MLB Bankroll Management: Sizing Your Staking Units

Staking Models: Flat, Percentage, and Fractional Kelly Systems

The first MLB season I bet from the UK I had no bankroll plan. I had a betting account with a number on it, and when I liked a game I put on whatever felt right. Sometimes £5. Sometimes £20 if I felt confident. Once, gallingly, £80 on a “lock” that lost in extra innings. By August my balance was a fraction of where it started and I could not work out, looking at the bet history, whether I had been unlucky or just bad. The stakes were all over the place.

The whole point of bankroll discipline is to remove that fog. When every bet is sized as a fixed percentage of your bankroll, the question “am I unlucky or bad?” becomes answerable. Variance shows up as a smooth curve around zero. Edge shows up as a slow, steady drift in one direction. Mix the two together with random stake sizes and you have no signal at all – just noise.

The UK gambling market exists, in part, because most punters do not run bankrolls. Andrew Rhodes, the Chief Executive of the UK Gambling Commission, pointed out in early 2025 that total gross gambling yield had reached its highest ever level at £15.6 billion, with participation stable at 48% of the adult population. That headline number is built on millions of casual stakes placed without any unit framework at all. A serious MLB punter operates entirely differently, and the difference shows up in the bottom line after a year.

To ensure your staking plan works long-term, you must maintain an accurate MLB bet tracking spreadsheet throughout the entire 162-game season.

Defining your unit against a 162-game schedule

A “unit” is one fixed percentage of your bankroll, used as the standard stake on a typical bet. Most serious bettors set their unit between 1% and 3% of bankroll. I work with 1% for normal bets and stretch to 2% for higher-conviction setups, and that is at the more conservative end. New bettors should anchor at 1% and not move from it for at least a full month.

Why these percentages? Because MLB is a high-variance sport. Across a season of betting, even a punter with a real 55% win rate will experience losing streaks of seven, eight, nine bets in a row. The maths is uncomfortable. With a 1% unit, a nine-bet losing streak costs you roughly 9% of bankroll – a manageable drawdown. With a 5% unit, the same streak costs you 45%. A 50% drawdown is psychologically devastating and effectively ends most punters’ attempts at disciplined betting.

For a £200 starting bankroll, a 1% unit is £2. That number sounds laughably small to anyone whose mental reference point is the high-rolling US picks community. The point is that the unit is a percentage, not an absolute number. When your £200 grows to £400 over six months of careful betting, your unit grows to £4 automatically. When it shrinks to £150 after a rough patch, your unit shrinks to £1.50. The system contains its own protection.

The 162-game schedule matters here because it shapes the betting volume realistically available. If you select carefully – say, three to four genuine value bets per day across the slate – you will place roughly 500-700 bets across a six-month regular season. At 1% per bet, that is a turnover of five to seven full bankrolls. The edges have time to compound. With a 5% unit, the same turnover hits 25-35 full bankrolls of action, and variance will eat you alive before edge has time to show up.

Flat staking vs percentage staking

Flat staking means you stake the same absolute amount on every bet – say, £5 – regardless of how your bankroll changes over time. Percentage staking means you stake the same percentage of your current bankroll. Both have a place, but they behave very differently in practice.

Flat staking is the simpler system and a perfectly defensible choice for a beginner. The advantage is that your record-keeping is cleaner: every bet is the same size, so ROI calculation is trivial and comparison across bets is direct. The downside is that flat staking does not protect you on the way down or compound for you on the way up. A losing streak that takes your bankroll from £200 to £100 leaves you betting the same £5 stakes – now 5% of bankroll instead of 2.5%. The risk has doubled without you doing anything.

Percentage staking – recalculated weekly or monthly – fixes this. After every Sunday, you check your bankroll, recompute 1% of it, and that is your unit for the coming week. The system is self-correcting. The downside is more bookkeeping and a slight psychological friction: your stake sizes change every week and it can feel less stable.

The compromise I use is monthly recalculation. The first of every month, I look at the bankroll, work out 1% of it, and that becomes the unit size for the next thirty days. Bets within the month are flat at that unit. This gives you most of the protective benefits of percentage staking without the constant fiddly recalculations.

Whichever system you choose, write it down. The system has to exist outside your head, in a document or spreadsheet you can show yourself. Because three months in, when you are on a five-bet losing streak and convinced the next game is a sure thing, you will absolutely try to talk yourself into a bigger stake. Writing the system down is what lets your sober Tuesday-morning self overrule your impatient Saturday-night self.

