There is a simple fix for all blind-play problems, and that fix is this: whenever the game has n number of players, nobody can join or leave the game until xn hands have gone by. That's it.

For example, if there are four players in the game, the software makes everyone play an equal number of BB, SB, BN and CO before anyone can join or leave and make it 3-way or 5-way. If you started in the BN, you finish after your SB. If it is down to two players the software makes both players play their button the same amount of times before being allowed to quit.

It's so incredibly simple. It can be baked into the software readily. It guarantees no one will get buttoned or extra-blinded. You can take your BB without fear with this system, which therefore means games are much more likely to keep going regardless of whether certain players quit.

I think these are reasonable concerns, but are almost completely unwarranted.

The best reg probably isn't the best reg playing 4 people at once who are each only playing him. So if you can't play 4 players as the best reg, then when you are playing 4 players you aren't the best reg anymore.

If someone knows you are going to give up the table and has no intention of playing you, then call their bluff and 3bet their first open and make them look silly.

However, if we go with KoTH, I strongly advocate a system of one player per open table. I think one player/table is superior both in terms of fairness and in terms of giving fish a reasonable selection of opponents to play.


I don't think this example is particularly plausible or remotely likely to become a large scale problem. However, if it did, it can be handled by contacting Stars' support. At the very worst though, the sitter gets to play at least one hand with the bumhunters where he is +EV.


Again, I don't think this is plausible.

At the very worst though, reg4 can sit in with reg1 and play 1 or more hands where he is +EV, at which point reg10 swoops in again. If this cycle continues, reg4 can leave his table when reg10 sits, and then grim reg1 when reg10 cedes him his table, making reg10 and reg1's collusion unprofitable.

Sure, this is annoying. At the very worst though, the slow-reg cedes some small amount of EV per hand to the sitting-reg. In addition, the sitting-reg should complain to support.

Simple. Give the KoTH the first button.

This is pure speculation on your part, and what's worse you present it as being obviously true.

I find the view counter-intuitive. Your position seems like this: 'Fish hate losing EV fast much more than losing EV slowly.' I don't see why. Fish are concerned with entertainment, not expected value. Also, fish don't know what their expected value in a given game is for the most part, that's one of the reasons why they are fish. So, they don't even have a precise idea of the thing you think they care so much about.


You're right. Eliminating bumhunting will make it more difficult (if not impossible!) to move up the ranks by bumhunting.



I'm not trying to be overly critical here, just to engage in fun and healthy debate and to try to arrive at a reasonable solution we can agree upon.

I think some form of KoTH has a lot of merit, and I'd like to see the idea continue to be on the table.

While this is complete trolling, I think it's worth pointing out why it's wrong.

Poker (without rake) is a zero sum game where betting decisions effect the EV of the players. Every (or at least most) players play differently. So, there will be winners and losers in poker.

If any member of the player pool is eliminated, this effects the winrate (or lossrate) of the rest of the player pool. So, if the winners are eliminated, the small losers become winners and the small winners become big winners. Similarly, if the losers are eliminated the big winners become small winners and the small winners become losers. Point being, there is no way to eliminated winners.

Another (maybe simpler) way of putting this is to say that eliminating winners won't change the fact that poker is a zero sum game.

By "one player per open table", do you mean "one open table (him waiting for a match) per player"? Also, how would you approach the number of non-full tables per player (if you have a specific suggestion)?

Sorry, that was unclear. I meant to say 'one non full hu table per player.' And then infinite full hu tables per player with a checkbox installed in the client to add another table once one is running.

Well, why would support do anything about it? Should playing slow be made against the TOC? Seems a bit far fetched.

Another issue I thought about is that people worse bumhunters might learn to play a decent shortstack game. They buy in shallow and you are forced to play them or give up the table. I'm not saying this is necessarily a problem, but just pointing out that there are a lot of ways in which KotH could turn out different from the heaven for stronger regs people are making it out to be now.

And one more: say you are a 25/50 reg and you constantly battle for tables there. Because of koth, you can't just sit at a 10/20 table, you have to battle for it. But you'd rather battle 25/50. Is it 'fair' that you can not have 10/20 table to wait for fish at? I don't battle regs at 3/6 but I do wait for fish there, should that behaviour be discouraged when I am better than virtually everyone willing to play 3/6?


So if a weaker player happens to have a table and I want it he can grim me before he gives up the table?



