- Joined
- Jul 11, 2008
- Messages
- 1,571
- Reaction score
- 184
- Points
- 63
This guide is for bettors who are using variable staking strategies and wondering why their bankroll is more volatile than it should be, or anyone who wants to understand why the simplest staking approach is usually the best one.
What Flat Staking Actually Means
Flat staking is betting the same amount on every bet. Not the same percentage of your current bankroll that adjusts as you win or lose. Not larger amounts on "best bets" and smaller amounts on longshots. Just the same fixed stake every single time you place a bet.
If your unit size is £20, you bet £20 on a 5.0 shot, £20 on a 15.0 shot, £20 on a 50.0 shot. Every bet gets the same stake regardless of odds, regardless of how confident you feel, regardless of tournament importance or anything else. One unit per bet, that's it.
The reason this works is that it removes all the ways you sabotage yourself by thinking you're smarter than you actually are. You can't bet three units on something because you're "really confident" and then watch it lose while your one-unit bets win. You can't convince yourself that this particular 80.0 shot "feels different" and deserves a bigger stake. The system makes decisions for you so your gut feelings can't wreck your bankroll.
Why Confidence Is Worthless as a Staking Guide
You feel really confident about a bet. You've done the research, you've watched the player, you've checked the course fit, everything lines up perfectly. So you bet three units instead of one because this is clearly a better spot than your normal bets. Then the player shoots 73-76 and misses the cut and you've just lost three times what you needed to.
The problem isn't that you were wrong about the bet being good. Maybe it was good, maybe the player just had a bad week. The problem is you convinced yourself you could measure how much better it was than other bets, and you can't. You don't know if this bet is 3% edge or 8% edge or 2% edge. You're guessing based on how confident you feel, and feelings are terrible proxies for actual probability.
I see this constantly on forums where someone posts their "lock of the week" or "max bet" or whatever they're calling their most confident play. Half the time these lose at worse rates than their regular bets because the confidence came from confirmation bias or recency bias or just being excited about a narrative. Confidence and edge are not correlated the way you think they are.
Flat staking removes this entire problem. You can still feel confident about bets, you can still do your research, but you're not acting on that confidence by risking more money. Every bet gets the same stake, which means over a large sample your actual edge will show through without being distorted by the bets you happened to feel most confident about.
Golf Has Enough Variance Without You Adding More
Golf tournaments are decided by tiny margins. A player three-putts once in four rounds and finishes second instead of first. Another player makes an eagle on Sunday and jumps from 15th to 5th. A gust of wind catches a ball on 17 on Saturday and it goes in the water instead of landing on the green. This variance is inherent to the sport and there's nothing you can do about it.
Variable staking adds another layer of variance on top. Now you're not just dealing with which bets win and lose, you're dealing with whether the bets you sized larger happened to win or lose. You can have a week where you go 3-2 on your bets but lose money because the two losses were three-unit bets and the three wins were one-unit bets. That's not bad luck in the normal sense, that's you creating extra variance by varying stakes.
Flat staking doesn't eliminate golf's variance but it stops you from compounding it. Your results are purely a function of which bets win and lose at what odds, not a function of which ones you arbitrarily decided to size differently.
Some people think this means flat staking leads to slower bankroll growth. That's only true if you actually have the ability to correctly identify which bets have bigger edges and size them larger. Most bettors can't do this reliably. They think they can, but when you track it properly their "best bets" don't outperform their regular bets by enough to justify the extra variance.
Kelly Criterion Sounds Smart Until You Try Using It
Kelly criterion is a mathematical formula for optimal bet sizing based on your edge and the odds. It tells you to bet a percentage of your bankroll proportional to your edge. If you have 5% edge on a 3.0 shot, Kelly says bet X percent of your bankroll. Sounds perfect, right? Mathematical certainty about optimal stake sizes.
Except you don't know your edge. You think you know your edge but you're guessing. If you guess wrong - and you will guess wrong constantly because edge estimation in golf is nearly impossible - Kelly either tells you to bet too much or too small. Overestimate your edge slightly and Kelly has you betting way too much and your bankroll gets shredded by variance. Underestimate and you're betting too conservatively to make meaningful profit.
People use fractional Kelly (half Kelly, quarter Kelly) to address this problem, which basically just means "bet less than full Kelly recommends because we don't actually trust our edge estimates." At that point you're just using a complicated formula to arrive at something that looks suspiciously like flat staking with extra steps.
The math behind Kelly is sound. The problem is the inputs are garbage because you don't have accurate edge estimates. Garbage in, garbage out. Flat staking sidesteps this entirely by saying "I don't know my exact edge, so I'm betting the same on everything and letting the results tell me if I have an edge."
Edge Estimation Is Basically Impossible in Golf
How do you know if a bet has 3% edge or 6% edge? You've done your research, you think the player is mispriced, but translating that analysis into a specific percentage edge is guesswork dressed up as precision.
Maybe you think a player should be 12.0 but he's available at 15.0. Is that 25% edge? How do you know he should be 12.0 and not 13.0 or 11.0? You're basing this on your assessment of his form, course fit, field strength, and about fifteen other variables that all have uncertainty baked in. Pretending you can estimate edge to within a percentage point is absurd.
This is why Kelly doesn't work in practice for golf betting. You'd need superhuman ability to estimate win probabilities accurately, and you don't have that. No one does. The best you can do is identify bets where you think the odds are wrong, bet a consistent amount on all of them, and hope your process is sound enough that you show profit over time.
The Units Argument Misses the Point
Some bettors say they flat stake but they're betting in "units" where different bets get different unit sizes. A confident bet is three units, a normal bet is one unit, a speculative bet is half a unit. They call this flat staking because each "type" of bet uses a flat stake.
