This article is based on original research carried out for GameIntel, a subscription-based poker industry publication. If you’re interested in getting the inside scoop on such stories in future, click here.

In case you’ve been living under a rock, cryptocurrency is all the rage in the poker community these days. People joke about adjusting their calling ranges based on how Bitcoin’s doing that week, Doug Polk is now more famous as a self-styled crypto guru on YouTube than as a poker player, and whether the markets are way up or way down at any given moment, the schadenfreude on Twitter is palpable.

One natural result of the craze is that there’s a lot of interest in the possibility to use cryptocurrency technology for poker. The most significant effort in this regard so far has been CoinPoker, which uses Ethereum’s “smart contract” technology to run its poker tables and has its own cryptocurrency CHP, now available for trade on no less than five crypto exchanges.

CoinPoker went live for real money play a few weeks ago – or “real” money, depending on your opinions about cryptocurrency. During its Initial Coin Offering (ICO), 125 million CHP were set aside for the company itself; some of these went to pay the team, some went to exchanges in return for being listed, and a big chunk of the rest will be given away through freerolls and promotions to try to encourage people to use the site.

At the moment, CoinPoker is rake-free, and the first of these promotions is running now: a set of weekly and monthly leaderboards with three stake tiers, which will give away a total of 5 million CHP. At current exchange rates, that’s about $125,000 USD.

The leaderboards are essentially a rake race, except that there’s no rake. Instead, it’s just the number of hands played that’s being counted, with the added requirement that players must voluntarily contribute chips to at least 10% of pots in order to qualify.

No rake and whopping piles of strings-free crypto cash being given away sounds like a very appealing deal, and if you open up to CoinPoker client, you’ll see that there is in fact a lot of action going on, at stakes as low as 0.01/0.02 CHP up to 5/10 CHP.

And yet, if you scratch just a little below the surface, you’ll find that things are not right. At all.

Breaking the story

These days, most of the work I do is not for PartTimePoker, but rather another company called GameIntel. Most readers will probably not have heard of GameIntel, both because it’s new and because it’s a B2B company serving the poker industry, rather than poker players. You may know of PokerScout; GameIntel provides the back end data service for PokerScout’s website, and until this year, they were the same company.

My job with GameIntel is to create a twice-weekly newsletter called the Poker Update, with a focus on site traffic data. Technically, my title is “managing editor,” but as we don’t have any other writers, I essentially create all the content myself with input from the owner, Dan Stewart.

Looking for a story this week, I opened up the CoinPoker client and was surprised to see how busy the tables were, and pitched that to Dan. He agreed that it was big news that we needed to cover, but as soon as he started poking around, he told me the numbers seemed fishy. After taking a closer look, I concurred.

Abnormal traffic patterns

Virtually all poker sites and networks exhibit some of the same patterns. Traffic follows a relatively smooth up-and-down 24-hour cycle, with the peak coming sometime in the late evening for whatever time zone the core customer base is located in. For sites with concentrations of users in two distinct time zones, occasionally a double peak can emerge, but there’s still a smooth, recurring pattern which is generally only disrupted by special events like holidays or technical problems.

Furthermore, it’s generally the case that most active tables are full, and that the percentage of less-than-full tables is fairly steady. Tables with empty seats either have those seats filled, or break before long.

In both these regards, CoinPoker’s traffic is very unusual. Although there is something of a peak, whose timing suggests a user base primarily located in Asia, it’s quite broad, and traffic makes sudden jumps of 100 seats or more, upwards or downwards, often with very little change in between.

These jumps also coincide with drastic changes in the percentage of full tables, from a fairly normal 80% or so, down to less than 50%. During those periods when the percentage is high, all the full tables have significant waiting lists. During those periods when it’s low, most of the partially-filled tables have the same number of users, which is often 5/6 seats filled during peak hours.

Massive multi-tabling

Both of these unusual trends point to the same story: that there are few, if any CoinPoker users behaving in a normal fashion, browsing the lobby for empty seats and starting new tables if they can’t find one. Rather, the sudden jumps in both traffic and filled-table ratio are consistent with users, or groups of users, either attempting to sit at, or leaving huge blocks of tables all at once.

Opening a few tables, the cause becomes obvious; at virtually any stake level and game you decide to look at, you’ll find almost exactly the same lineup of players at every table. The only time you see some variation is when there are waiting lists, and even then, most tables will comprise, say, six players from a pool of eight, with the other two presumably being the ones on the waiting list.

At first, I wasn’t convinced this was all that unusual, because you wouldn’t expect CoinPoker to attract a conventional crowd of recreational players. Rather, the overlap between cryptocurrency enthusiasts and poker lies mostly with the more serious, professionally-minded crowd. It’s natural that multi-tabling would be the norm on CoinPoker, even if the majority of users at other sites play only one table at a time.

Then I decided to count just how many tables individual users were playing, and then I definitely knew something was up. Virtually every player who was multitabling was at more than 20 tables. I even found one user with the screen name “1st” who was at 42 tables for at least eight hours straight.

