Tracking Your Bets: The Metrics That Matter
What to record for every bet, which metrics are worth calculating, and why win rate is the most misleading number in sports betting.
5 min read
Human memory of gambling results is not neutral. Wins are specific and retold; losses compress into a vague sense of having been unlucky. Ask a hundred bettors how they did last season and the average answer will be comfortably above the true average, which is arithmetically impossible.
A record fixes this, and it is the only thing that does.
What to log
Eight fields, and the first six are non-negotiable.
| Field | Why |
|---|---|
| Date | Lets you segment by season, month or form |
| Sport and market | Shows where your results actually come from |
| Selection | What you backed |
| Odds taken | Everything else is calculated from this |
| Stake | Required for ROI, and for honesty about sizing |
| Result and return | Won, lost, pushed, voided, and what came back |
| Closing price | Enables closing line value, the most useful early signal |
| Reasoning | One line. Invaluable six months later |
The reasoning field is the one people skip and later wish they hadn't. "Backed the under because both defences rank top five" is a testable claim. A record of selections with no reasoning can tell you that you lost on NBA totals, never why.
ROI, not win rate
Win rate is wins divided by settled bets. It is the number everyone quotes and it is nearly content-free on its own, because it says nothing about price.
- 65% on -250 favourites:
(0.65 × 40) − (0.35 × 100) = −$9.00per $100. Losing. - 40% on +200 underdogs:
(0.40 × 200) − (0.60 × 100) = +$20.00per $100. Strongly profitable.
The 40% bettor is doing far better than the 65% bettor. Win rate cannot see that.
ROI — also called yield — can:
ROI = total profit ÷ total staked
$640 profit on $16,000 staked is a 4% ROI. It accounts for both how often you win and what the wins pay, and it is comparable across bettors with different staking sizes.
What a good ROI looks like
Lower than the internet suggests.
| ROI over a large sample | Interpretation |
|---|---|
| Below 0% | The overwhelming majority of bettors |
| 0% to 2% | Possibly a small genuine edge, possibly noise |
| 3% to 5% | Strong, and sustainable by a disciplined bettor |
| 5% to 10% | Exceptional, usually in niche or slow markets |
| Above 10% long-term | Almost always a small sample or a selective record |
The reason the ceiling is low is in the pricing. A -110 market has about 4.5% of margin in it. To net 5% after paying that, you have to be roughly ten percentage points better than the price before costs — a very large claim.
Closing line value
The problem with ROI is that it takes thousands of bets to become reliable. CLV is the workaround.
Compare the price you took to the price the market closed at. Took +150 on something that closed +130? You beat the close. Took -110 on something that closed -125? Same.
It works because the closing line — particularly at high-limit books, after all the information and money has arrived — is the best publicly available estimate of a true probability. Consistently taking better prices than that is very hard to do by luck.
Crucially, every bet produces a CLV reading, including bets that lose. That turns a sample of dozens into something interpretable, rather than needing a sample of thousands.
Segmenting the record
Once there is enough data, break it down — carefully.
Useful cuts: by sport, by market type, by odds range, by whether you bet early or late, by favourite versus underdog.
The warning that has to come with this: slicing a small sample produces findings that are not there. Split 300 bets eight ways and the best-looking slice will look excellent for purely random reasons. Treat any segment under a couple of hundred bets as a hypothesis to watch, never a conclusion to act on.
The point of the exercise
A record is not a scoreboard. Its value is that it makes specific questions answerable: am I actually profitable on NHL totals, or do I just remember the good ones? Is my parlay habit costing me what the maths says it should? Are my late bets worse than my early ones?
None of those can be answered from memory, and all of them change what you do next.
Related reading
- Expected value in sports betting — what the record is ultimately measuring against.
- Line shopping: why the best price matters more than the best pick — the easiest way to improve the prices your record stores.
- How to remove the vig from betting odds — the margin your ROI is fighting.
Frequently asked questions
- What should I record for every bet?
- Date, sport, market, the selection, the odds you took, your stake, the result and the settled return. Adding the closing price and a short note on your reasoning makes the record far more useful later.
- Is ROI or win rate the better measure?
- ROI. Win rate ignores price, so a 65% win rate on heavy favourites can lose money while a 40% win rate on underdogs can be strongly profitable. ROI accounts for both how often you win and what the wins pay.
- What is a good ROI in sports betting?
- Sustained ROI above about 3% to 5% over a large sample is genuinely strong. Claims of 20% or more over long periods are almost always either a small sample, a selective record, or untrue.
- What is closing line value?
- Closing line value measures whether the price you took was better than the price the market settled at before the event started. Consistently beating the closing line is the strongest early evidence that a betting process is sound.
Tags
About the author
BestSlip Analytics Desk
Odds & Analytics
The analytics desk covers the measurement side of betting — implied probability, vig removal, expected value, return on investment and sample size. The work is descriptive: it explains how a number is calculated and what it can support, never what anyone should wager.
All articles by BestSlip Analytics Desk