Year-End Review and Optimization

Once a year you have enough sample to draw a real conclusion. Most bettors skip the review anyway, because the most useful thing it produces is usually an answer they did not want.

Monthly reviews, as covered in Post 85, can only assess process. A month is not a sample and results across one are noise.

A year is different. Several hundred bets, multiple sports, a full cycle of drawdowns and recoveries. That is enough to answer questions a month cannot: which segment actually produced, whether the edge is real, and whether the hours were worth it.

This post is the procedure. Ten steps, one sitting, once a year.

01When and Why

Do it at the end of your betting year rather than the calendar year, if those differ. For most American bettors the natural point is after the NBA and NHL finals in June, when the sports calendar is at its thinnest.

Do it in one sitting, with the ledger open, before the next season pulls your attention forward. A review deferred until October is a review that does not happen.

And do it in writing. A review you conduct in your head will conclude that things went roughly as expected, because that is what memory produces.

02Step One: Compute the Real Numbers

From the ledger, not from memory:

• Total bets placed. Your actual sample size.
• Total staked in units and dollars. This is your turnover, and it is the number the arithmetic in Post 92 runs on.
• Net profit or loss in units and dollars.
• Return on investment. Profit divided by total staked.
• Win rate, excluding pushes.
• Average price taken. Which gives you your actual break-even threshold.
• Gross winnings and gross losses, separately. These are the tax figures from Post 83, and you will need them.

Compare the ROI to the 2 to 4 percent band this series has used throughout. Above 4 percent across a large sample is exceptional and worth examining for whether it is real. Below zero is common and is not a reason to stop, but it is a reason to keep reading.

03Step Two: Segment Everything

The aggregate number hides the useful information. Break it down by:

Two cautions. Segments need sample too. Forty prop bets at 62 percent is noise, and treating it as a discovered edge is the overfitting problem from Post 76 applied to your own history.

And resist slicing until something looks good. If you cut the data enough ways, some cut will appear profitable by chance alone. Decide your segments before you look.

04Step Three: The CLV Analysis

The most important step, and the one that tells you whether anything else in the review means what it appears to.

Compute the percentage of bets where you beat the closing number, overall and by segment. Then read the combination:

The second row is the one that costs people the most, because a profitable year with poor closing line value feels like validation and is actually a warning. Per Post 91, that is a drawdown you have not experienced yet.

05Step Four: Process Compliance Audit

If you kept the compliance field from Post 90, this is quick. If not, add it this year.

What percentage of bets followed your own rules on sizing, leagues, caps, and the evaluation sequence? Then, more usefully: did the non-compliant bets perform differently?

Almost universally they do, and worse. Seeing that in your own numbers is more persuasive than any argument, and it is the single most behavior-changing output of an annual review.

Also look at when compliance slipped. It usually clusters, and the clusters usually correspond to drawdowns or to periods when you were betting outside your normal leagues.

06Step Five: The Drawdown Review

From the cumulative units chart:

• Deepest drawdown, in units and dollars.
• Longest drawdown, in weeks.
• How many drawdowns exceeded ten units.
• What you did during the worst one. Sized up? Changed method? Held?
• What your CLV did during it. This is the retrospective version of the diagnostic in Post 91.

Record these permanently. Across several years this becomes your own reference for what normal looks like, which is worth more during the next bad stretch than any general reassurance.

07Step Six: The Hourly Rate

The calculation most bettors avoid, and the most clarifying one available.

Estimate the hours you spent across the year: research, projections, line monitoring, execution, record keeping, account administration. Be honest. Then:

Net profit ÷ hours = your effective hourly rate.

A bettor who cleared $3,000 across a season while spending ten hours a week for eight months has worked roughly 320 hours for about nine dollars an hour.

That number is not an argument against doing this. Plenty of worthwhile activities pay nothing, and the intellectual return is real. But it should be known rather than assumed, particularly by anyone entertaining the questions in Post 92.

