Replay vs Microstakes Cash

Lately I have been playing microstakes cash games on a US-based poker site, finally have a decent sample size, and want to share the results for the sake of comparison to Replay. For players who wonder if Replay is a worthwhile learning experience, how it stacks up to real money poker, or who are just huge data nerds like me, hopefully this comparison is useful for the Replay community. And just for reference, the site I’ve been playing on is known as one of the toughest and is full of eastern European nit-regs, as opposed to some of the other US sites which are known for being extremely soft.

I played nearly 250k hands on Replay over several years across various game types and have 847m chips (no purchases). I was a winning reg at the elite stakes 50k/100k and 100k/200k, but never really made it consistently above that. Unfortunately, there is no way to calculate a win-rate on Replay across stakes and game types, but if someone won 500m chips in 50k hands of 50k/100k they would have a win-rate of 10bb/100 hands.

From over 10.5k hands (a much smaller sample) of 10NL 6-max cash games regular tables, I have a win-rate of over 15bb/100 hands. However, I also played 10NL 6-max blitz tables, and over 11k hands of blitz I was a losing player at slightly better than -5bb/100 hands. For those who are unfamiliar, on blitz tables you move to an entirely new table after every hand, so you are in a different position with a different set of opponents. It makes it easier to play more hands, but my heads up display doesn’t work on the blitz tables and it is much harder to take advantage of table dynamics or reads on opponents. So, the first takeaway from these data: do not play blitz unless you just want to mindlessly gamble or you are a GTO robot who can crush the games without trying to exploit specific opponents. Another fun stat, my red line (non-showdown winnings) on regular tables was -60bbs while on blitz tables it was -890bbs.

The second takeway would be, that from my experience the elite stakes of Replay is a decent parallel to 10NL. Sure, it’s all play money and nobody cares, or so players here have said, but at the elite stakes it takes a fair amount of persistence or skill edge to accumulate hundreds of millions of chips that would cost thousands of dollars to buy. Playing against these players is good practice for adjusting and beating thinking players, especially nit and loose-aggressive player types.

The caveat of this comparison is that the games themselves play differently, so the exploitative adjustments will be different. Because there is effectively no rake on Replay (and because it’s play money) people love to limp, call, and generally play passively preflop, while at 10NL (with 5% rake that is effectively uncapped) there is a lot more 3-betting and 4-betting. In my experience on Replay, 3-betting was a sign of extreme strength from most players (like QQ+/AK) and 4-betting meant KK+. At 10nl players 3-bet closer to an optimal range. My strategy has definitely changed since leaving Replay to include more 3-betting, and I can see how it would be a huge exploitative advantage to 3-bet in position a lot on Replay even though players call too much. Sure, you don’t like getting called when you 3-bet with A5s or 76s or KJs, but if you are playing in position with the initiative, you are going to generate huge profits in the long run instead of just flatting.

So, overall, I think my experience on Replay was extremely useful to gain experience in different spots and against the specific player types that are common here. If you can learn enough to consistently beat 100k/200k then you can easily beat regular table 10nl for a few bucks an hour and possibly 25nl as well (that’s my next step now that I’ve quit playing blitz to focus on building a bankroll at the regular tables). From what I have seen, 25nl has plenty of fish and isn’t that different from 10nl, so I reckon the players with multiple billions here on Replay could beat it fairly easily. This site really is a good opportunity to learn to beat cash games without risking any actual money. Just make sure that you keep trying to play strong balanced poker even if it’s possible to win while taking less risk.

If you have any questions about lessons learned from Replay, what real-money online games are like, or strategy in general, feel free to jump in!


Great Post Joe, I was interested too in how players found Replay vs Real Money. You are the first to go into some depth on your experience. Thanks for that.

imho, and I think confirming what you have also said, I have found that Replay tends to be loose passive playing, certainly at the stakes that I play (not yet elite). While the real money sites tend to be tight aggressive imho, with a lot of shoving with pocket top pairs , with far more 3betting and 4 betting too.

Thanks once again though.


I played a lot of micro stakes for real money around 15 years ago, and think back then it was considerably softer than elite stakes here (and I didn’t even use a HUD until I moved up to higher stakes, was considerably weaker then than I am now, and still crushed the games). Those were the good old days, lol.

I’ve heard from many reliable sources that the games are much, much harder now. Real money online games have been illegal where I am for a long time now, and so I haven’t been able to personally confirm that, but I have no reason to doubt what I’m hearing from many independent, competent sources.

I suspect live $1/$2 and $1/$3 cash games are softer than elite stakes also, though my sample in Vegas is pretty small, at less than 10,000 hands, I’d guess. My experience was that most tables were primarily made up of reasonably competent players that were either regulars or struggling professionals, and about 20% to 60% very weak players, depending on the table. But I’d say my win rate was well over 100bb/100 hands during that sample. But again, over a sample that size, I may have simply been running good (and suspect I was at least to some extent).