There was also very little variation between stake levels, except for the fact that only tables at the minimum stakes for each leaderboard bracket (0.01/0.02, 0.50/1 and 5/10) were heavily populated. The terms of the promotion specify that each user can only win prizes from a single leaderboard, so naturally the users were different at the different stake levels, but if there were, say, 27 0.01/0.02 tables open, there would be something close to that number at 0.50/1 and 5/10 as well.

What’s going on here?

None of this adds up in any non-sketchy way. There are some people out there capable of playing 20+ tables at once, but I’m not convinced anyone can play 42 successfully. That’s with software assistance, and to my knowledge there’s no decent third-party HUD available for CoinPoker or TonyBet, on which its software is based.

Moreover, no one who can play that many tables would be interested in playing 0.01/0.02 CHP, where the big blind equates to roughly one-sixth of a penny. Nor would you be likely to find six high-stakes players all sufficiently confident of their edge that they’d be willing to play twenty-plus simultaneous tables of 5/10 against each other.

The most obvious answer would be that most if not all of the players are bots, but after watching them play, I’m not convinced that’s the case. It could be, but it’s not necessary to explain how and why they would be playing so many tables.

Last night there were 27 tables of 5/10 running, all of which had five of the same players at them: “sikoslovake,” “IgotTricks,” “upay4me,” “ado”, “Efte” and “pracka.” The remaining seat was usually filled by either “muchloveforu” or “Zantonsus,” though a few tables had a sixth player who I didn’t think was part of the team.

I opened sixteen of those tables at random, the most that could fit conveniently on my screen, and watched for about half an hour. The preflop play was tight, but not unreasonable. Postflop, however, almost every single pot was either checked down or won by a single bet. The two exceptions I saw were one hand involving a player who definitely wasn’t part of the group, and one where a King-high flush lost all-in to an Ace-high flush on an unpaired board. That latter case may simply have been a token effort to not be completely obvious.

Furthermore, all of the players bought in for 40 blinds whenever necessary, and none ever got much about 200, despite playing 27 tables for hours and hours. In other words, the players were clearly soft-playing each other, just swapping chips around in order to satisfy the promotion’s 10% voluntarily put-in-pot requirement while racking up an insane number of hands.

We’re only 17 days into the month, and the Low and Mid leaderboards both have multiple users with over 100,000 hands played, including three over 140,000, while the High tier is led by “IgotTricks” with 65,000 hands played.

Don’t know or don’t care?

The terms and conditions of the promotion claim the following:

“We will be reviewing the most active players every day to ensure transparent and fair games. No abuse of any kind will be tolerated.”

It’s hard to see how that could possibly be the case, however, when the soft play is so transparent on even casual inspection, and players are sitting at so many tables at once against the same opponents, yet not very much money is actually changing hands in the end.

I got in touch with a representative from CoinPoker, who was initially polite though not very helpful, and who stopped responding when I asked about “1st” and his 42 simultaneous tables.

I also inquired about CoinPoker’s “Milestones” timeline, where October is specified as the target date for their crowdsourced, peer-to-peer security system, which will task players with policing their own games. In particular, since October is still many months off and real-money games are already running, I asked whether there was an in-house security solution in place in the interim, or whether things were effectively unsupervised for the time being. That query also got no response.

Sadly predictable

When CoinPoker was first announced, I was skeptical about the business model. The ICO raised capital for the company in the form of first Euros and then Ethereum, but although the division of this money as laid out on the website included payments to the team and its advisors, there was none clearly earmarked for salaries going forward.

I pointed out at the time that there was a huge moral hazard at play, in that the CHP purchased during the ICO would only retain any value if CoinPoker succeeded, yet the people tasked with its success were receiving their compensation up front and would not have much committing them to the endeavor if it were not immediately successful.

The way things are playing out so far, it looks like my fears may be realized. The company is adhering to its milestones and delivering the product it promised, but doesn’t seem to be doing very much to support it or ensure its integrity. The lights are on, but no one is home.

Unfortunately, this laissez-faire attitude may bring the house of cards crashing down sooner than those involved with the company expect. It doesn’t look like very many people bought CHP during the ICO with the intention of competing seriously with it; rather, some are likely sitting on piles as a speculative investment, while others apparently bought in so as to be able to take advantage of the company’s loss leaders like this 5M CHP leaderboard promotion.

The trouble with that is that these players probably have little intention of keeping the CHP balances once the free money runs out. Instead, they’ll try to convert their CHP to ETH and from there into other cryptocurrencies, or back into conventional money. Unless there miraculously turn out to be buyers who want to play crypto-poker for real, all that CHP being put on the market at once is going to cause the price to plummet, possibly leading to panic among those who invested speculatively. Depending how quickly they get out, even those who win the leaderboards may find their winnings worthless and may not recoup the money they shelled out for their CHP bankroll in the first place.

I’d like to be wrong about this, but at this point, all signs point to this endeavor ending in tears for everyone involved. Everyone, that is, except the founders, who can take the Ether and Euros they raised and ride off into the sunset.

Alex Weldon (@benefactumgames) is a freelance writer, game designer and semipro poker player from Dartmouth, Nova Scotia, Canada.