While you are right that it is speculation, do you really think this is totally true? Do you think fish are totally indifferent to their EV? That they don't care at all about their losses? That they are just as likely to keep depositing when they lose it at -15ptbb (with a much worse $-per-hour-of-entertainment-had ratio, and with much less winning evenings to keep them hooked on the game) than when they lose it at -5ptbb?

I never meant they are concerned with their EV in that way (although I think anno 2012 a big amount of fish has read/watched strategy in one way or another, just not to the extent we do). I meant that they like to win every once in a while, and that the winning nights are part of what keeps them coming back. If I was a fish I'd have much more fun vs a mediocre bumhunter that I have more winning nights against, than vs a very good reg who just reads my soul every single hand. I think most fish don't find it fun to get raped on a consistent basis, I think they like to have a fighting chance, maybe play an opponent that is very tilted too sometimes, that they can sometimes outthink/that one trick that worked they remember the whole evening. Playing better regs on average seems like it would mean less fun for a bigger $ cost for fish. They don't even have to consciously think about to stop depositing.

The fish who lose at 30ptbb and keep depositing are very small in numbers (and it's not a given what I described above doesn't matter for them). Most fish have a winrate like -8ptbb, and seem to have read some sort of strategy (limping is less popular than 3 years ago and 3betting more popular, for example). I think they are more alike to the NVG reader than to the oblivious-to-everything fish you describe, and I think they would notice if they were only playing very good regs now and the effect that has on their enjoyment and the price they pay for their enjoyment.




It will also make it more difficult to move up the ranks by a combination of reghunting and bumhunting, and it will also make it more difficult to move up the ranks by purely reghunting as the skill level of average regs sitting will go up - even hardcore bumhunters play sometimes if you sit them (not often, but it does happen).

I find it weird people seem to take it for granted the current system is broken (has anyone pointed out why btw? other than that fish don't like the lobby which is 'pure speculation' too and the weird idea that the best regs deserve to win an even bigger part of the cake that they do now), but if we must change it one-table-a-player-reasonably-amount-of-tables koth is my favorite too, for sure much better than anything else. But then again, how much will having 30 mostly empty tables in the lobby versus 130 really matter in the eyes of a fish?

If this is an issue, which it probably wouldn't be for PS, then it's an issue with the cap on buy-in amounts.

I think this question stems from all the confusion the bumhunter apologists have been trying to stir up. The object of the changes suggested here is not to create a money making safe haven for all the tough regs. The only objective is to create a system where playing others is incentivised. You can have all the tables you are willing to battle for. The scenario you bring up is actually a great example of how more action might be generated: A 25/50 regular is battling for his table at 25/50 and 10/20 vs. two different people and a 3/6 regular sees this. He feels that if the game was straight up, he would probably be at a disadvantage towards this player, but with the 25/50 player being kept busy by two others, he fancies his chances and comes to challenge him for the 3/6 table. Now the 25/50 player has to decide how confident he is in his multitabling abilities, or if he would rather just give up the table. This seems extremely fair to me. The 25/50 guy already has action on two tables, which is great for him. Why would he be entitled to any more tables that he is unable to defend?

I don't like sauces suggestion either, but the issue with Party you were complaining about can be completely eliminated by making people post their blind before high-carding for the button.

Poker was a zero sum game in 2005. Poker is a zero sum game in 2012. Are the games better or worse now? If they are worse what does worse mean?

which blind do they post if they have not yet high-carded?

By "post", I just meant a commitment to pay the blind no matter which blind it happens to be. This is the way it works in Entraction, where there's not even a prompt for "Do you want to post your blind?", as the game starts immediately when there's two players sitting in. However you can still have the prompt, but just make sure that the button is highcarded afterwards.

Rather than building up straw men or continuing to argue against the motives of the proposers rather than focusing on the issue at hand, I challenge anyone to come up with a good reason why the most agreeable proposed fix so far, the One Empty Table Per Player KoTH, would NOT result in more action.

