If you want to play in a good game, there aren't many of them. If you want to play in several games, they won't be good. I want to make a poker site that has many, good, games.

My general strategy is to keep a similar retention plan to Unibet, add a deeper understanding of what liquidity is, automate away a lot of the work that B2C sites normally carry out, and invest the remainder into marketing.

As well as this account, I was also UnibetAndrew. I've worked in gambling, mostly poker, for 19 years. I know how to make a poker site, I've done so in the past. I know how to project/product manage it, how to hire/manage a team, etc: basically, how to get it all done.

But what I don't know how to do is raise the money to start. I need: 10 people (mostly developers), 12 months to reach beta, and a little under €700k. After that, I'd need about €2.4m for a 3 month open beta, and as low a risk launch as I can model. I'd break even in month +2, and be overall in the green by month +11 (running at about 30% margin, though I'd want to spend that on marketing, there's no need to stop at that size).

I figure I'd 'sell' some percentage of the company for that initial dev cost, then some more for the launch cost - it's easier to sell something that exists, rather than something I've only modelled/written the specs for.

I've built a pitch. I have a CTO (as long as he doesn't find anything else meanwhile), and we have good ideas of who we'd like to hire first.

I've approached a couple of people who might be interested in investing, but no luck so far. Ideally, I'd talk to people who know gambling and poker, and would be interested in investing on a longer term basis than e.g. a VC firm would (from my understanding of them, anyway).

Does anyone have any leads or advice?

I'd obviously love to get started, and I'd like to be as open as possible about what we're doing/how it's going, as that's the kind of thread I'd be interested in reading myself. If and until there's any funding, I'd prefer not to talk about my precise strategy in this thread, but I'm happy to discuss basically anything else related to this.

Which licence / jurisdiction(s) are you shooting for? Presume you are doing real money, not social / freemium.

Breakeven after 2 months; I guess you don’t want to give specifics but are you targetting existing player pools or new player pools?

2400 dev days to get a product to market. Doesn’t feel like enough.

The last guy to try this was smart and we’ll funded and spat out a very good poker site, but with a lot more dev days as I understand it.

There’s a poker site launching shortly (currently being built in stealth mode) and they’ve apparently been working with a team a multiple of your size for two years: will be interesting to see how they do.

Why have you decided not to just go b2b and get to market cheaper and more reliable?

I'd do IOM & UK, both for strategy reasons.

As you know, you need a lot of players in for day one. The minimum liquidity requirements for a site are hard to build so quickly. So you have little choice but to do both.

Relax built v1 of their poker product in nine months with a similar number of developers. They built v2 and v3 in a year each, each from scratch, with fewer developers. I worked on the product side of two of those launches, and our CTO worked on two as well. We know the architecture that's required, and how to get it built.

Alongside the tech simplifications we have learnt from building several sites before, I also intend to make several major product simplifications compared to most sites. People have spent a lot of time developing features that just aren't that important anymore.

If your last guy was RIO: as far as I'm aware, nobody there had done it before from a tech perspective, and I assume they had SOME kinds of development problems along the way (given what I heard before launch - e.g. a change of CEO). The site is indeed very nice, but our strategy would be rather different.

There are big efficiencies from controlling both sides yourself - e.g. fewer levels of bureaucracy is important for iteration speed, or having control of the whole KYC/fraud/RG/AML process saves many manhours. By controlling the 'front office' as well as the back, you can integrate your processes much better. I want to automate a lot of traditional B2C work away too, which only works when you're integrated.

I also don't think it's trivial to get to market more reliably. Let's say your run rate is ~100k/month, roughly in line with what I imply with my timelines before. You charge 10% fee to someone, and you need them to make €1m/month just to survive. Most operators can't do that, and due to liquidity requirements, most people smaller than that are a LOT smaller. You're at someone else's mercy, and if they can make it work, they're now the ones making all the profit - which they may choose to not reinvest in marketing. The incentives aren't perfectly aligned.

And a network - a shared liquidity network - is for sure out. The network model doesn't/can't work without falling afoul of anti-competition laws, or being severely crippled by intra-skin cannibalisation.

Some kind of B2B system where e.g. several sites can operate your software independently is possible, but you need several such customers lined up pretty soon (run rate problems). I'd prefer to have this as an option post-launch, e.g. if someone wanted to offer poker alongside their casino in a closed liquidity US state or whatever in 2023-4. I have some leads along those lines, on those timescales.

