Satellites
How Many Chips Lock Up a Satellite Seat?
The stack that locks up a satellite seat depends on what share of the players left will miss out. With 11 players left for 10 seats, about 1.3 times the average stack is 95% to get one, and twice the average is 98.9%. With 10 left for 5 seats, 95% takes 3.8 times the average. With 9 left for 3, it takes 5.4 times, 60% of all the chips.
Each number is your chance of finishing with a seat, worked out by ICM with every seat worth the same. The average stack is all the chips divided by the players left. Your stack is shown as a multiple of it, and every other player has the same stack. The satellite bubble calculator does the same sum for any real table.
Chance of a seat by stack size
| Your stack | Chance of a seat | |||
|---|---|---|---|---|
| 11 left, 10 seats | 12 left, 10 seats | 10 left, 5 seats | 9 left, 3 seats | |
| 0.25× average | 47.5% | 36.8% | 15.6% | 9.3% |
| 0.5× average | 71.4% | 59.7% | 28.9% | 17.9% |
| Average | 90.9% | 83.3% | 50.0% | 33.3% |
| 1.5× average | 96.9% | 93.0% | 65.3% | 46.5% |
| 2× average | 98.9% | 97.0% | 76.3% | 57.5% |
| 3× average | 99.9% | 99.5% | 89.5% | 74.5% |
When everyone has the same stack, each player's chance is the share of players who get a seat: 10 of 11 is 90.9%, 3 of 9 is 33.3%. Each setup opens in the calculator at the average: 11 left, 10 seats, 12 left, 10 seats, 10 left, 5 seats or 9 left, 3 seats.
How many chips is 95% to get a seat?
| Setup | 90% stack | 95% stack | 99% stack | 95% stack, share of chips |
|---|---|---|---|---|
| 11 left, 10 seats | 0.96× | 1.27× | 2.03× | 11.6% |
| 12 left, 10 seats | 1.29× | 1.69× | 2.63× | 14.1% |
| 10 left, 5 seats | 3.05× | 3.82× | 5.33× | 38.2% |
| 9 left, 3 seats | 4.50× | 5.38× | 6.84× | 59.8% |
The number follows the share of players who miss out. At 11 for 10, one in eleven goes home empty and a little above average is plenty. At 12 for 10 it's one in six, and 95% takes 1.7 times the average, the same as a small satellite with 6 left for 5. Three players for two seats takes 2.1 times. When half the table misses, as at 10 for 5, you need 3.8.
Does the total chips divided by seats rule work?
A rule from Barry Carter and Dara O'Kearney's book Poker Satellite Strategy: divide the total chips by the number of seats to get a target stack, and slow down once you reach about 70% of it. On the bubble the target is just over the average. Eleven players for ten seats puts it at 1.1 times the average, a 92.7% chance of a seat, and the 70% point, 0.77 times, is already 84.7%. On a 10-for-5 table the same rule asks for twice the average, which is only 76.3%, and its 70% point is 62.6%. Treat it as a floor, not a lock.
The other figure you see on strategy sites is 1.5 to 2 times the average. On the 11-for-10 bubble that's 96.9% to 98.9%, more than you need. At 10 for 5 it's 65.3% to 76.3%, and not safe.
Is doubling up worth it on a satellite bubble?
On the bubble, it hardly matters. An average stack at 11 for 10 already has 90.9%, and calling another average stack's all-in and winning bursts the bubble: a seat, up 9.1 points. Losing costs the whole 90.9%. That call needs 90.9% equity, more than pocket aces have against a random hand (85.3%). The satellite bubble formula gives that number for any table size.
Three seats for nine players is where chips start to matter. Bust an equal stack there and you go from 33.3% to 58.3%, a gain of 25.0 points, and an average stack has to win chips to have a real chance.
When is a satellite stack safe to fold?
Once you're past the 95% stack, fold to all-ins from anyone who covers you or can cut you down to size, and keep opening. The players behind you can't afford to call you either, so a safe stack picks up the blinds every orbit. A shove from a very short stack is different: holding 1.3 times the average at 11 for 10, calling a stack a quarter of the average needs 59.8%, and half the average needs 71.0%, because winning ends the bubble and losing barely dents you. Calling a stack twice the average needs 96.4%. The satellite bubble calculator opens on a spot where folding aces is right, and the lessons The satellite bubble bully and Folding aces in a satellite play out each side of it.
What the table doesn't show
ICM knows chips and seats, nothing else. Carter and O'Kearney say the average stack usually has about ten big blinds on a satellite bubble, which puts the 95% stack at 11 for 10 around 13 big blinds. If the blinds are high enough, a stack that size can still be blinded down while the short stacks fold, and the table above can't see that. Count your stack in big blinds as well: the M to big blinds tables say how many orbits you can sit through. ICM also treats everyone as equally good, and the tables here assume the other stacks are all the same size, which at a real table they won't be.
How do satellites award seats?
One EPT satellite format made the lock-up a rule: you won your seat the moment you reached 100,000 chips, from a 10,000 start, and your chips left play. Most satellites leave it to the bubble. A single-table satellite has ten players and one seat, sometimes two, to an event costing about ten times the buy-in, and the WSOP's $1,060 mega satellite in 2007 gave one Main Event seat for every ten entries. Step satellites chain five or six single tables, with the winner moving up.
When two players bust in the same hand on the bubble, GGPoker ranks whoever started the hand with more chips ahead. Some online satellites pay a little cash to the next finisher: in one PokerStars six-handed satellite for two $109 seats, third place got $2.56.