Betting Handle Explained: Reading Money and Ticket Splits

Handle means the total amount of money staked within a stated game, market, sportsbook, or period. Also called turnover, it counts stakes regardless of whether those bets win or lose. It is not sportsbook profit, bettors’ losses, deposits, account balances, or the number of bets.

To interpret a handle figure, first check which wagers it includes and the period it covers. The same term can describe one game’s betting activity or a sportsbook’s monthly total.

Originally published: March 17, 2026.

Why can handle exceed deposits?

Handle can exceed deposits because the same money can be wagered more than once. Each accepted stake adds to turnover; it does not have to come from a new deposit.

Consider this hypothetical example without bonuses or voids:

1. You deposit $100.
2. You stake $100 at decimal odds of 2.00 and win, receiving a full payout of $200, including your stake. Your balance is now $200.
3. You wager $100 again and lose, receiving a payout of $0. Your balance is now $100.
4. You place a third $100 wager. It is still pending, leaving $0 available and $100 staked on the unsettled wager.

The three stakes produce $300 in handle. That is not $300 in new money or $300 in losses: there was one $100 deposit. The $100 profit from the first wager offsets the $100 loss from the second, and the third wager's outcome is still unknown. The first payout allowed funds to be reused for the subsequent wagers.

This is why deposits and handle are not interchangeable. Also check how the provider treats voided wagers and refunds, because reporting rules can differ.

A historical, undated operator guide, page 4, defines handle or turnover as the total amount wagered and separately explains hold. It is used here only for those definitions, not for information about current offers.

How does handle differ from tickets, revenue, and hold?

Handle measures money staked, tickets measure the number of wagers, and gross gaming revenue measures stakes minus full payouts in a simple settled-bet model. Realized hold expresses that retained amount as a percentage of handle; net profit also accounts for expenses and taxes.

What do $1,000 in stakes and $930 in payouts mean?

With $1,000 in handle and $930 in full payouts, gross gaming revenue is $70 and realized hold is 7%.

This example assumes settled cash wagers and excludes bonuses, voids, taxes, and operating costs. Full payouts include the returned winning stakes, not just winnings above the original stake.

The calculation is:

• Gross gaming revenue: $1,000 − $930 = $70.
• Realized hold: $70 / $1,000 × 100 = 7%.

If payouts instead totaled $1,100, gross gaming revenue would be −$100 and realized hold would be −10%. The sportsbook would have paid out more than it accepted in this sample, while handle would still be $1,000.

Actual reports may include adjustments. Do not relabel GGR as taxable revenue or net profit unless the source’s methodology supports that interpretation.

Why is overround different from realized hold?

Margin, or overround, describes the structure of the offered odds; realized hold describes the actual accounting result. They are different measures even though both can be expressed as percentages.

Take a two-sided market priced at −110 on each side. Each price corresponds to decimal odds of 1.90909. The sum of their reciprocals is 104.7619%, giving an overround of approximately 4.76%.

Now assume the sportsbook accepts $110 on each side and one side wins, with no push or void. Handle is $220, and the winning wager’s full payout is $210. The sportsbook retains $10, or 4.55% of handle.

The difference between 4.76% and 4.55% is not an error: the calculations use different denominators. Realized hold depends on the distribution of stakes, results, and accounting rules. For more on pricing and odds formats, see how betting odds work.

How do you read handle and ticket splits?

Handle % shows a side’s share of the money, while Tickets % shows its share of the wagers. Comparing them can reveal differences in average stake size, but it cannot predict the winner.

Consider a hypothetical sample from the same two-sided market, source, and observation time. It contains no bonuses, voided wagers, or parlays: there are 100,000 tickets and $10,000,000 in total stakes.

Side A has fewer tickets but more money. Its Handle % exceeds its Tickets % by 27 percentage points: 68% − 41% = +27 pp. This is a difference between two shares, not a 27% increase.

The calculations are:

• Handle % = amount staked on the side / total amount staked × 100.
• Tickets % = tickets on the side / total tickets × 100.
• Average stake = amount staked on the side / tickets on the side.

For A, $6,800,000 divided by 41,000 gives $165.85 after rounding to cents. For B, $3,200,000 divided by 59,000 gives $54.24. Using the unrounded values, A’s average stake is approximately 3.06 times B’s.

What can you calculate from percentages alone?

Percentages alone let you compare average stakes relative to one another, but they do not reveal the dollar averages without the total handle and ticket count.

For this example:

Average stake A / average stake B = (68 / 41) / (32 / 59) = 3.0579268.

The total money and total ticket count cancel out in this ratio. They are still required to answer “What was the average stake in dollars?”

The same 68% Handle and 41% Tickets can occur in a tiny sample or a very large one. Also, 41,000 tickets do not mean 41,000 people: one bettor may place several wagers.

Finally, 68% of the money on A does not mean A has a 68% chance of winning. It describes the distribution of recorded stakes, not an outcome probability.

Do larger stakes prove sharp action?

Larger stakes do not prove that the bettors are professionals or have better information. A gap between Handle % and Tickets % describes the sample’s structure, not the reason behind it.

Possible explanations include large recreational bets, money concentrated among a few bettors, promotions, and the timing of wagers. Public shares alone cannot tell you which explanation applies.

Reverse line movement does not identify its cause either. If odds or a spread move against the more popular side, public betting shares cannot establish whether informed betting or another factor caused the change. See who sets sports betting odds for more on how prices are formed.

Before comparing splits, check:

• The same market: moneyline, spread, and total wagers are separate samples.
• The same source: one sportsbook’s activity does not automatically represent the whole market.
• The same timestamp: later shares should not be compared with earlier prices without acknowledging the mismatch.

The current price may differ from the price at which the larger wagers were placed. Even if an earlier price was favorable, that does not establish that the same opportunity exists now.