The point I've seen some people make is that "People don't want to play in -EV games, and you can't force them to", but this is a very blanket statement to make. The problem is not that people are unwilling to knowingly play in -EV games, but rather that the vast majority of players have become so risk averse that they are not willing to engage in situations where their edge is unknown, because they can always wait for the sure thing instead. This is a problem not with the players but with the system, because there is no incentive that rewards people with higher levels of risk tolerance. I think this is the key thing that a lot of people keep misunderstanding, because the debate is not even so much about higher skilled players trying to oppress the less skilled ones, but about rewarding people who are willing to gamble. Unlike some of you might claim, figuring out edges in poker is extremely complicated and takes a large sample of hands, and it's the biggest reason why people are willing to gamble money on poker in the first place. In a KoTH like system it's entirely plausible that a solid regular might get dethroned by a player who is -EV against him, but who is willing to make the game enough of a headache for the other player, so that in the end he is awarded with the right to wait for action. I think this is great because it also rewards people who have very annoying and flashy styles of play. What is the reason why PokerStars would not want to reward people who have more gamble in them?

Suppose that this semi-KoTH was enabled tomorrow at 5/10 and now instead of 100 tables we only have 25 (completely arbitrary number, this should be tweaked and determined by PS with time). True, a lot of people who used to play 5/10 are now going to have to find another game to play, but there is no way that some of the people who now open up the list of tables only to discover that there are no empty tables left anymore, are not going to look at a few of the names sitting and think to themselves "With limited amount of tables available, each table has now gone up as real estate because of the higher likelihood of getting a fish. I played this guy two years ago at 6-max and he wasn't that great. I can probably make him quit and get his table and get some of that fish gold as well". Consider that people aren't robots, so we are much less reactive towards mathematics than the whims of our emotions and our ego, so I think that once those two get a game going, if they are even close in terms of skill, then that game is going to go on for a while.

Why do we keep dropping numbers for KotH like '25' or '30' or whatever? What happens if we make it 8? When was the last time you saw 8 unique HU matches going? This should be plenty.

because most of us (i hope) feel a too strict koth system is very bad for the games long term

+1 to "semi-KotH," that is: for every distinct game type, place a cap N on the # of empty tables and allow each player one empty table.

This solution:
- Is trivial to implement.
- Adds no complicating features to the user interface.
- Creates more action by rewarding risk-seeking behavior. (see kotkis's post above)
- Passes the "common-sense" test -- it's how a brick and mortar hu lobby might look.
- Is completely fair to careful game selectors, provided a reasonable N is chosen.

As raptor alludes, the discussion simply becomes one of how big the table cap should be. If necessary, different games/stakes could have different N's. To me, something in the 8-15 range seems reasonable, but even a larger number would be a huge improvement over the status quo.

I second kotkis's challenge to come up with counter-arguments to this system.

You arrive at this conclusion by placing an awful lot of emphasis on something that's unsubstantiated -- namely, the notion that good regs beat rec players MUCH faster than bumhunters. You give estimates of the average rec player's "winrate" vs a top player (-15ptbb/100) and vs a bumhunter (-5ptbb/100):
This is a ridiculous exaggeration. Sure, there's probably some difference, but as long as we're picking numbers out of the air, I'd guess the top player beats the rec player on average 20%, 30%, MAYBE even 50% faster than the bumhunter -- not 300% faster.

The skill gap between a great hu player and a mediocre bumhunter is significant, I agree. But this gap manifests MOSTLY in a high(er) level match between the two, not in the games of each vs a rec player. Weak/rec players require a much simpler strategy to beat, and there is just a lower skill ceiling in these matches. I.e. it takes much less skill to maximize one's EV vs a rec player.

There is some merit to Ben's point awhile back -- that someone who battles regs a lot might even be worse vs rec players. Someone who primarily battles other regs puts most of their time and energy into studying GTO play, which has very little application vs a donk. As someone who loves to battle, I know that I've made plenty of plays that are horrible vs rec players just because I was in "GTO-mode" mentally while simultaneously 4 tabling some reg.

I'd argue that even if donks lose their money slightly faster to the top regs, the costs of allowing things to stay the way they are (ridiculous looking lobbies, less action) outweigh the benefits of having donks play slightly weaker players on average. I say "slightly," because under semi-KOTH, it's not like they'll be playing the #1 crusher every time. If the number of empty tables is capped at something like 20, the rec player is going to end up playing the 10th best hu reg sitting at that moment on average. Not that big of a deal.

Can you expand on that? Let's assume 1 open table per person instead of 1 dude clogging up all 8. I could see that being an issue, but you really think there should be more than 8 people waiting around for fish?

