How to bid smart in a calcutta
A calcutta rewards judgment more than money. The room sets the price on every team; the pot pays back by finish. Your job as a buyer is to find teams whose expected share of the pot beats what you paid for them.
Most guides on this site are written for the person running the auction. This one is written for the person sitting in the seats with a bidder's number, trying to decide whether the horse for sale is worth what the room is bidding it up to. The mechanics of running a calcutta are covered elsewhere; the point of this piece is to give you a way to think about the auction that isn't "raise until my heart rate is too high."
The math nobody says out loud
Say twelve teams sell in a calcutta, and by the time the last team is hammered the pot is $2,400. The payout is 50/30/20 on the top three, so first pays $1,200, second pays $720, third pays $480 — with everyone else winning nothing.
The expected value of buying a team is the probability that team finishes in each place, multiplied by the payout for that place. If a team has a 40% chance to win, a 20% chance to place second, and a 15% chance to place third, its expected value in this pot is:
(0.40 × $1,200) + (0.20 × $720) + (0.15 × $480) = $480 + $144 + $72 = $696
If the room bids that team up to $500, you paid $500 for a $696 expected return — a good buy by about 40 percent. If the room bids it to $800, you paid $800 for the same $696 expected return — a bad buy, and no amount of hoping fixes it. The number that determines whether a bid is smart is the one you never see printed on the sale card.
Nobody actually does this math with a calculator during the auction. But the point of writing it out is to notice that the room is doing an approximation of it every time it stops bidding. When the room stops on the favorite at $500 and stops on a middling team at $75, the room has priced its own opinion of the winning chances. Your job is to notice where the room has priced badly.
Three patterns that show up in almost every calcutta
These aren't laws of the universe. They're patterns worth watching for, because they generate mispriced teams often enough to be useful.
1. The favorite tax
The team everyone thinks will win almost always sells for more than the math justifies. This is not a strange result — the whole room is bidding for the same feeling of "having the winner," and that feeling costs money. In a 50/30/20 pot, a team that wins one out of two contests it enters is only worth 50 percent of the first-place payout plus a smaller expected share from the runner-up finishes it will occasionally have. When the room bids the favorite up to 40 or 50 percent of the pot, it has often overpaid.
2. The middle discount
The teams the room can't quite decide about — good but not great, dark horse but not sentimental favorite — routinely sell for less than their expected share of the pot. These teams don't inspire bidding wars, and they don't attract the "I'll take a flier" money that goes to true longshots. They just sit in the middle and get hammered down cheaply. If you can identify one team a night whose real chance of a top-three finish is meaningfully higher than what the room is paying, that team is where the value in a calcutta lives.
3. The bottom trap
The temptation with a longshot is to think "for $25 I might win a $1,200 pot." That's true, and it's also true that you have to be right about which longshot. Most of the very cheap teams sell cheaply because they're going to finish out of the money — that is precisely the market working correctly. Buying two or three longshots for the price of a middle team can be reasonable if you actually know something the room doesn't. Buying seven longshots as a shotgun is a way to spend money quickly on the same expected return, worse.
Read the room before you read the field
Two rooms with identical fields will produce different pots and different bidding dynamics, and the difference matters. Some things to notice before the first team goes on the block:
- How many bidders are actually here to buy. A room of forty spectators with six real bidders behaves nothing like a room of forty spectators with thirty. Fewer bidders means slower runs and cheaper hammer prices; more bidders means everything gets bid up.
- Whether local knowledge is concentrated or dispersed. A room full of people who all know the same three top hands will drive those three teams into the sky. A room where knowledge is spread across many teams tends to have flatter pricing.
- The mood of the pot. A charity calcutta where nobody is trying to make money bids differently from a private club pool where everyone is. Neither is wrong — but the same team fetches a different price in each.
- Whether syndicates are forming visibly. If four friends are pooling on the favorite, that team is going to sell for more than one person alone would pay. If you plan to buy the favorite, look for people to go in with; if you plan to bid against the syndicate, don't try to outlast four checkbooks with one.
Set your budget before the first team sells
Decide before the auction starts what you're willing to spend in total, and — separately — the most you'll pay for any one team. Both numbers matter for a reason. A total budget stops you from being the person who spent a car payment on lot after lot because "the next one might be the one." A per-team cap stops you from getting drawn into a war on a favorite when the room's price crosses what the math justifies.
Write the numbers down. Rip up the paper. But have them.
A quick heuristic that isn't a formula
If a team's hammer price is more than 1.5× its share of the total pot at first-place value, you are almost certainly overpaying — that team has to win its class outright to break even, and better teams than it don't win outright half the time. If a team's hammer price is less than 0.5× that same number, and you have any reason to think its chances are ordinary, it's a bargain worth taking.
What to do when the auction gets fast
Late in the sale, the auctioneer is moving quickly, teams are hammering in twenty seconds each, and every price feels reasonable relative to whatever just sold. This is the point where more money goes wrong per minute than any other. Two disciplines help.
First: let one or two teams sell without bidding. Not because those are bad teams, but because sitting out resets your sense of what the room is paying. When you jump back in on the next lot, your read is based on the whole night, not on the last thirty seconds.
Second: when a team you wanted goes past your number, let it go. The person who wins that team paid too much; you didn't. That is the trade being offered, and it's a good one. The temptation is to think "well, I was going to pay $200, what's another fifty" — and to keep thinking that four more times.
Two things to do with a syndicate
Buying with three or four other people lets you win teams you couldn't afford alone, and it changes the math in your favor by spreading your money across more of the field. Two habits to build:
Agree on the per-team cap before the auction. A four-person syndicate that hasn't decided in advance whether their cap on the favorite is $400 or $800 will find out at the auction, at speed, in front of a live auctioneer, and will not enjoy the answer.
Write down what each person pledged, at the moment of purchase. Not later. Later is when the arithmetic goes wrong and people remember pledging different amounts than what they actually pledged. Every argument about who owes what after a calcutta traces back to a moment when the clerk wasn't watching.
The one true rule
Everything above is a heuristic. The one thing that isn't a heuristic is this: you cannot outbid the pot. If a team is bid to more than the first-place payout will return, you cannot make money on it even if it wins. The most any team can pay is 100 percent of first, and in most structures first is only half of the pot. If the price of any single team is climbing past a plausible payout, someone in the room is deciding to lose money to own the winner. That's a legitimate choice. It's just not a smart bid.