From what I’ve seen (and generally heard) live $1/2 is much softer than 10nl, though I also haven’t played much. In my experience it was loose-passive, a bit like replay, so you could try to hit a big hand and get paid without having to make any hero calls or bluffs. I would compare live 1/2 to maybe 10k/20k here. I think my winrate in live cash is also like 100/100 but it’s so slow, I’ve probably only played 1000 hands.

The game has definitely gotten much tougher since the moneymaker days of the early 2000s. It seems to have evolved a lot even since 2017 as everything is based on solvers. Making a 1/3 cbet seemed strange then, and now it might be the most common size (at least at 10nl and from videos I’ve seen of online pros). I learned to play back in the 00s but it would have been nice to take it more seriously back then.

It’s definitely possible to beat 10nl with very little bluffing or bluff catching, but there’s a lot more profit to me made by doing both (well).

1 Like

Thanks, this is an interesting commentary. It is kind of offputting for any RP player who is thinking of playing for money.

I think Holdem Poker was always designed as a game where the strong could predate on the weak to make money, so you always want to be the pro who pretends to be a tourist and find the weakest opposition so as to make money. Preferably rich tourists in Las Vegas who have had plenty to drink and are trying to impress the wife or mistress. Maybe they don’t exist now that the whole world is woke.

You don’t really want to be playing against Eastern European card sharks and out of work pros in games with a rake.

On RP it is a game of wits, but in real money poker you want opponents who play like the lowest levels on RP, but think they are as good as the pros on TV or YouTube.

Maybe the future of online poker really is play money, especially in tournaments which people enjoy playing for their own sake as recreation.

My own recent observation is that there seem to be players on RP who are actually buying chips so that they can enter 1-million play money chip tournaments, which would work out at about $30 per tournament,. so they must enjoy it a lot.

1 Like

I agree that poker players often focus on maximizing profit at the expense of enjoyability, and Replay is great for just having fun. I have been missing the sense of community and the table chat (there is no talk in the chat on real money sites except the occasional insult, although those are frequent on Replay too).

But poker can be about fun and learning even on real money sites. I invested $80 and don’t mind losing a few buy-ins here or there. It is worth it for the fun and experience of facing off against players who are more difficult to beat and aggressive than the players in other games. There are definitely easier ways to make money, but it’s not bad having a hobby that also pays a few bucks an hour.

From a bankroll management perspective, you also don’t have to stress because if you have $40 to invest you can play .1/.2. I started out with the $80 at 5nl (16 buy-ins) but won another 5 buy-ins in about 500 hands and moved up. On the other end, after 750 hands of 25nl I am down a buy-in (recently lost a stack with KK vs QJ that called 3-bet, flopped top two, turned a boat), but still need to increase the bankroll to be comfortable playing that level without thinking about going broke.

There are definitely people on Replay spending hundreds or even thousands of dollars buying chips. To some people that may not be a lot of money, and they are paying to have fun (hopefully) without caring about winning real money. Unfortunately for them, they don’t usually last long in those 1m chip tournaments, though the same is true for real money tournaments, which are much softer than the cash games. People will go nuts and call off their entire stack preflop with 43s or QTo, just like on Replay.

However, in real money microstakes tournaments, the stacks get very short very quickly during the 2+ hour long rebuy periods, so even if you have a huge skill edge it becomes a crapshoot. I’m about breakeven from playing about 15 mtts ($1.5 to $10 buy in). It is easy to min-cash but difficult to build a stack to challenge for the final table without being a complete maniac and running super good. They can be more exciting than cash games, but require a huge time commitment.

1 Like

I would guess that most online players are using software that saves every hand to a database and presents your opponent’s data on a HUD. I know I used such a program 15 years ago, and can only imagine how advanced they must be these days.

This is probably the biggest reason online games are so tough compared to live poker, and won’t be getting any easier.

1 Like

I use a HUD, though not a top of the line one, and it definitely helps a lot to identify fish and for hand review. The HUD is very useful for identifying when players get far out of line, like playing 50% of hands or never folding versus 3-bets.

It is also useful to spot regs, but less useful to actually exploit them. Regs generally have stats like 25/19/8 (VPIP/PFR/3bet), so beyond that it can tricky to exploit them unless they fold too much vs cbets or 3bets.


One of these reasons I like Replay is that it does not support HUDs, which I believe reduces the skill aspect of the game. I like to play poker (recreationally) because it is a form of mental exercise, and using a HUD would defeat the purpose. I know that HUDs are capable of calculating all sorts of frequencies (e.g., VPIP, c-bet rates), and I have to assume that some are also capable of calculating equity, which is a massive advantage versus players without a HUD.


Is that 25/19/8 at six-max?

How many tables do you play? I know a lot of microstakes players multitable so that might skew the numbers since its hard to multitable on replay.

Fun/rec players have NO chance online given how advanced the available HUDs are. Fun players don’t know or care what a HUD is, they just want to put money in and have some fun chasing their cards. But they lose so much now so much it’s impossible for them to having fun. Online poker will eventually be just GTO AI bots playing zero sum against each other, if it isn’t already.


