Spawn of Grimstarr: PokerStars Sit-out Trick Proliferating
Among the news topics of lesser importance but greater interest to emerge in recent days is another proof of the old truism that no matter how fair one tries to make a rules system, someone out there will still try to shoot an angle. What’s unusual in the story that follows is how prolific the use of the ploy has become – in this case, the use of a flaw implemented within PokerStars tournament seating rules that allows players to sit out to increase their odds of being moved to a softer overall table. It’s called the PokerStars sit-out trick, and its use is increasing.
Not aware of this one? I wasn’t either, until recently, though that’s partly due to the fact that as a US player who hasn’t attempted to try to play via an offshore account and a VPN, I haven’t participated on PokerStars since 2011. Still, this should be a topic of interest to all MTT (multi-table tournament players) around the globe, since if the world’s biggest online-poker site can get it wrong — and, seemingly, refuse to address the situation for a couple of years – then the optimal rules of poker aren’t nearly as uniform as many people think.
The issue received a publicity boost in recent days thanks to high-stakes online pro Steve O’Dwyer, a prominent force in the game for most of online poker’s existence. O’Dwyer, best known as online pro “MrTimCaum,” recently took to the 2+2 discussion forums to express his growing displeasure with PokerStars over the proliferation of the ‘sit-out trick’ on Stars.
All online sites in the modern era use a seating algorithm to move players from one table to another to rebalance tables or to account for new players during the late-registration period. As O’Dwyer relates it, the time-honored practice at PokerStars differs from accepted tournament-poker practices… and it’s been that way for a long time.
Here’s how it normally works, and how it differs at PokerStars. According to most MTT tournament-operations directives, players from tables that are fuller are moved to emptier tables as part of the balancing process. The widely accepted “best practice” method is to move “next big blind to next big blind, meaning that the player selected from one of the fuller tables is taken from the seat which would post the big blind on that table’s next hand, and the player is then moved into the vacant seat at the new table which is as close to the next big blind as possible.
Some minor variants exist with that, but the idea behind the practice has always been to prevent players from gaining large positional advantages within the context of a single table orbit, or to luckily evade paying the blinds altogether for one or more orbits while moving between tables.
The high speed of online play and the massive number of tables in large MTTs make an algorithm for handling all this a complex procedure. Still, that’s how it done, and even a poor system is a far sight better than how it was handled in online poker’s earliest years.
As part of my digging into the background of the old UB scandal, by way of digression, I learned first-hand that in the old days, 2003 or so, player moves were handled manually by a site overseer whose job it was, literally, to monitor MTTs and move players table to table for balancing.
At the very least, online poker’s come a long way from that.
At PokerStars, however, the optimization chosen by the site’s engineers took a different track – the algorithm was written to favor pulling players from fuller tables who were sitting out in addition to their proximity to the next big blind.
What this means in practice is that MTT regulars on PokerStars have widely recognized that sitting out increases their chances of being bumped to a different table. Since many high-stakes tourneys often feature tables full of regulars, players are reported to be intentionally sitting out during the earliest levels of major MTTs in the hopes of being shuffled to a softer table.
Whether this represents an angle shot or not is up to debate, because there’s no intentional deception of other players involved. Instead, it’s more of a situation where a knowledgeable few have taken advantage of an different many, not that that doesn’t mean Stars shouldn’t address the issue. In other words, the process may well be a historical anachronism, but given the modern-day abuse, it may be time to chuck it.
Though Stars reps have not offered public comment on the situation on poker discussion forums, there appear to be two possible reasons for why their method was adopted: (1) It may have led to less disruption among players who were actively participating in every hand, and (2) it may have prevented unintentional sit-outs (affected by such things as internet disconnects), from having all their chips pillaged by a single table’s players in high-speed formats such as turbo and hyper MTTs.
Stars’ current table-seating algorithm, however, has now been exposed as sub-optimum, and O’Dwyer and others are now reporting that large numbers of online players have begun exploiting it; one poster reported (though this has not been confirmed and seems somewhat of an exaggeration) that as many half of the early entrants in the recent WCOOP main event may have been doing this, when the relative costs of missing hands and paying modest blinds were minimal.