08Step Seven: Account Health

From Post 84, an annual inventory:

1. Which accounts remain at full limits?
2. Which have been restricted, and roughly when?
3. What is your actual maximum bet across the portfolio today?
4. Are there operators in your state you have not opened?
5. Is your bankroll distributed sensibly across what remains?

This is also the moment to decide, deliberately, where you sit on the extraction-versus-longevity trade from Post 84. Another year of data should inform that choice rather than leaving it to drift.

09Step Eight: Decide What to Cut

The most valuable output of the entire exercise, and the one people resist.

Candidates for elimination:

• A sport with negative ROI and poor CLV across adequate sample. Not a bad year. A demonstrated absence of edge.
• A market type that consistently underperforms. Props are the most common offender.
• Recreational betting mixed into the ledger. Parlays and novelty markets belong in a separate budget, per Post 86.
• A league you follow out of habit rather than edge.
• Tools and subscriptions whose cost exceeds their demonstrable contribution.
• The lowest-return time sink identified in Step Six.

Cutting is harder than adding because it feels like retreat. It is the opposite. A bettor who eliminates a segment that lost money for two consecutive years has improved their expected return without learning anything new.

10Step Nine: Add One Thing

One. Not a list.

Multiple simultaneous changes make the next review uninterpretable, because you will not know which change caused what. This is the same discipline as the modeling advice in Post 76: change one variable, observe, then change another.

Reasonable single additions:

• A new league within a sport you already cover well
• A market type you have been avoiding, at reduced size
• An improvement to your projection method
• An additional book, if account access is your constraint
• A tracking field you wish you had this year
• The accountability relationship from Post 93

11Step Ten: Set Next Year's Targets

Targets should be about process, because process is what you control. Results are downstream and noisy.

Useful targets:

• CLV percentage above a specific threshold
• Process compliance above a specific percentage
• Shopping gap at or near zero
• A bet volume range, with a floor and a ceiling
• Monthly reviews completed on schedule
• Records maintained without gaps

Notice that ROI is absent. You cannot control ROI in a season. You can control every item above, and doing so is what produces ROI eventually.

12The Emotional Difficulty

Worth naming, because it is the actual reason most people do not do this.

An honest annual review can tell you that a year you felt good about was luck, that a segment you enjoy loses money, that your hourly rate is poor, or that after two years you have no demonstrated edge.

Those are unpleasant findings and they are the valuable ones. A review that only confirms what you believed has told you nothing.

The reframe that helps: the numbers describe the past, not you. A negative year with strong CLV is information about variance. A negative year with weak CLV is information about a method, which is a thing you can change. Neither is a verdict on whether you are capable of doing this.

And if the finding after several honest years is that there is no edge, that is genuinely useful too. Knowing that lets you decide whether to keep going as recreation, sized and budgeted accordingly, which is a perfectly good outcome. The alternative is continuing to believe an edge exists while paying for the belief indefinitely.

13Common Review Mistakes

• Skipping it. The default, and it guarantees the same year repeats.
• Reviewing results only. Without CLV, profit tells you little.
• Slicing until something looks good. Decide segments before you look.
• Drawing conclusions from small segments. Forty bets is not evidence.
• Changing five things. Next year becomes uninterpretable.
• Adding without cutting. Cutting is where the return is.
• Skipping the hourly rate. The most clarifying number available.
• Setting result targets. Process is what you control.

14The Bigger Picture

Everything in this series has been aimed at making betting a process with feedback rather than a sequence of opinions. The annual review is where that feedback finally arrives with enough sample to be trusted.

It is also the only mechanism by which this activity compounds. A bettor who reviews honestly each year and cuts one weak segment while adding one improvement is meaningfully better in five years than they were at the start. A bettor who does not is running year one repeatedly with different teams.

That difference is invisible in any single season and enormous across a decade, which is a reasonable description of most things in this section.