1. You need a thorough business plan with 5 yr pro formas and industry comparisons, marketing plan, corporate layout and structure.

2. Business mgmt. You mention technical aspects but investors are not dumping money into a group of techies without actual experience of running a business successfully.

It might be cheaper, and surely much faster, to try and buy existing software. Operate the company to generate some revenue while you reverse engineer it to be what you want it to be.

V.C. will eat you. You'll probably need to raise seed money internally to get some sort of prototype if you want to start from scratch.

I was an investment banker, ran/operated a hedge fund for 25 years. I've been an investor for longer. You will have no other option than to sell off a fair percent of the company if you seek 100% outside funding.

You're best bet is to start working with a consultant that has extensive I.B. background and someone experienced with writing business plans. If you've never written an SEC, or its local equivalent, compliant business plan I would highly recommend you find someone to do it for you. Simple phraseology can be the difference between "oops, we were wrong sorry you lost your money" and class action against the company and yourself, jointly and severally.

Yep, done all that, particularly the modelling (I like modelling). Business plan isn't yet of the formal kind that I'd need to run by e.g. a VC firm - it's models + a pitch + a rather extensive (private) wiki. So the content is there.

We've both been independent consultants, but more importantly, have both run teams of around this size, with a lot of latitude. We've also both been moderately senior in small & successful companies.

I've looked at this a bit, but most poker software is ancient and takes almost as much effort to run these days as it would to make something new from scratch.

A large part of why I think this will work is that very factor: most people's software is terrible, particularly behind the scenes. Another is the general strategy: it's not something you can easily build on top of existing software. The only software I'm aware of that might fit would be Unibet's and perhaps RIO's. The former isn't for sale, and the latter is probably out of my price range if it were.

Not that you imply it, but to extend my response a bit to a hypothetical person who did: I'm not interested in running a mediocre site that fails for similar reasons most of the industry fails/has failed. I think it can be done a lot better than anyone currently is: tech stack, strategy, implementation. If this were ten years ago, I'd be saying 'disrupt' a lot.

Yes, no doubt. I have a couple of avenues for advice - in one place from someone who has managed 20+ raises and built a poker site before, back in the day. When we get anywhere near money, I will involve them, as I don't know what I'm doing on this side.

I'm rather hoping to avoid having to go the VC route, because I'd find it hard to align our interests properly. But if I do, I shall follow your advice.

Vulture capitalists will come in and analyze your biz plan. Determine a range of valuation and then demand a big discount off the bottom of the valuation. Terrible way to fund.

If you can get seed money going and some level of prototype, then best option is to seek an investment bank that works in the sector. They will do private and public raises and if they like your deal, honestly like, they'll get your funding.

Like VC, some banks are very greedy and charge extortion like fees. Banks that have been active members of the syndicates for other deals are probably your beat option. The lead on deals is just as likely to take your plan and try and use it to help existing deals if applicable as they are to sign you on for real.

The amount of money you're looking for is relatively small. A good way of finding funding would be to look at your competitors that are public, go to the large shareholder list, and make a note of the funds that own 5% or more. You can get their direct contact info from a search most of the time. Then cold call them and see if they will look at your deal. If they've had big success in the past, they are probably very open to more investment in the space.

The first monies in are your most expensive, so make sure that raise equals the needs of the level. Its lightyears easier to get 5 MM in pieces like 750K, 1.2, then the balance as important milestones are reached. IF you were to get it all at once and upfront, you will pay far too much for it .

If you get to that stage, happy to give an outside opinion on it from the side of a fund manager, not for a fee or any compensation.

I know you have a ton of experience in the industry but it surprises me that you think you will be break even in/by month 2 of launch and have all costs paid for by month 11. As you fully well know its insanely expensive to run a poker site and most poker sites operating today are at an intentional loss to cheaply convert players to their casino vertical where normally the CPA is many X higher. I also think your costs are way too low for a ground up build. Plus marketing spend will be many X higher than dev build.

You will also need more than just dev imo. You will need a solid backend admin (maybe you have seen enough to mock up an entire one yourself and have it built for you). You will also need to have game security come in early on to start building their tools and integrating them into the client - but I see you are in Tallinn now, so maybe you are working with someone who previously worked with MPN on this already

There are lots of projects around of people trying to build the next big site. I did some hours for one site that just seem to disappear during discussions. Another company recently paid me upfront for a proposal, which I sent to them and then just never heard back - they had previously contacted me last year too and then went silent also. I think people who do not have many years in the industry just do not realize how expensive a project like this is

shark tank

Hungover so incoherent, but let's try anyway:

Good idea, thanks.

