I Kept a One-Week CLV Diary and Became Less Fun

I named the spreadsheet final-final-v3 and felt the way professionals must feel #

I made a tracking sheet on a Sunday night and named it, with no particular justification, clv-diary-week-v3-FINAL. The previous two versions were variously corrupted by my unwillingness to read about pivot tables, but version three survived contact with reality long enough to record a week. The fields I committed to were nine columns long: timestamp, sport, market, book, price at placement, my fair-line estimate, stake, closing price, result. The last column was added on Sunday night, after the games settled, and was the one I most wanted to skip.

The first lesson of the diary, which I will admit arrived before I had filled in a single row, is that the act of opening the sheet at the moment of placement is itself the discipline. I had been keeping a kind of mental CLV diary for months, the way a person keeps a mental fitness regime. The actual diary, with timestamps and price fields and an explicit column for fair-line estimate, is a different artifact. The mental version preserves the bets I am proud of and quietly disposes of the rest. The actual version preserves everything and refuses to be flattered.

The mechanic that makes the diary worth keeping #

CLV — closing-line value — is the difference between the price you got and the price the market closed at, expressed in cents. The number is interesting because it factors out the noise of any single result. A bet that wins at a price worse than the closer is a bet that beat variance more than it beat the market. A bet that loses at a price better than the closer is a bet whose process was correct even if the outcome was not. Over a single week, CLV is the only column that tells the bettor something about herself that the W/L column cannot.

The mechanic of recording CLV is straightforward and unpleasant. You write the price down at the moment of placement, because the market starts moving immediately and the only honest comparison is the one anchored to the moment of decision. You return after the game, find the closing price, and subtract. The number you get is, in any reasonable American-odds format, a small integer with a sign in front of it. Positive is good. Negative is bad. The discipline is to record the sign before you record the win or loss, so the spreadsheet cannot be edited by knowing how the game went.

The bet I was proudest of, the one I most resented, and what the diary said about both #

The bet I was proudest of that week was an NFL moneyline I placed on a road dog I had been writing about for three days. The dog won. I felt, briefly, like a person whose opinions had been validated by the universe. Then I checked the closing price. The closer had moved against me by about four cents, which meant the market had agreed with my read and the bet was a positive-CLV win. This was, on inspection, the cleanest entry in the entire diary — the price was good, the read was good, the result was good. The pride was, just barely, earned.

The bet I most resented was a college-football spread that lost on a special-teams play I do not wish to describe in detail. The closer on that bet had moved against me by three points. The diary entry, when I filled it in, was a small humiliation: I had placed the bet at a price the market had quickly identified as wrong, and the result, while painful, had merely confirmed the price. The bet was a negative-CLV loss, which is the worst kind. The diary did not let me describe it as a bad beat. The diary described it as a bad bet that had also lost.

The quadrant view and what it teaches #

The most useful view I built off the diary was the quadrant chart: bets plotted on a CLV-versus-result grid. The top-right quadrant — wins that also beat the closer — is the easiest to feel good about and contains, on average, the cleanest decisions. The bottom-right quadrant — losses that beat the closer — is the one the diary was built for. Those are the bets the bettor would otherwise have remembered as unlucky and which the closer has classified, dispassionately, as correct. The bottom-left quadrant — losses that lost to the closer — is the one the bettor should be most willing to learn from. The top-left — wins that lost to the closer — is the most dangerous, because the result tempts the bettor to remember the process as sharper than it was.

My week broke down with seven entries in the top-right, three in the top-left, five in the bottom-right, and three in the bottom-left. The single most important fact the diary surfaced was that the three top-left bets — the ones I would have remembered as smart wins — were process bets at worse-than-fair prices. If I had repeated those decisions over the rest of the season at the same volume, the wins would have decayed faster than the price could carry them. The diary is the only artifact that would have caught this. The W/L column, taken alone, was sending the wrong signal.

By Sunday night I had fewer stories and better notes. The tracking sheet was not charming company, but it kept telling the truth after the scoreboard stopped talking.

The protocol that survived the week #

The protocol I will keep, going forward, is a short one. I will write the placement price into the sheet at the moment of placement, before any other action. I will write the fair-line estimate into the sheet in the same minute, even if the estimate is rough — the rough estimate is more useful than no estimate, because the difference between the two is the closest thing to a real-time edge measurement I can produce. I will record the closing price on Sunday night, in one sitting, before the W/L results have a chance to edit my memory of the bets.

The discipline is small and the friction is real. The first week was the worst, because the sheet was unfamiliar and the placement-price column kept reminding me how lazy my mental version of the same exercise had been. By the second week the friction had reduced to about ninety seconds per bet. By the third week the sheet had become the thing I check after a game in addition to the result, which is the moment at which a tool stops being an experiment and becomes a workflow. The transition is small. It is the part of the experiment I am most surprised by.

The closing argument #

The diary made me less fun in the way that flossing makes a person less fun. The week produced fewer screenshots and more rows. The rows, taken together, told a story I would not have arrived at by memory alone. They told me the bet I was proudest of was earned, the bet I most resented was deserved, and three of my favorite wins were sitting in a quadrant that signaled trouble. None of those conclusions are dramatic. All of them are useful. They are the kind of conclusion the W/L column cannot produce, which is why the diary needs the closing price to function at all.

I will keep the sheet. I will keep the protocol. I will keep being mildly insufferable on Sunday nights, with a screen open to a column called CLV and a small quiet pleasure when the column is positive and a small quiet humility when it is not. The bankroll has not yet noticed; the diary is too young. The bettor has noticed. The bettor is the part of the system the experiment was actually designed to update, and the update has, for once, taken.