PokerStars is aware of the situation, and has been so for quite some time prior to the sale of the company from the Scheinberg family to its current corporate owners, Canada-based Amaya Gaming. Stars’ Tournamant Team Manager, Mike Jones, even stopped by the 2+2 discussion thread to add a tiny bit of corporate input to the discussion. Wrote Jones:
The issue being discussed here is table balancing in tournaments.
Table balancing is the process of moving players from tables having more players to tables with fewer players so that all tables in a tournament are populated as equally as possible. This balancing consists of a sequence of operations, each of which moves a single player. At present the first factor is whether a player is sitting out. It was set up this way in order to minimize the impact to the playing experience. If a player is sitting out, they presumably are not playing, and therefore moving tables has no impact to their playing experience.
As implied in the original post, there turns out to be an exploit here that we didn’t foresee when the table balancing algorithm was designed. Some players have come to realize that sitting out increases their chances of being moved to a different table, and use this exploit when they want to be moved.
It has only recently come to our attention in PokerStars Poker Room Management that some people are exploiting the sitting out portion of the table balancing algorithm. Since it came to our attention a few weeks ago, we have been working on a change. I think it is worth noting here that our desire to change this has nothing to do with the fact that it was “made public” in the OP. Rather, our desire to make a change was and is driven by the desire to offer the best poker experience possible.
The idea that we’re currently pursuing is to simply change the sitting out portion of the algorithm to be x number of hands. With such a change, sitting out will only be considered for a table change if that sitting out has happened for x consecutive hands. X can be any number we wish. The higher the number, then the less exploitable this becomes, but we would want to choose a number low enough that sitting out status could still be considered, due to the aforementioned desire to minimize impact to the playing experience. If you have feedback regarding that number, please share it.
Like any software change, this one will require time to write, test, and implement, but the ball is rolling.
Tournaments Team Manager
PokerStars Poker Room Management
Whether or not the “X-hands solution” described in the above will fix the inequities the current algorithm allows for would remain to be seen, though for some players, “the ball is rolling” isn’t fast enough. O’Dwyer, despite a string of major successes in recent tourneys on PokerStars and elsewhere, continues to agitate for a speedy fix.
O’Dwyer even accused Stars’ software engineers and customer support of pooh-poohing the situation, posting the following not long after Jones’ post:
As I mentioned in the OP, a friend of mine, an extremely prominent and high volume player inquired about this issue with Stars MONTHS ago and was brushed off. By the end of WCOOP it was very clear to me and my friends that this exploit did in fact exist, and following my trip to UKIPT IoM, where I had multiple discussions with various Stars staff about the exploit, it was clear to me that fixing this was not a major priority. With that said, I find it disingenuous at best for you to say that you were only aware of it “a few weeks ago” and that my OP was irrelevant in regards to your “desire to make a change.”
O’Dwyer subsequently reported receiving an e-mail about the issue from Jones, but also relayed that Stars was treating it as a low-priority gaming issue, one that perhaps wouldn’t be “fixed” until some time in 2015. Other prominent players confirmed that the practice is ongoing and increasingly prevalent in higher-stakes MTTS. About a week ago, O’Dwyer took his complaints to Twitter, drawing even more attention to the algorithm’s unintended consequences.
Meanwhile, a practice that looks a bit “grimmish” in its implementing looks as though it will continue unimpeded. For those new to the term, the “grimming” reference in the title refers to the antics of a well-known online player named “grimstarr” in the mid-‘00s, who used sitting out as unethical tactic in various ways.
“Grimming” in online poker describes two unethical tactics, both originally practiced by the same player: either sitting and playing one hand at a new heads-up cash table and if having received favorable blinds/button position, then immediately leaving the table after that single hand; or pre-arranging a “flip” tourney with an opponent with the advance agreed terms of going all-in blind on the first hand (hence the “flip”), then failing to honor the all-in-push deal if receiving a subpar hand.
The so-called PokerStars “sit-out trick” at PokerStars isn’t quite a grimming play, at least to the extent it heads down that classic, unethical road, but the ploy at least takes a nod and a peek in that direction. Given the increasing awareness and use of the tactic, it’s in PokerStars’ interests to fix this sooner, rather than later.