The milestones I had in mind were alpha, open beta, launch. I haven't put much effort into thinking how long it would take to get to alpha yet (i.e. first hand dealt), but it makes sense that I do.

Thanks, if I get to this point I'll send you a PM.

It's possible that I've missed some costs in my model, but I (of course) don't think I have.

But the biggest risk in this project is hitting that critical mass early on. Open beta will never get to too many players, but the launch needs several hundred new players each day for the first few days. If you can get to the point you're hitting >1.5k actives per day, you should be fine.

That's why most of my cost is getting to that as quickly as humanly possible: if that fails - e.g. CPAs are 400 instead of 150 or whatever, or we can get players at 150 CPA, but only 100 a day - the project is in a lot of trouble.

I've spoken to a couple of people who have done poker acquisition at that scale before, and they both think it'll be fine with my spend/plan. It still worries me though.

I don't use LTV per se in my modelling (as it doesn't take win/loss or liquidity into account), but I'll use it here for illustration: I know/knew the LTVs of players at several sites, and it's definitely doable. I also know the casino CPA/LTVs relative to poker at one site, and the actual profit per player is very close to the same. The LTVs there are in line with the other sites I know it of, but I don't know e.g. what PokerStars are at these days, or what Party are at, so perhaps I am missing useful information here.

I agree that most sites are very expensive to run though, and many are losing money. But this is precisely why I think this will work: a modern tech stack and a lot of process automation means we can do it more cheaply. I heard that in 2006 Party had almost a thousand developers - if any of that infrastructure still exists for them, it would be impossible to run with ten.

I agree: I need something like 700k to get to open beta, 400k to run the open beta for three months, a further 400k as safety buffer, then 1.6m to launch - and we'd spend all of that before the first week is out. Then we'd aggressively reinvest revenue in acquisition. My model says this works without much risk (except for those I already discussed, and/or the site not working on day 1, which would also be a disaster - and is why there's a 3 month beta & safety buffer).

The plan includes one person on devops/infrastructure, and another on UX/design, alongside the normal QA + devs. CTO can do the former too, and we're faaaairly confident this will be enough to get done.

I designed the Relax bot/collusion detection systems (that Unibet use), so I'm not worried about this side. The plan is definitely to integrate the data collection/etc from day 1, though I'd probably be happier outsourcing the actual analytics - it's a big project. I know of at least two companies that do it, but I don't know how good they are. Obviously that's something that I'd need to put some time into if/when we get started. If all else fails, I can build it myself.

I also think the opposite is true to some extent - people who built sites 10 or 15 years ago underestimate what's possible these days. Relax have proven it's possible to do it in under a year, and we have learnt from what they've done, having both worked there at various points.

That said, a lot of people say it will take longer than I expect. It's definitely something I need to worry about/spend more time on, but I think I have good reasons to think as I do.

There's no space to reply with this on Twitter, so I'll use this thread instead.



To be honest, I don't understand how crypto poker sites hope to reach minimum required liquidity. You need (well) over a thousand actives a day to be efficient, and none of them do that I've seen. The key isn't the payment method, it's the acquisition. Not having to worry about local licensing laws is fine and all, but that means you can't do proper acquisition marketing there either. Banner ads don't cut it.

The only real exceptions I've seen are some of the, shall we say, secretive Asian sites. They have their own acquisition networks though, which you can't really replicate. Or you can do a deal with the owners of those networks I guess, but that's easier said than done. I don't know the early history of GG poker very well, and I hope I'm not slandering them here, so I'll be a bit careful with my words. My understanding is that they had links to that kind of network. My understanding is that they provided the software, then with the profits, set out on their own. I think the first few posts in their 2 + 2 thread refer to these times too. It's not just a matter of accepting players from black markets.

That said, I did hear of an asian-facing site that's paying out rakeback as crypto, which is kind of fun from a marketing perspective. "Select rakeback payout method: crypto, balance, GiveWell."

RIO: I agree, great product. But nothing like enough marketing spend early on. I would like to spend more on marketing in the first week than I would need to spend to develop the site in the prior 15 months. You can't get by on word of mouth alone.

