How the Calculator Works
Every quantity on this site comes from one model with no hidden steps. This page sets out each constant, in the order the calculation applies them, and then compares the result against two outside reference points. The constants below are read directly from the same files the page generator uses, so this description cannot fall out of step with the math.
1. How many guests are drinking
Not everyone at a party drinks alcohol, and the share falls as the guest list grows. The model uses a drinking rate by guest-count tier.
| Tier | Guests | Drinking |
|---|---|---|
| small | 10 to 30 guests | 80 percent |
| medium | 40 to 75 guests | 75 percent |
| large | 100 to 150 guests | 70 percent |
| xlarge | 200 or more guests | 65 percent |
For 100 guests that is 70 drinkers. Food uses a separate and higher rate, since nearly everyone eats: 95 percent for small, 90 percent for medium, 85 percent for large, 80 percent for xlarge.
2. Drinks per hour, and how that fades
The base rate is 1.5 drinks per drinker per hour. People drink fastest early and slow down, so each hour carries a decay factor rather than counting equally.
| Hour | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|
| Factor | 1.00 | 0.90 | 0.75 | 0.60 | 0.50 | 0.45 |
The last figure repeats for any hour beyond the list. A 5-hour reception therefore counts as 3.75 drink-hours, not 5. For 100 wedding guests: 70 drinkers times 1.5 times 3.75.
3. The event modifier
Each event moves each drink up or down from that baseline. A wedding lifts wine to 1.2 and trims beer to 0.9; a Super Bowl party does close to the reverse. These modifiers are per event and per drink, and they are listed in full in the site data.
4. Share of the bar
The headline number on each drink page is that drink's share of a full bar, not the amount you would need if it were the only thing served. Total drinks are fixed by guest count and duration; putting out four kinds of alcohol divides that total rather than multiplying it. Each event has a weighting across the four drinks, normalized so the shares sum to one. At a wedding, wine is 37.5 percent of the bar, which turns 473 servings into 177.
Every page also publishes the only-drink figure, so if you really are serving one thing, the larger number is right there.
5. Pour sizes
Servings per container, which convert drinks into things you can buy:
| Drink | Container | Servings |
|---|---|---|
| Wine | bottle | 5 |
| Beer | case | 24 |
| Champagne | bottle | 6 |
| Spirits | bottle | 17 |
6. The buffer
Every quantity is multiplied by 1.15 and rounded up, a 15 percent margin for heavy pours, spills, and guests who stay late. Running short is worse than having a few bottles left over, and unopened bottles can often go back.
7. The toast
Champagne pages carry a separate toast figure. A toast is the one moment where headcount matters more than drinking habits, because people who are not drinking still take a glass. So the toast counts 1 flute for every guest, not just the drinkers, at 6 flutes per bottle plus the same buffer. For 100 guests that is 20 bottles.
How this compares
Against one drink per guest per hour
The common rule of thumb is one drink per guest per hour. For 100 guests over 5 hours that is 500 drinks. This model lands at 447, about 11 percent lower, because it counts only the guests who drink and applies the decay curve instead of treating every hour the same.
Against The Knot's wedding list
The Knot publishes a sample bar for a 100-guest, five-hour reception of average drinkers, assuming one drink per guest per hour and a 50 percent liquor, 25 percent beer, 25 percent wine mix. Their figures beside ours:
| Category | The Knot | StockTheEvent |
|---|---|---|
| Wine | 20 bottles (750 ml) | 41 bottles (750 ml) |
| Champagne or sparkling | 17 bottles (750 ml) | 19 bottles (750 ml) |
| Liquor | 11 bottles (1 liter) | 6 bottles (750 ml) |
| Beer | 100 cans or bottles | 5 cases, 120 cans |
Source: The Knot, How to Stock the Bar at Your Wedding, updated September 30, 2025.
The champagne, liquor, and beer figures land close together. Wine is the one real gap, and the reason is the split rather than the pour rate: our wedding profile gives wine 37.5 percent of the bar against their 25 percent, and we count 5 glasses to a bottle before adding the buffer. If your wedding pours less wine than that, the share and only-drink figures on each page are there to adjust from.
What this model does not do
It does not know your guests. It has no data on your particular crowd, the weather, whether the bar is open or cash, or how heavy your pours are. It is arithmetic applied to reasonable assumptions, published in full so you can see where it would be wrong for you and shift it.