formats · spin and go
Spin and go.
Three players, a clock that rises faster than you can wait it out, and a prize pool that is drawn at random just before the first card. Those three facts pull the strategy away from everything you learned at a nine handed table, and they do it in different directions, so it is worth taking them one at a time.
one · three players
Two opponents cost far more than twice one.
Every hand you can be dealt has an equity against one unknown hand and a different, much lower one against two. The drop is not gentle. King two offsuit is a coin flip against a single random hand and clearly behind against two, and seven deuce offsuit, the hand with the worst reputation in the deck, goes from winning a third of the time to winning one time in five. This is the number people skip when they carry a heads up shoving habit into a three handed table.
| Hand | All in vs 1 | All in vs 2 |
|---|---|---|
| 69% | 50% | |
| 61% | 42% | |
| 58% | 38% | |
| 52% | 33% | |
| 51% | 31% | |
| 35% | 20% |
Equity when all the money goes in before the flop, against one and against two unknown hands. Simulated over 400,000 deals per figure with our own hand evaluator and rounded to the nearest whole percent.
The practical consequence is about who still has to act. Shoving from the button three handed puts the decision to two players, and the chance that both of them fold is far smaller than the chance that one does. That is your fold equity, and it shrinks with every player left behind you. It also means the reverse: once someone busts and you are down to a heads up match, your range should widen sharply, because every hand in the table above is suddenly worth fifteen to twenty points more than it was a hand earlier.
two · the clock
Waiting costs half a big blind per hand.
An orbit costs one and a half big blinds wherever you play: the small blind plus the big blind, once per lap. Three handed a lap is three hands, so folding everything costs you half a big blind per hand. Say you sit down with twenty five big blinds. Fold every hand and you are down to nothing after about sixteen orbits, which is fewer than fifty hands, and long before that you have stopped having a stack anybody has to respect.
That is why the interesting part of a spin and go happens between roughly ten and fifteen big blinds. At that depth a normal raise commits you to the pot anyway while giving your opponent a cheap price to call, so the choice collapses to all in or fold. Which hands go in depends on your seat and your exact stack, and the push fold calculator has the full ranges for every depth. What decides the size of the pot is not your own stack but the smaller of the two in the hand: sitting on twenty against someone holding eight, you are playing an eight big blind hand. At these depths what is left behind after the flop is small enough against what is already in the middle that the flop is usually where the hand ends.
Position matters more here than at a full table, not less. Three handed the button acts last after the flop and only two players can wake up behind a raise, so the seat you are in changes which hands are playable more than one card of difference does. If your opening ranges are still fuzzy, the ordinary starting hand order is worth fixing first on the preflop chart.
three · the prize is drawn first
Winner takes all is the one case where ICM has nothing to say.
Everywhere else in tournament poker, chips and money come apart. Doubling your stack does not double what you are owed, because you can only finish first once, and ICM is the model that puts a price on that gap. In a spin and go the size of the gap depends entirely on what the draw produced.
If the whole prize goes to the winner, there is no gap at all. Take stacks of 1,000, 300 and 200 and put the entire pool on first place: run that through the model and the three players own 66.7, 20.0 and 13.3 percent of the prize, which is exactly their share of the chips. A chip won and a chip lost are worth the same, so you can think in chips all the way down and never be wrong.
Change nothing except the payout, so that second and third also get something, and it separates immediately. With the same three stacks and a pool split 80, 15 and 5 percent, the leader with two thirds of the chips owns 57.7 percent of the money, while the short stack with 13.3 percent of the chips owns 18.0. Chips at the top of the table are worth less than they look and chips at the bottom are worth more, which is the whole of ICM in one line. You can run your own stacks through the ICM calculator and see the same effect.
There is one more consequence of the random draw, and it is about how you read your own results. When the size of the prize is decided by a lottery before the cards come out, your balance after a few hundred games says very little about how well you played them. It is the clearest case in poker of variance drowning the signal, and a good reason to judge yourself on decisions rather than on the number at the bottom of the screen.
next
Get the shoving ranges right, and most of this format is solved.
Almost every spin and go is decided somewhere between eight and fifteen big blinds, and at that depth the correct play is a lookup rather than a judgement call. Enter your stack and your seat and the calculator gives you the hands that go in and the hands that do not. Free, no account.
questions about spin and go
What is a spin and go?
A three handed sit and go with a very fast blind structure and a prize pool that is drawn at random just before the first hand. You pay one buy in, three players start, and the game ends when one of them holds all the chips.
How do you win at spin and go?
By being comfortable with short stacks and by folding far less than you would at a full table. Three players means you are in a blind on two hands out of every three, so waiting costs you half a big blind per hand. Most of the money is decided between roughly ten and fifteen big blinds, where the real choice is all in or fold rather than raise and see a flop.
Is spin and go profitable?
It can be, and it is one of the harder formats to be sure about, because the prize multiplier is drawn at random. A long run of ordinary prize pools can hide a perfectly good win rate, and a single large one can hide a bad one. Judge the format on your decisions over a large number of games rather than on the balance after twenty.
How many big blinds do you start with in a spin and go?
Shallow, and it varies by operator and stake, so check the lobby rather than trusting a number you read somewhere. What matters more is that the levels are short: whatever you start with, you are usually inside push or fold territory within a handful of orbits.
Does ICM matter in a spin and go?
Only when more than one place is paid. If the entire prize goes to the winner, there is nothing for ICM to correct and a chip is worth exactly what it is worth: winning them and losing them balance out. The moment the structure pays second and third as well, that stops being true and the chip leader owns a smaller share of the money than of the chips.
Should you shove any two cards three handed?
No, and the equity table above is why. Against one random hand the worst holding in the deck still wins about a third of the time, but against two it wins about one in five. A shove that only one player has to fold is a very different proposition from one that two players have to fold.
What is the difference between a spin and go and a normal sit and go?
Three things: three players instead of six or nine, a much faster clock, and a prize pool that is unknown until the game starts. A normal sit and go has a fixed payout table you can plan around, which is exactly what makes ICM the central skill there and short stack play the central skill here.
Back to all five poker formats, or read how the same short stack thinking plays out over a whole evening in a multi table tournament.