Interesting thread, although I'm interested to hear your strategy to achieve "I want to make a poker site that has many, good, games" - to me, that requires either a ton of marketing capital to attract recs, or some promotion of non-vanilla NLHE games, or other games full stop, as the increasingly dull, solved nature of NLHE prevents the very possibility of good games. Gl with your project

You definitely need a lot of acquisition marketing at the start, and after that I'd spend a high proportion of revenue on marketing too: the plan is not to attract players by word of mouth (though I'll take it if I can get it - if I get close to raising money, I'll be as open about what we're doing as possible).

I have three pillars to my strategy, and one of them is about keeping weak players around for as long as possible. For example, most (but not all!) of your retention marketing spending should be going towards weak players. This was the plan we used at Unibet, and I have a few improvements I'd like to add. Stronger players will come in order to play against the weak.

Dull/solved NLHE is something we can deal with via product improvements. It's too complex for launch I think - the first step is to do the basics really well. But I have the outline of a plan for this in my product wiki, so that we can build our initial architecture with it in mind. These kinds of ideas would be ruled out by system limitations at Unibet.

https://anw.ee/2020/11/19/poker-rta-solutions/ has some stuff I was thinking about a year ago on this topic.

Hey Andrew,

Looking through that link - I disagree with showing burning cards being a useful measure to protect players. The weak players would be the least likely to correctly use information by showing burn cards, and while some pros would learn to incorporate it into their strategy --- we all know someone would [instantly] build a HUD that could give pros adjusted odds in real time.

That being said, from a site management perspective ----- exposing burn cards or begging the question of why we need burn cards for online poker is bad. [There is no NEED for burn cards.]

Since most poker players are experts in random number generators, you should have a customer support staff team just to deal with those questions and suggestions. ---- I personally wouldn't bother making the type of RNG be part of a sites marketing strategy, but Provably Fair is always nice to haven. [I believe near zero players will inspect the seed/hash, but its one of those things that will make them cozy either way.]

I think you definitely need to certify your RNG. It'll be part of the licensing process anyway, and I agree there's no point making it part of the marketing strategy. People who think online poker is rigged will always find some hands that trigger their confirmation bias.

The idea/example of showing a burn card isn't about showing the site is fair, or about needing a burn card like you might in live poker. It's about breaking the solver's solution. It's GOOD if someone has to think about how to play their hand, rather than do whatever they learnt the solver told them to do. I'd be quite surprised if players interpreted it as something RNG related, but perhaps you're right.

If a weak player adjusts to the revealed burn card worse than a good player - well, that's still better than a weak player vs. a player who has memorised a solver position. It also raises the question of what a good player actually is: someone who has memorised more positions, or someone who can think on the fly? I obviously prefer the latter, and I think most players do too.

There's no real need for a HUD - you can't re-calculate the solver's solution on the fly, and if we show the burn card to all players, you don't need a program to remember it for you. You could pre-calculate 52 more solutions, one for each burn card, but people aren't likely to remember 52 different solutions for each preflop spot (and it'd be very, very, suspicious if someone made next to no GTO errors preflop even with burn cards).

I haven't played on Unibet in a while. Looks really similar Cubeia. Do they have games other than NLH / PLO? PLO5?

PL Drawmaha / Sviten Special seem like easy wins for them and OFC

Hey!

I am always impressed with this thought process. Seen it a lot.
Why do ppl think that they can save recs from gto players. There's no need for it.
Tbh to recs it is the best if a reg played gto this way they lose the slowest way.

Real regs go the expoitive route and destroy weak players really fast.

Just my 2 cents!

That rule changes section of what you linked to is interesting stuff. Would say that they tried to introduce a fifth suit to games like bridge back 80+ years ago and it was universally hated, so wouldn't try that, but yeah, more things like chucking in wild cards would throw people off. One thing I've thought would be interesting is to have a 53 card deck where one card is duplicated, which changes every 30 minutes or something. Would have an interesting dynamic of possible information assymmetry (wait, I folded the seven of hearts and there's one on the flop, let's play hands with sevens and hearts more) and then the adaptation if/when it becomes open knowledge due to the duplicated card appearing twice in a flop/on flop and in players' hand at showdown. Becomes super fun if two people start smashing it in when an ace is duped and they both think they have the nut flush

Of course, the correct answer is that there are hundreds of games that are poker that aren't NLHE, other software has had them for years, they've just never tried to promote it. Stars is the biggest culprit for this, rather than promote all the mix games they already spread, they just keep tweaking what is essentially NLHE

Just to nitpick on this one There is a misconception that regs are just learning off GTO solutions and that NLH is close to solved. It's very difficult for a human to do a good job of replicating GTO solutions. At 200NL/500NL the guys with the biggest winrates tend to play quite exploitatively which would indicate that the pools are still quite far away from GTO. Below 200NL, virtually no one is playing close to GTO.

