Many online poker sites now have a profound understanding of how online poker players‘ play relationships affect their bottom lines. One player’s actions affect not only other players’ spending power but also the margin, or rake revenue per deposited dollar, on those players’ spend. So who plays who where for what matters. The old doctrine that sites don’t care who wins or loses is long since dead and buried. As a result, online card rooms are busy trying to come up with ways to stem the (too) rapid conversion of deposits to withdrawals and create optimum conditions for maximum value play while still offering a fair and credible game. Forced player segregation where players are prohibited through algorithms in the software from playing against certain other players is the latest attempt to do so.

While actively working towards improving the profitability of play is a welcoming development, it is important to remember that forced player segregation is only one of several solutions available to improve raking conditions. I’d personally explore other avenues. Forced segregation has a lot of potential but releasing that potential is complicated. Implementing segregation secretly behind the scenes breaks a sacred boundary of intervention that sites should not cross (it’s the same boundary I think they fail to uphold by covertly permitting the use of third party player aid tools). Implement it transparently and the resulting gameplay must be compelling and marketable. What good does it do to have better margins on play if play no longer makes sense? Someone has to sell the idea to the players. And the story they tell must be both credible and intriguing.

It’s a perfect example of when business design and game design must marry.

To make forced player segregation work, there are three sometimes conflicting goals that must be met

1. It must stall the conversion of deposits to withdrawals
2. It must create a backdrop upon which fun and rewarding gaming experiences  can be had
3. The solution must be packaged so that it is intuitive to understand and thus easily communicated

Failure to address either one of these is bound to lead to implementations that prove the critics right.

It’s not an easy task. Each chosen factor upon which to base the algorithm that groups players has its pros. None, based on my own experience of trying to patch together a working solution, is without its distinct drawback.
Even the most intuitive solution is far from straight-forward.

One of the causes behind the rapid deposit-to-withdrawal conversion is of course the skill gap. So closing the skill gap will create desired effect in terms of reducing the pace at which money flows from depositors to withdrawers.
Enter skill-matching.

Skill-matching, on paper at least, also meets the two other demands really well. It is perfectly natural to design a game that matches up players of equal skill level. Lots of games use it. It’s easy to grasp and it can generate a engaging narrative from which great player stories can be weaved. From a competition standpoint it makes perfect sense. It’s not unfair as long as it is transparent. It’s a great way to measure players against each other to see who really is the best. By incorporating skill matching every game served will be relevant. The stage for tough and intensive games where only those who play up to their very best will succeed is always set.

It’s still no walk in the park to convince players to play poker competitively in this manner, but skill-matching is at least a coherent concept. Again, on paper.

Unfortunately skill in poker is very difficult to measure. Especially if it has to be measured in a manner that actually improves rake conditions.

Before it is possible to make accurate predictions regarding a player’s skill level, most of the damage may already been done – all new good could players create carnage before being adequately labelled. Five hundred played hands to make a reasonably accurate call is a number I hear quite often. If you know the data on player losses that sites are so concerned about you know that five hundred hands is a long time. Base your segregation on mere indications of skill, general assumptions or primitive rules like that bankroll surplus players may not play with net neutral or negative players and the skill-matching story would not fly. A skill-matching site can’t have Jerry Yang champions.

How would the solution change if one instead of skill based it entirely on site activity or loyalty? What impact on the profitability of each deposited dollar would saying that “you get access to more opponents the more you play with us” have?  Would that argument fly?
What would be the playing community’s reaction if the basis for segregation was in fact simply that players with short-term gains above x cannot play players with short-term losers above y but it wasn’t framed as a skill-matching setup?  How could it be framed instead? And how would gameplay change as a result?

To stand any chance of successfully introducing forced player segregation I think one has to start with gameplay and the marketable story. An algorithm that does as fine job as possible of backing that story follows next. The fact that it may be far from optimum from a monetary flow perspective is something that has to give. The revenue gains should primarily come from the increased attractiveness of the product as a result of the implemented changes – not the increased efficiency of the rake engine. Forced player segregation can enable a site to shape a differentiating marketing message. In a streamlined market like online poker, that’s not a laughing matter.

Only once forced player segregation is discussed from multiple angles is it possible to hail or fail it. It remains to be seen whether someone can figure out the right brew. I personally don’t like the taste of it much, but the idea deserves more scrutiny than a traditionalist knee-jerk reaction that it is discriminatory and not “poker”.