No verified strategy backtest has been supplied for these figures. There is no basis here for a reliable rule to “always fade the majority” or “always follow the money.” Betting splits are not a substitute for evaluating a betting edge.

Reporting methods matter too. Bonuses, voids, cashouts, and parlays may receive different treatment across providers. One parlay ticket is not automatically counted as multiple tickets, one for each leg.

Handle vs Tickets Analyzer

The Handle vs Tickets Analyzer compares manually entered stake amounts and ticket shares. It can help check arithmetic, but it does not identify professional bettors or estimate the chance of winning.

The screenshots below were captured locally using the hypothetical example. They do not show live betting data or real-money testing.

Analyzer inputs showing stake amounts for A and B, A’s ticket percentage, and A’s American odds

Hypothetical example inputs: $6,800,000 on A, $3,200,000 on B, 41% of tickets on A, and American odds of −110.

Important before using the widget: its automated sharp labels use unvalidated thresholds. They do not identify professionals or establish an edge. The widget’s claim of historical 54–55% ATS wins has no verified supporting dataset and must not be used for a strategy decision. ATS means against the spread.

Important after using the widget: treat the result as an arithmetic comparison, not a betting signal. Its automated sharp labels are unvalidated, and the claim of historical 54–55% ATS wins lacks verified supporting evidence. A disclaimer inside the tool does not validate its statistics.

How do you reproduce the example?

Enter the four available inputs: the dollar amounts for A and B, A’s ticket percentage, and A’s American odds. B’s ticket percentage is calculated automatically; there is no total-ticket input.

1. Enter 6800000 in the side A dollar-amount field.
2. Enter 3200000 in the side B dollar-amount field.
3. Enter 41% in the side A ticket-percentage field. The tool calculates 59% for B.
4. Enter −110 in the side A American-odds field.

For these inputs, the result shows:

• Handle %: 68% for A and 32% for B.
• Tickets %: 41% for A and 59% for B.
• Average stake on A: $166.
• A’s difference between shares: +27 pp.

Analyzer result showing 68% Handle and 41% Tickets for A, a 166-dollar average stake, and a 27-percentage-point difference

The arithmetic matches the hypothetical example. “Sharp on A” is an automated interface label, not evidence of professional betting or an edge.

What are the analyzer’s limitations?

The displayed $166 average uses whole-dollar rounding and a hidden fixed assumption of 100,000 tickets. If the real sample has a different total ticket count, this output does not give its actual average stake.

In our hypothetical example, the assumption matches the stated sample: 41% of 100,000 is 41,000 tickets on A, and $6,800,000 / 41,000 rounds to $166. Because the tool has no input for the real total ticket count, the displayed average cannot be applied to an arbitrary sample.

The odds field is unused in these calculations. In the local check, changing −110 to −250 produced identical metrics. Its presence does not mean the tool calculates probability, expected value, or the current line.

The analyzer requires manual data and has no live feed. It does not calculate win probability, expected value, or current odds. Reset restores the example values.

If your source supplies percentages only, do not invent dollar amounts to fill the fields. Use the formulas above to compare the shares and relative average stakes, and mark the absolute amounts and averages as unknown.

Where can you verify betting data and its scope?

Check the source description, provider, update time, and market before drawing conclusions from betting figures. A public splits table and an official regulator report answer different questions.

How should you use VSiN betting splits?

VSiN’s betting-splits page describes DraftKings action, not the whole betting market. It states that updates occur every five minutes, but access to the live table may require a subscription or suitable plan.

To check a specific sample:

1. Open the VSiN betting-splits page.
2. Read its data description and confirm the DraftKings provider.
3. Check the update or observation time.
4. If the table is accessible, select the relevant game and a single market.
5. Compare Handle % and Tickets % only within that sample.

VSiN’s explanation published February 5, 2022, discusses reading money and wager counts within a particular sample. Those shares should not automatically be extended to other sportsbooks or all bettors.

What does the official Pennsylvania report show?

The official report shows aggregate amounts for a reporting period, not game-level splits or betting advice. Pennsylvania’s historical September 2025 report lists $850,549,347 in handle and $29,735,588 in taxable revenue.

Fragment of page 4 of Pennsylvania’s official September 2025 report showing handle of 850,549,347 dollars and taxable revenue of 29,735,588 dollars

Original English fragment of the historical report. Handle and taxable revenue are different metrics; their ratio should not automatically be called gross hold.

The full official report, page 4, belongs to the release dated October 17, 2025. The corrected file’s metadata is dated October 22, 2025. This is a historical example, not a claim about the latest 2026 data.

Dividing $29,735,588 by $850,549,347 and multiplying by 100 gives 3.496045%, rounded to six decimal places, or approximately 3.50%. The numerator is specifically taxable revenue. The result is therefore the taxable-revenue-to-handle ratio, not necessarily gross realized hold and not net profit.

When using a report, identify the period and the exact label for each monetary figure before dividing them. An official source establishes where the numbers came from; it does not make different accounting metrics interchangeable.

What can handle tell you—and what can’t it?

Handle and betting splits let you check what a metric means, repeat a calculation, and identify missing data. They cannot predict a winner on their own.

Before drawing a conclusion, ask:

• Does the figure measure money, tickets, revenue, or realized hold?
• What source, market, and period define the sample?
• Are both total handle and total ticket count available?
• How are bonuses, voids, cashouts, and parlays treated?
• Does the price available now correspond to the same time as the published splits?

If an answer is missing, retain that limitation rather than filling the gap with a guess. Heavy betting on a side is an observation about a sample, not a promise of a winning wager.

If you choose to bet, set an affordable loss budget in advance. Do not increase it because of a high Handle %, a large average stake, or an automated sharp label.