Kelly criterion in plain English

The Kelly criterion is a stake-sizing formula designed to maximise long-term bankroll growth. The maths is two centuries old in spirit and seventy years old in its modern form. It is also dangerous if applied naively, which is why almost every serious bettor who uses Kelly uses a fractional version of it.

Full Kelly says: stake the percentage of your bankroll equal to (probability of winning multiplied by decimal odds, minus 1) divided by (decimal odds minus 1). So if you think a bet has a 55% chance to win at 2.00 decimal odds, full Kelly says: (0.55 × 2.00 – 1) ÷ (2.00 – 1) = (1.10 – 1) ÷ 1 = 0.10, which is 10% of bankroll on a single bet. If your probability estimate is wrong by even a few percent, that stake is enormous.

This is why fractional Kelly exists. Half-Kelly means you stake half the full-Kelly recommendation. Quarter-Kelly means a quarter. Most serious bettors I have spoken to use a quarter to half. With quarter-Kelly, the above example becomes 2.5% of bankroll – closer to a normal unit. The trade-off is slower growth in exchange for dramatically more stable equity curves.

The deeper reason fractional Kelly works better than full Kelly for MLB punters is that your probability estimates are not as accurate as you think they are. Full Kelly assumes your edge calculations are exactly right. Real probability estimates carry uncertainty – sometimes you are off by 2%, sometimes by 5%, sometimes by 10%. Fractional Kelly stakes shrink proportionally, which gives you a margin of safety against your own estimation error. It is a discipline that punishes overconfidence and rewards humility.

A practical compromise: do not run Kelly on individual bets. Use a flat 1% unit for everything except your highest-conviction setups, where you bump to 1.5% or 2%. This gives you most of the value of stake-sized confidence without the false precision of plugging probability estimates you cannot really defend into a sensitive formula.

Monthly reviews: ROI, CLV and emotional drift

The discipline is not just in placing the bets. It is in reviewing them honestly once a month. Without a monthly review you are flying blind. The review takes maybe thirty minutes and asks four questions.

First, what was your ROI? Total profit divided by total stakes. A casual MLB punter with no edge will hover around -3 to -5% ROI long-term. A genuinely sharp bettor might run +2 to +5% over a year. Anything above +5% sustained for an entire season is exceptional. If your monthly ROI is +25%, you are not a genius – you are running hot, and the regression is coming.

Second, what was your closing line value? For every bet, your decimal price taken versus the decimal closing price. If you consistently beat the close on most bets, you have positive CLV. If you consistently get worse prices than the close, you have negative CLV. CLV is a more honest verdict on your selection process than results, because results are tangled up with variance.

Third, which kinds of bets generated CLV and which did not? Filter your log by market type. Did your run line bets beat the close more than your moneyline bets? Did your strikeout props perform better than your home run props? Are there market segments you should drop entirely because you cannot demonstrate edge there?

Fourth, did you size every bet correctly? Sort your log by stake size relative to unit. If you find bets at 1.5x unit that you cannot remember explicitly upgrading, your system is being eroded by emotional drift. Every bet at a non-standard size should have a written justification at the moment of placement. If half your “upgraded” bets do not have one in the notes column, you have a discipline problem more than a model problem.

I keep my monthly reviews in a single spreadsheet that goes back years. The cumulative picture is more sobering than any single month. Bad habits show up as recurring patterns. Real edges show up as slow upward drifts. Both pieces of information are invaluable, and you cannot get either without writing things down. The spreadsheet itself is what we cover in the bet log template UK punters actually need.

Protecting your funds is just as important as finding value on the best MLB betting site in the UK.

The patience that the bankroll demands

The cruellest thing about bankroll discipline is that it works on a longer timescale than most punters have patience for. A 1% unit feels small. A 1% edge per bet looks unimpressive on a spreadsheet. Across one weekend, none of this seems to matter. Across a season, it is the entire reason you finish in profit or in losses. The MLB regular season is roughly six months long, almost exactly enough time for variance to wash out and edge to show. A punter who survives that six months with stake sizes intact and a complete bet log can, for the first time, ask honestly whether they are good at this. A punter who breaks unit sizing the first time they go on a losing run will never get to that question, because the bankroll will be gone before the answer arrives.