Stars went through a phase where they invested heavily in introducing new variants and nothing really caught on (except for maybe short deck) presumably because people just tend to gravitate to the games they see on tv or play at their casino.

Anyways, GL with the site.

The first problem is that if you offer a game, it has to run. Are there enough players in the world of (xyz) game to run? How do you reach them? Or are you trying to convert existing NLHE/PLO players to the game? If you do, are you gaining by doing so?

The second problem is that development work that touches deep parts of your game engine is much, much, slower than the normal work. You'd be spending a ton of time - and taking some risks of major, company-breaking bugs - to make that change.

That's why I'm interested in 'RTA breaking' temporary rule changes, long before we've built anything. We can build the original game engine with this stuff in mind, ensuring we can make changes without the development downsides that e.g. Unibet would (probably) have.

I'm always impressed with people being rude.

I like that idea as it's a consistent rule. It does a lot of what a burn card would, but it probably introduces more thinking about hands.

My general day to day plan is to give people more of what they already want, rather than try to give them something they don't know they want yet. You CAN do the latter for a couple of key parts of your software, but you can't run it all the way through it. That's largely what I wanted to do at Unibet - no table selection, alias changes, etc. These are things that many players didn't know they wanted, but it means that almost everything else has to be more of what they already know they want.

That makes it very hard to introduce new game variants/sell to e.g. your own existing mixed games. I think Stars also have problems with the innovators dilemma - or, if it ain't broke, don't fix it.

First, you're right that I should be more careful with how I phrase this.

Yes, but it's trivial for a bot/assistant to do it. That's what I'm the most interested in going after, so my earlier posts about 'regs' vs 'recs' (both are terms I don't like) aren't ideal.

But even for a human, it's relevant to mess up solver solutions that someone has memorised. Sure, you can exploit opponent's mistakes. But you're STILL playing the majority of your hands inline with what a solver would suggest. It's not possible to play 100% of your decisions exploitatively, and the more that GTO knowledge permeates the playerbase (which it will even if just by osmosis & natural selection), the more that's true.

It's just a matter of time. Look at limit holdem - basically dead because of GTO. Or at how ICM knowledge is so widespread in SNG these days. It started in $60+ SNG and in a couple of years it was in $1 games.

To be fair, Rush & Expresso are both variants of NLHE and did pretty well. I don't think it's a fool's errand to try to find a third, as the rewards are so high. Though notice that both of those games kept much the same rules, but changed the mechanics of how they're played.

But that isn't what I'd go for - having a NLHE game where e.g. a different card is duplicated each 30 minutes, is pretty much NLHE but with a little more thinking required.

may i mention i highly doubt it!

6max 100bb+ NLHE is way more compicated than those other variants. in limit holdem you can not chose infinite betsizes. in sngs they play really short stacked (ofc icm brings some trouble).

due to it's complexity majority of ppl who play those low limits will never put in enough work/time to get even close gto. and most of ppl can try their best if they are not wired that way they can't understand the concepts.

otherwise if your prediction came true it would not worth the time to learn gto just to earn pennies on those low limits.

i guess time will tell.

ps: i did not mean to be rude but i admit life/truth is tough. anyway i wish u the best with your project/business!

Nor would I. The problem with an RNG is the software can ignore it. You prove the RNG is fair, the riggies will move to claiming you are rigging the deal. And then the distribution of the deal.

When/if you get to that point, I know someone that might be a good fit for you. They have software that detects RTA and perfect/near perfect play. Currently, it's used after the fact to study action but it ca be integrated I to the software to detect it live in real time.

I love mixed games and would love to see a site spread them. It would have to have a fairly strong marketing budget. I'm in a private club and we have tables run 24/7 but its typically no more than 45 or 50 players. It could be done with substantial marketing budget. I am looking at buying a skin on a major site, no clue how to develop from scratch. Even with having built in traffic, support, etc...I still figure it's at least 100K in monthly advertising/marketing.

Hope you get it off the ground and document along the way. Be interesting to follow. 100000%

Riggies do riggie shizz. That is the exact progression of riggie babbling in the riggie thread. There is always one thing a site doesn't do that is their "proof". There is a cadre of them that bring the fair deal whine up every couple of months.

They are, they just don't know it. Below that, there's still going to be players making mistakes, and GTO would tell you to exploit the **** out of what they're doing