Edit for clarity: I should say more about the scenario I am imagining. Let's say that I want to 4 table Jman. I sit at an empty table, he sits with me. There is an 'add tables' button that we click a few times to get up to our magical number of 4. There are now 7 other tables in the lobby. Let's say I also want to 4 table Urindanger because he is a fish at NL. I sit with him and we click 'add tables' to get up to 4. There are now 6 open tables. Assume that we all want to sit and wait for fish. Now there are 3 open tables. Assume that Jungleman, Durrr, and Trex313 are sitting at those. There are 6 people waiting for games, sitting all alone not playing.

Maybe 12 is a better number, though I don't really see what difference it makes. We could also do something where regardless what games are running, there will always be 8 tables where people can sit alone. There could be 15 HU matches going but still 8 different options for people to play against in the lobby. I think I like this better.

I should also note that I think the KotH system isn't all that great. I am still very much in favor of the option to only have 6 max tables to start, then if people want to play HU they can do this:

Player A sits alone at a 6max table. It fills up. Player A thinks he is very good, and each other player has a box next to their name called 'play HU'. He clicks this box next to each player. Player B joined the table, and after a few orbits, he loses a big hand to Player A. Player B now clicks the 'play HU' button next to player A and a HU table is made! They have the option to click 'add more tables' and if they both click it, another table is added!

Can someone punch some holes in this? This seems like a pretty good way to handle this without hurting the HU games (possibly even helping it, I really like the idea that after I suckout vs someone they can challenge me HU, whereas they may not want to go search through 200 people in the lobby in order to find my HU table).

well go ptr all the bumhunters in a lobby and compare that to some winrates shown by the very best regs that spent the majority of their time playing other regs - or just note that a couple of bumhunters (who are really **** good at it) have winrates of 15ptbb over big samples whereas the vast majority of bumhunters that you ptr hardly have over 5ptbb

ofcourse you could say i'm pulling numbers out of thin air, but i have more hu experience than most in this thread (not trying to appeal to authority/not meant like that, but i'm not just randomly guessing like you seem to think i am). i know my winrate, i can ptr must bumhunters, and i know what some of the very best bumhunters are capable of. i know how hard the average 5/10 pure bumhunters suck (because they do play every once in a while) and you are kidding yourself if you don't think they leave a ton of money on the table, even vs fish

not at 25/50, but at 5/10 yeah. a strict koth would channel more fish action to very good players, i think that leads to less fish depositing for reasons i explained already. in the end this leaves even the very best reg worse off

8 tables at 5/10 is a fairly strick koth system imo given that the 8 best players online at any given time per site are probably all top 10% of the population of ppl sitting 5/10 now at that site (or to phrase that differently, most sites typically have more than 80 5/10 tables open). for 5/10 at stars i'd say say 25 tables?

Don't most fish tend to just have a buyin or 2 in their account and quit when they lose it?

Isn't that the point? I thought we were trying to get rid of all the bumhunters or whatever. A strict KotH system does that. Also, I edited my first post for a little more clarity. Again, I don't think KotH is the best way to go about this, but if it has to be done, allowing 25 open tables doesn't really change the lobby system much.

yeah but like I said i think the negatives of koth in terms of fish experience/future deposits outweigh the positives of a cleaned up lobby (once again terms of fish experience). i don't care for the argument that the best players deserve more fish, and i don't care for the site's bottom line/arguments that there should be more action (although i guess i'm open to that if my self interest dictates it - but i get enough action now)

You seem to think that "improving the games" means "Lets make it so there are more fish for the regs" or "Lets lower the average skill level of the player pool". This is not true. Improving the games means, stopping people from turning the games into a scummy environment that isn't an enjoyable place for regs and recreational players. The solutions put forth in this thread are an effort promote good ethics and to stop: table-blocking, buttoning, sitting out when a fish sits out, clogging the hu lobby etc. I think you would be hard pressed to find someone who will argue that all of this scummy behaviour isn't terrible for the games.

The fact that all of the solutions in this thread are anti-bumhunter and pro-reg is simply because it's the bumhunters who are causing all of these problems. I don't think anyone thinks that these solutions will make it so that it will be like the golden ages again. What these 'fixes' will do is make the games a much less predatory environment and therefore more enjoyable for regs and recreational players. If stars changes some things for the better and it happens to attract more recreational players then that's just a bonus.

Can you elaborate on this? Would be great to hear your non-KOTH structure. Sorry if you already posted it and I missed it.

