If you do not track your bets, you do not know whether you are good. You have a story you tell yourself, and that story is almost always wrong. Memory inflates wins and discounts losses. The bet you "knew" was a lock you forgot about until you hit it. The 0-4 Sunday gets blamed on bad luck instead of bad lines. A spreadsheet does what your brain refuses to do: it tells you the truth. This guide walks through the columns every bettor's log needs, the metrics it should compute automatically, and the mistakes that make most bet logs useless.
Why tracking matters more than you think
Casinos run on the assumption that the average customer cannot tell whether they are winning. That assumption is correct for the average sports bettor too. Without a log:
- You will overrate your favorite sport and team because winners are sticky in memory.
- You will underestimate parlay losses because they happen quietly.
- You will not know if your edge is real or a hot streak.
- You cannot adjust unit size, fade weak markets, or double down on real edges — because you do not know which is which.
Tracking is not about discipline for discipline's sake. It is the data that makes every other improvement possible.
The columns every bet log needs
You can use Google Sheets, Excel, Notion, or a paid app — what matters is the columns. The minimum useful schema:
- Date — when the bet was placed (not the game date)
- Sport — NFL, NBA, MLB, NHL, soccer, MMA, etc.
- Market — spread, total, moneyline, player prop, parlay, futures
- Selection — what you actually bet (e.g., "Lakers -3.5", "LeBron over 26.5 pts")
- Line/Odds — the price you got
- Closing line — the line at game start (critical, do not skip)
- Stake (units) — in units, not just dollars
- Stake ($) — actual dollar amount
- Result — win/loss/push
- Profit/loss ($)
- Source — model, tout, gut, friend
- Notes — weather, injuries, why you played
Without the closing line, the log is decorative. With it, you can answer the only question that matters: am I beating the close?
Closing line value (CLV): the metric that decides everything
The closing line is the sharpest version of the line — every piece of information that hits the market is in there. If you consistently beat the close, you are getting better prices than the market consensus, which is a strong leading indicator that you are a long-term winner. If you consistently lose to the close, you are paying a tax on every bet, edge or not.
Example: you bet Eagles -3 (-110). The line closes at Eagles -3.5 (-110). You beat the close. Even if the Eagles lose by 4 and you lose this bet, your process got you a better number than the market settled on, and over hundreds of bets that translates to profit.
Two ways to express CLV:
- Points beaten: simple subtraction. You took -3, line closed -3.5, you beat by 0.5 points.
- EV beaten: implied probability of your line minus implied probability of the close, devigged. Slightly more accurate, especially on moneylines.
Logging both lines and the difference is the highest-leverage two cells in your spreadsheet.
The metrics your spreadsheet should compute automatically
- Total bets by sport, market, and source
- Win rate by category
- ROI = total profit / total amount risked
- Units won/lost
- Average CLV by sport and market
- Performance vs unit size — do your one-unit plays beat your two-unit plays?
- Bet count by stake size — are you flat betting like you think you are?
If your tracker does not compute these, it is a list, not a tool.
A worked monthly readout
After a 100-bet month, your sheet might say:
- Bets: 100
- Risked: $2,000
- Profit: +$140
- ROI: 7.0%
- Win rate: 56%
- Average CLV: +0.4 points
That is a good month. But the slice that matters: NFL spreads were +0.6 CLV and 60% win rate. NBA player props were -0.2 CLV and 49% win rate. The aggregate looks great, but you are bleeding edge on props and printing on spreads. That is the actionable insight a log gives you. Without it, you have "had a decent month."
Common bet log mistakes
- Skipping the closing line. The single most-skipped column, and the only one that proves edge.
- Logging in dollars only. Units make performance comparable across bankroll changes.
- Lumping all parlays into one row. Track each leg separately if possible — it changes which legs are actually profitable.
- Forgetting to log losses. Skipping the ugly ones is the most common form of self-deception.
- Reviewing weekly, sizing daily. Look at the data monthly. Don't tweak unit size after one bad weekend.
Tracking model bets vs gut bets
If you are running models on Tinker, your model picks are already logged with predicted probability and CLV automatically tracked. The bigger lift is logging your off-model bets — the gut plays, the "I just liked it" wagers, the friend's tip you took. You will almost always discover that your gut bets are dragging your overall numbers down. That data is worth its weight in bankroll.
You can sanity-check your model's calibration directly on the leaderboards before betting it.
What to do with the data
After 200 to 500 logged bets, you have enough sample to act:
- If a market category is below water on CLV, stop betting it.
- If your high-confidence plays don't beat your routine plays, go flat-bet.
- If one sportsbook gives consistently better prices, route more action there.
- If you are net positive on CLV but negative on profit, your sample is just too small — keep going.
Bottom line
Tracking your bets is the cheapest, highest-EV thing you can do as a bettor. Build a sheet with the columns above, especially the closing line. Compute ROI, win rate, and CLV by slice. Review monthly, not daily. Log the ugly bets along with the pretty ones, and let the spreadsheet tell you what is and isn't working.
If you'd rather skip the spreadsheet entirely for model bets, the dashboards on NFL picks, NBA picks, and MLB picks already track results, CLV, and ROI per model. The hardest part — being honest with yourself about your numbers — is no longer optional.
Bet responsibly — set limits, never chase losses.