Learning goals and lesson guide
**Estimated reading time:** 7 min read.
**Level:** Intermediate. **Study time:** Allow 30–40 minutes with a timestamped worksheet. **Prerequisites:** Odds, implied probability, sources and expected value. All quoted markets in this lesson are fictional.
A team opens at 2.20 and later appears at 1.90. The direction of the price change is clear, but its cause is not established by those two numbers. A confirmed lineup update, market activity, a changed margin or other information could contribute. An analyst should describe the movement before claiming to explain it.
This lesson teaches you to define opening and closing observations, distinguish price changes from line changes and make like-for-like comparisons. You will also learn why a favourable comparison with a later market price is a useful research signal in some settings, yet not proof of a profitable forecast.
## 1. Define which opening price you mean
“Opening odds” can mean the first price offered by one operator, the first price captured by a data service or the first price available in your own records. These need not be the same observation. Name the source and timestamp rather than treating an opening label as universal.
Some markets open before key information is available and at different limits or participation levels. That context can affect how informative a price is. Do not assume an early number is inherently weak or that a later one must be correct.
For a paper study, define opening operationally: “The first available observation from the selected source after the market appears.” If earlier observations are missing, say so. A reproducible imperfect definition is better than an impressive label that cannot be audited.
## 2. Define the close consistently
A closing price might be the last pre-event quote, a quote at a fixed cutoff or a data provider's designated close. If you use the final pre-start observation, record how event delays and suspensions are handled. A quote recorded after play begins is not a clean pre-match close.
Choose a rule before examining performance. For example, use the last recorded valid quote at least one minute before the scheduled or actual start, with a documented policy for delayed events. The exact convention is a research choice, not an industry-wide constant.
Keep event identity, market, line, period and settlement conditions the same. Comparing a regulation-only close with an overtime-inclusive earlier price is not a valid like-for-like measurement.
## 3. Separate price movement from line movement
A handicap can remain at minus 3.5 while its decimal odds change from 1.95 to 1.80. That is a price change on the same line. Alternatively, the handicap may move from minus 3.5 to minus 4.5 while the displayed price remains 1.90. The threshold has changed.
Record both fields. Saying “the odds moved” without specifying which part changed can obscure the comparison. A total moving from 210.5 to 214.5 is a different proposition, not merely a new payout for the same event outcome.
To compare unlike lines, you would need a model of the underlying score distribution or available prices on a common line. Do not divide two decimal odds on different thresholds and claim the ratio measures an advantage by itself.
## 4. Describe shortening and drifting precisely
For an unchanged selection, a move from 2.20 to 1.90 is a shortening of decimal odds and a smaller conditional payout per unit. Its reciprocal implied percentage rises from about 45.45% to 52.63%.
The difference is about 7.18 percentage points. It is not proof that the true chance rose by exactly that amount. The market may contain a changing margin, and the quoted prices are observations of an offer, not direct measurements of a hidden probability.
A move from 1.90 to 2.20 is a drift to a larger conditional payout. Again, the cause needs evidence. Avoid language such as “insiders know” unless there is reliable reporting establishing relevant facts. Price direction alone cannot identify who acted or why.
## 5. Compare the full market where possible
For a two-outcome no-push example, record both sides at the same time. At prices 1.90 and 1.90, reciprocal probabilities sum to about 105.26%. At 1.80 and 2.00, they sum to about 105.56%. The distribution and total have both changed.
A proportional normalisation would divide each reciprocal by the sum to create a baseline adding to 100%. Label this as a convention, not a recovery of true probabilities. Other margin assumptions can produce different estimates.
If only one side is archived, you can report its raw movement but cannot reconstruct the entire contemporaneous market reliably. Do not fill the other side with a later quote and pretend both were simultaneous.
## 6. Interpret comparison with a closing price cautiously
Suppose a paper record captures 2.20 and the same market later closes at 2.00. The ratio 2.20 ÷ 2.00 − 1 is 10%. This is a raw price comparison under a stated convention, not automatically a 10% true expected profit.
The close might contain margin, be poorly observed or reflect information unavailable earlier. A later price can be a useful benchmark across many records, but it does not prove that every earlier selection was well founded. Nor does losing the individual event make the price comparison arithmetically false.
If a paper study reports closing-price comparisons, provide the source, cutoff, matching rules and missing-data rate. A benchmark is only useful when it is consistently measured. Do not select whichever closing source makes each record look best after the fact.
## 7. Do not infer causation from timing alone
An injury announcement followed by a price movement can be consistent with a causal explanation, but other information may have arrived at the same time. State the evidence proportionately: “The market shortened after the announcement” is an observation; “the announcement caused the entire move” is a stronger claim.
Check whether the announcement was genuinely new. It may confirm information already widely reported. A market can move before an official statement without proving wrongdoing or secret access. Avoid reputational claims based only on a chart.
Use an event log: price timestamp, line, full market where available, confirmed news timestamp and any data-quality flags. This makes the analysis inspectable and keeps narrative assumptions from becoming hidden facts.
## 8. A paper tracking exercise
Track a fictional unchanged winner market at three times: 2.40 at opening, 2.20 six hours before the event and 2.00 at the defined close. Record the reciprocal percentages and describe the direction without guessing a cause.
Now introduce a handicap example moving from plus 4.5 at 1.90 to plus 3.5 at 1.90. Explain why unchanged decimal odds do not mean unchanged terms. A four-point defeat would settle those two lines differently.
Finally, imagine the closing observation is missing for two out of ten records. Disclose the missing cases and avoid silently reporting an eight-record average as if it covered all ten. Missingness may be related to unusual events and therefore matter to interpretation.
## 9. Review questions and answers
1. Must every operator share the same opening price and timestamp? 2. What changes when a total moves from 200.5 to 204.5 at the same price? 3. Does shortening prove the true probability changed by the reciprocal-price difference? 4. What is the raw ratio comparison between 2.10 and a 2.00 close? 5. Does that ratio alone prove true expected profit? 6. Why should both sides of a market be recorded at the same time? 7. Can a post-start quote be silently used as a pre-match close?
### Answer key
1. No. Define the source and observation rule. 2. The outcome threshold, which changes the proposition being evaluated. 3. No. Margin and other pricing factors remain, and the true probability is unknown. 4. 2.10 ÷ 2.00 − 1 = 5%. 5. No. It is a benchmark comparison whose interpretation requires assumptions. 6. To inspect the contemporaneous market and avoid mixing incompatible observations. 7. No. It may include information from play and must be labelled separately.
## 10. Completion check
Produce a small price-history table with exact market definitions, timestamps, sources and missing-data notes. Your conclusion should distinguish what moved, what information was confirmed and which causes remain uncertain. That is a useful analysis even without a prediction.
Review [odds](https://betting.crazywingo.ph/article/read-betting-odds-calculate-payouts), [probability and margins](https://betting.crazywingo.ph/article/implied-probability-bookmaker-margins-beginners), [market types](https://betting.crazywingo.ph/article/moneyline-spread-total-parlay-beginners), [market identification](https://betting.crazywingo.ph/article/sports-betting-basics-for-beginners) and [budget boundaries](https://betting.crazywingo.ph/article/betting-budget-limits-avoid-chasing-losses) as needed. Next comes comparing prices when fees and settlement conditions differ.