this edge. You will not maximize your advantage of you bet too little. But you risk ruin faster than the edge can reward you if you bet too much.
The Kelly Criterion assumes your edge figure is accurate. If you are playing imperfect strategy, your true edge is lower. Kelly-based sizing will lead you to overbetting a situation that was never favorable.
Card Counting and Bet Spreading
Card counting in blackjack only functions correctly when basic strategy is mastered first. The player gains an edge as the deck composition shifts toward high-value cards (tens and aces). The counter raises their bet. The house gains an edge as the deck shifts toward low cards. The counter lowers their bet. This is called bet spreading. A common spread used by recreational counters is 1:8, meaning they bet one unit at the minimum and eight units at the maximum advantage.
| True Count | Player Edge | Recommended Bet (1 unit = $10) |
|---|---|---|
| -1 or lower | House has edge | $10 (table minimum) |
| 0 | Neutral | $10 |
| +1 | ~0.5% player edge | $20 |
| +2 | ~1.0% player edge | $40 |
| +3 | ~1.5% player edge | $60 |
| +4 | ~2.0% player edge | $80 |
| +5 or higher | ~2.5%+ player edge | $80+ |
This system generates a mathematical edge only because the bet sizing is tied to an accurate assessment of the situation. A counter who deviates from basic strategy during a high-count hand can easily give back more than the count advantage provides.
Flat Betting vs. Variable Betting
Many players debate whether it’s better to flat bet (same amount every hand) or vary bet sizes based on some system. The answer depends on whether you have a genuine edge.
- Without an edge (no perfect strategy). Varying your bet size doesn’t change the expected outcome. Every hand has the same house edge against you. You are just deciding how much of that edge applies to each hand.
- With an edge (perfect strategy + card counting). Varying your bet size based on the count is superior. You put more money at risk when the odds are in your favor and less when they are not.
| Betting Style | Strategy Applied | Hands Played | Avg Bet | Expected Result |
|---|---|---|---|---|
| Flat betting | No strategy | 500 | $25 | -$250 (2% edge) |
| Variable betting | No strategy | 500 | $25 | -$250 (same result) |
| Flat betting | Basic strategy | 500 | $25 | -$57.50 (0.46% edge) |
| Variable betting | Basic strategy + count | 500 | $25 avg | +$150 to +$300 (estimated) |
Variable betting without strategy produces identical expected results to flat betting without strategy. The bet variation itself does nothing. Variable sizing can change the expected outcome only when combined with perfect strategy.
The Martingale Myth
The Martingale system is one of the most popular bet sizing strategies in existence. The idea is to double your bet after every loss, so that when you eventually win. Then, you can recover all losses and make a small profit. It sounds reasonable. But here’s why it falls apart mathematically.
The Martingale does not require perfect strategy because it’s not trying to exploit an edge. Rather, it is trying to overcome a house edge through bet escalation. It fails for two key reasons.
- Table limits cap your progression. Most tables have a maximum bet. You hit the ceiling and can no longer double after just a handful of consecutive losses.
- The math of ruin is brutal. The chance of losing 8 hands in a row at blackjack with basic strategy is approximately 0.4%. But you may encounter this scenario at least once over 500 hands. When you do, the required bet to continue the Martingale may exceed the table maximum.
| Losses in a Row | Required Next Bet ($10 start) | Cumulative Loss if Stopped |
|---|---|---|
| 1 | $20 | $10 |
| 2 | $40 | $30 |
| 3 | $80 | $70 |
| 4 | $160 | $150 |
| 5 | $320 | $310 |
| 6 | $640 | $630 |
| 7 | $1,280 | $1,270 |
| 8 | $2,560 | $2,550 |
A standard $500 table maximum stops this progression cold at around hand 6. The Martingale is not a bet sizing strategy built on perfect play. Instead, it is a bet sizing strategy built on hope, which the numbers do not support.
Positive Progression Systems
Positive progression systems work in the opposite direction to Martingale. It allows you to increase bets after wins and return to base after losses. These are less dangerous because they don’t require escalating bets to recover losses. But they still don’t create an edge.
The only meaningful application of a positive progression is when combined with a game where perfect strategy lowers the house edge close to zero. In this scenario, you structure your session to make the most of winning streaks and limit exposure during losing ones.
| System | Direction of Bet Increase | Risk Level | Works Without Perfect Strategy? | Effective With Perfect Strategy? |
|---|---|---|---|---|
| Martingale | After losses | Very high | No | No |
| Paroli | After wins | Low | No | Partially |
| 1-3-2-6 | After wins | Low to medium | No | Partially |
| Kelly Criterion | Based on edge | Calibrated | No | Yes |
| Count-based spreading | Based on count | Calibrated | No | Yes |
The pattern is consistent. Every system that works performs better with perfect strategy underneath it. The systems that claim independence from strategy don’t hold up under mathematical scrutiny.
Doubling Down and Splitting
Perfect strategy doesn’t just affect how much you bet before a hand starts. It also determines when you increase your bet mid-hand through doubling down and splitting pairs. These are bet sizing decisions that only make mathematical sense when applied correctly.
Doubling down means doubling your bet in exchange for exactly one more card. Basic strategy identifies the specific moments where this is mathematically profitable, typically when you hold a total of 9, 10, or 11 and the dealer shows a weak upcard.
Splitting pairs means dividing a pair into two separate hands, each with its own full bet. This can turn one unfavorable hand into two profitable ones when done correctly. The expected value difference between correct and incorrect doubling/splitting decisions is significant.
| Situation | Correct Action | Expected Value | Incorrect Action | Expected Value |
|---|---|---|---|---|
| Player 11 vs Dealer 6 | Double down | +54% | Hit only | +35% |
| Player A-A vs Dealer 6 | Split | +63% | Hit only | +18% |
| Player 8-8 vs Dealer 10 | Split | -48% | Stand | -54% |
| Player 10-10 vs Dealer 5 | Stand | +68% | Split | +56% |
Every incorrect mid-hand bet sizing decision chips away at your expected value. Over hundreds of hands, these small differences compound into significant losses. And since doubling and splitting involve larger amounts of money on the table, getting them wrong is especially costly.
Conclusion
Systems, progressions, and gut-feel adjustments can attract attention because they feel like they are doing something. They may seem to work sometimes. But the math can be unwavering.
Bet sizing adjustments that hold up only function correctly when built on a foundation of perfect strategy. Thankfully, this perfect strategy is learnable. Basic strategy cards are available at most casinos and can be used openly at the table. Optimal video poker holds can be practiced online. Craps odds bets are simple to understand. The players who combine available tools with thoughtful, mathematically grounded bet sizing can give themselves the best possible chance at the table.