Indeed, there is not much point in playing for fun against a computer.

This gets to the whole issue of whether poker can be considered a recreational sport/pastime or just a business.

I learned to play whist as a child and in the village where I grew up in Europe there would often be “whist drives” where you paid a coin or two to enter and there was a prize like a fruitcake or a chicken at the end of the night for the gentleman winner and the lady winner.

All players played the same cards at the same table, and at the end of each hand the gentlemen moved north and the ladies moved south to another table. When you met your original partner again, then it was all over and the scorecards were tallied and the winners announced with one winner who played the gentlemen’s card and one who played the ladies cards. Sometimes the winning gentleman was a lady as there often seemed to be more unpaired ladies than gentlemen.

I then moved on to learning bridge, which is a more skillful version of whist, (also with resemblances to the game of Spades, but more sophisticated in the scoring, and involving 2 players playing as partners.)

Bridge is also played competitively in clubs usually in the form of Duplicate Bridge–a system similar to a whist drive, except that you keep your partner, but the same cards are played at each table. At the end of the night there are two prizes–one for the best north-south couple, and one for the best east-west couple. (Obviously you cannot play any one set of cards from both directions as players would remember the cards.)

What NL holdem really needs is some kind of tournament format that would allow players to compete against other players who get the same deals, but the details of how this could be implemented are beyond me.

In such a manner the skill of playing the game could be divorced from the idea of making money, and the test would be to outwit other players of one’s own standard rather than to exploit weak players. The luck of the cards element would also be eliminated.

Replay Poker comes as close to achieving this as any poker site I know, because you can play the same tournament every day against roughly the same set of opponents, and over the long term your skill level will be reflected in your bankroll, which will be either increasing, declining, or remaining about the same and you will see the same in the rankings of opponents.

If you are successful, you can move up to the next level–a MTT that is more expensive to enter–until eventually you reach a point where there is nowhere else to go. The 1-million, 2 1/2 million, and 5-million chips tournaments in RP have fields drawn from exactly the same subset of opponents.

1 Like

Yeah that is for 6max. There are almost no full ring cash games on the site I play.

As far as huds go, they shouldn’t be able to provide equity, as they shouldn’t be reading hand data until after the hand is over. If a program is providing real time assistance, that would be cheating, although players certainly could have such programs.

I agree that HUD takes away a lot of the memory aspect of the game, but it still requires a slightly different skill to apply the data at the table. If a player has a VPIP of 50 they play too many hands, but does that mean you want to tighten up or loosen up against them preflop? Does that mean you can blow them off hands postflop or do they call too much postflop too? Huds obviously provide a huge advantage, but sorting through the wilderness of data is a fun way to learn without being subject to the bias of memory.


I thought that if a HUD is capable of storing hand histories, it would be able to construct your opponent’s range (based on shown hands), from which equity could be calculated.

1 Like

I play here for fun when i play for real i dont play the same - our poker games have been closed since COVID but if i was going to play a MTT or a cash game then would play 100K sit and go before as they felt like a good warm up

1 Like

The one I used 15 years ago stored every hand I played and every hand I saw. I could replay any hand at any time, but the accuracy of the tendency/frequency data was directly tied, not to millions of hands from the general player pool, but to each specific opponent.

In practice, this meant I never had a large enough sample size to make the information all that useful. Complicating matters was that losing players either learned fast and changed their game or gave up and either quit or dropped to lower stakes. The data showed how they had played historically, not how they were playing now.

Since there is some table banter, no HUDs, and less churn, Replay probably plays more like live games than online real money games.


Another factor that could be related to tougher games these days online is that online poker is more or less illegal in the US, so millions of players have left the scene and don’t want to be involved in anything that borders on the illegal.

Although I am not currently in the US and am in a country where pokers is legal (I think), my credit cards are US based and I would not want to get a card canceled, even if that is not very likely.


Often in poker articles, I have read that a great winrate in microstakes cash games would be 5 to 10 bb/100 (big blinds per 100 hands). But, after 100 hands with 9 players at the table, 11 orbits would have completed. 11 orbits X 1.5 bb per orbit would come to 16.5 bb posted during the 100 hands.
So, you have this awesome 5 bb/100 winrate, but you’re paying 16.5 bb during each block of 100 hands. How does anyone win at this game?
To me, it looks like you need an average winrate of 16.5 bb/100 just to break even.

Quoted win rates include the blinds you are paying. So if my win rate at 10nl is 15/100, that means I’m winning 15bbs beyond the blinds and the rake.

But you can’t really count the blinds that way because they aren’t dead money. You can still win them back in the hands and they affect your pot odds to continue to get involved and try to win the pot.

Rake on the other hand just takes 5% from your winnings and is effectively uncapped. So while my winnings should only be 5% lower, the effect compounds over time because you have less winnings to build more winnings from. But the rake is also built into that winrate. The site also has a rakeback program, so a small percentage of that rake goes back to me (which is not included in the winrate).

1 Like