+1, well said.

Raptor- I agree with you that semi KoTH is not the best solution. The best solution is that the hu lobby be eliminated completely, and a 'play hu' checkbox be implemented at 6m tables. In addition:
-There should be a 'request hu' feature installed in the client such that if two players mutually request each other's screen names a private hu table is started.
-There should be an 'add table' checkbox once a hu table is started so that more tables can conveniently be added.
-There should be a cap on the number of 1 player 6max tables, and this number should be pretty low, < 5.

Unfortunately, I don't think this idea will gain support as it will detract from so many people's bottom line (including my own). I think it is the best long term solution though, both from the point of view of fairness and "what's good for the games", and so I hope people will consider it seriously.

A second best solution is semi KoTH. It seems like this is something we can come to a reasonable consensus for, and so I'm pretty happy to support it. The major point of discussion left is agreeing on a reasonable table cap. I think setting the table cap relies on three criteria:
1) Having an uncluttered lobby attractive to fish
2) Having enough non-full tables so that it's as easy as possible for most players to get action
3) Having enough non-full tables so that a player who wants to play hu has a reasonable selection of opponents.

These criteria are pulling in opposite directions. On criterion 1, we want exactly one open table per limit. On 2, we want one open table per limit as well, although maybe two would be OK if games start frequently. On 3, we want as many tables as possible.

I think it's reasonable that regs and recreational players might have a preference not to play one or two regs at a current limit. For example, say I'm a fish wanting to play hu but the one regular sitting has ran great against me and berated me in chat. I probably won't want to play. However, I don't think it's plausible that a fish wanting to play hu will have these kind of grievances against more than 2 or 3 regulars at a given limit.

So, on a balance of criteria, I think the table cap for semi KoTH should be set fairly low, probably to 3 or 4 tables.

The current system is flawed. Absolute KOTH is flawed. I think a KOTH system with 1/2 table cap per person is a good compromise. Maximum aggregate action should be the ultimate goal.

The key is that regs should be encouraged to battle, but there should be some room for those who aren't necessarily the final bosses of their limit, to still give action to each other.

For example, I played for about 3 hours today vs a reg at 25/50 2 tabling on Party, and then took a break to eat dinner.

When I came back, all the tables were being held by someone who neither of us wanted to play, but they would not give up even 1 table for us to continue our match as we did for several hours, despite them having no action. No one is really at fault here but it results in less action.

Playing at fullring or 6m tables is incredibly frustrating as people will continue to sit out until a fish joins, and they will often force reverse blinds OVER and OVER due to a bug in the 6max tables. Especially when you don't want to play 6max this can be very frustrating.

I do not think a communist type system with table-cap like Stars, or Ipoker is a good solution in fact I think it is the second worse model. The worst model are the random sites that allow bumhunters to sit at infinite tables.

HU Rush poker (global lobby) could be great, up to limits of 5-10. Regs should also be able to invite an opponent somehow to play headsup at multiple tables, when you want to start a 4-tabling match for example. Since there are dramatically more fish at SSNL and MSNL and a far lower % of the aggregate action is regs battling each other, I think rush poker would be the solution to create the largest number of hands played / rake paid to the sites.

1020+ lets say, I think it should be KOTH, with a table cap per player, and a total cap on empty tables.

At 1020+ there is a reasonably wide range of players who range from 2-4 bumhunters to guys slumming it who would rather be playing 50-100 but are bored and want whatever action they can get. I don't think Rush HU (global lobby) would be practical at high-stakes due to a much smaller player pool. I also think a stricter KOTH system would operate better with a smaller table cap, which is fair given these are the larger games. I also think this model makes more sense for 1020+ since often regs want to play multiple tables vs same opponent and deep often for long periods.

Obvious solution to grimming: (should have been implented years ago)
Sitting at table automatically posts blinds, and you are forced to play 1 hand, which high-cards for dealer button. You are free to sit out at any time after that, including after playing your button.

3 hand Button'íng people is still allowed, which is unfortunate, but the result is that the BB would keep the table which should be some form of compensation. I think any system which FORCES people to play a later hand, eg quit on their big blind, would possibly run into legal problems with their responsible gambling laws these sites have to abide by. EG. problem gambler fish loses a big pot on his button, but MUST play his big blind. For this reason I would submit its quite likely the sites must allow players to sit out at any time for legal reasons.