That's just variable staking with semantic games. You're still varying your stake based on confidence, you've just created categories to make it feel more systematic. The problems remain - you're risking more on bets you feel confident about without actually knowing if those bets have bigger edges, and you're creating extra variance in your results.
True flat staking means one bet equals one unit, always. Not one bet equals one to three units depending on how you're feeling about it. The discipline comes from removing your ability to make stake size decisions in the moment based on feelings that are probably wrong.
Bigger Stakes on Longshots Makes No Sense
There's a school of thought that says you should bet more on longshots because if they hit you make more profit. Someone betting £10 on a 50.0 shot makes £500 if it wins, but that's too small to matter, so they bet £25 instead to make it worthwhile.
This logic is backwards. You don't bet more on longshots to make the win matter more. You're betting longshots because you think the odds are wrong and the player has more chance than the market thinks. If you're right about that edge, the same unit stake works fine. If you're wrong, you've lost a bigger stake on a bet that was probably wrong anyway.
Betting bigger on longshots also means you need much better win rates on those bets to break even. If you're betting three units on 50.0 shots instead of one unit, you now need to hit at 6% instead of 2% to break even. Have you done the math to verify your longshot selection is good enough to handle that threshold? Probably not.
Flat staking avoids this trap. Bet one unit on the 5.0 shot, one unit on the 50.0 shot. Let the odds do the work. If you're selecting good longshots, you'll profit over time without needing to size them larger. If you're not selecting good longshots, betting bigger on them just loses money faster.
When Your Bets Win Doesn't Correlate with Bet Quality
Variable stakers tend to remember the times they sized up a bet and it won big. "I knew that was a great spot so I bet three units and he won at 25.0, made 75 units in one bet." They forget the three times they did the same thing and lost, or the times they sized down a bet that would've won huge.
Your memory is biased toward confirming that you're good at identifying which bets deserve bigger stakes. You remember the successes and dismiss the failures as bad luck. Flat staking forces honesty by making all bets equal weight. You can't cherry-pick which results to emphasize because every bet counted the same.
This also makes analyzing your results easier. If you're flat staking and profitable, you know your process works and you can stick with it. If you're variable staking, you have to figure out if you're profitable because you're good at selecting bets or because you got lucky with which bets you happened to size larger. That's much harder to untangle.
Psychological Benefits Are Underrated
Flat staking is easier mentally. You don't agonize over how confident you should be about each bet. You don't second-guess yourself after a loss wondering if you should've bet less. You don't torture yourself after a win thinking you should've bet more. The stake was one unit, same as every other bet, move on.
This reduces decision fatigue. Every time you make a bet you're making one decision - is this bet good enough to place? That's it. You're not also deciding how good it is and whether it deserves a bigger or smaller stake. Removing that second decision means you're not burning mental energy on something you're probably bad at anyway.
It also reduces tilt. When you lose a three-unit bet you told yourself was a "lock," the emotional hit is worse than losing a one-unit bet. You feel stupid, you question your process, maybe you chase losses to make up for it. Flat staking smooths out the emotional swings because no single loss is devastating even if the stake size was the same as every other loss.
Look, losing still sucks. I'm not saying flat staking makes losses painless. But it makes them more manageable because you haven't attached extra significance to specific bets by sizing them larger.
It Forces You to Improve Bet Selection
When you can't vary stakes to express confidence, you have to get better at only betting when you actually have an edge. Variable stakers sometimes place marginal bets at small stakes figuring "it's only half a unit, worth a shot." Flat stakers can't do this without noticing their results are getting worse.
If every bet is the same size, every bet matters equally to your bottom line. This incentivizes being more selective. You can't hide weak bets behind small stakes. Either the bet is good enough to place at your standard stake or you shouldn't place it at all.
Over time this improves your process. You get better at identifying which spots are actually good versus which spots just feel good. You stop betting because you're bored or because a tournament is on TV. You bet because you've found something worth betting, and you bet your standard amount because that's what the process says.
The One Reasonable Exception
Each-way betting in golf sometimes requires adjusted stakes because the structure is different. If you're betting each-way with 1/5 odds for top-5, you're essentially placing two bets - one on the player to win, one on them to finish top-5 at 1/5 the odds. Some bettors adjust their stake size for each-way bets to account for this split.
That's fine. The principle of flat staking is about not varying stakes based on confidence or perceived edge size. Adjusting stakes for each-way versus win-only betting is about structure, not psychology. As long as you're consistent about how you handle each-way stakes, you're not violating the core idea.
Everything else though - betting more when confident, betting less when uncertain, sizing up longshots, using Kelly criterion with made-up edge estimates - that's all adding complexity and variance without improving results.
FAQ
Doesn't flat staking mean slower bankroll growth?
Only if you can actually identify which bets have bigger edges and size them appropriately. Most bettors can't do this reliably. When you track it properly, their "confident" bets don't outperform enough to justify the extra variance. Flat staking might grow slower in theory, but in practice it often grows faster because you're not creating extra variance that destroys your bankroll.
What about betting more on better odds?
The odds already account for value. If you're betting a 20.0 shot at one unit, you're getting more potential return than a 5.0 shot at one unit. You don't need to bet more on the longshot to make it worthwhile. If you think it's a good bet, the odds will reward you if you're right.
How do I know what unit size to use?
Standard recommendation is 1-2% of your bankroll per bet. If you've got a £1000 bankroll, betting £10-20 per bet is reasonable. Adjust based on how many bets you're placing and how much variance you can tolerate, but once you pick a unit size stick with it for every bet.
Last edited:
