In 2016, Roman Romanovsky (better known as Romeopro) began a charity marathon on GipsyTeam that lasted three years.
Initially, the goal was $600,000 profit, and completion in 666 days. Two thirds of the profit was earmarked for charity.
He exceeded his goal, showing $1.5 million in profit on Sharkscope and donating $1 million to charity.
In September 2019, Romeopro announced that he was playing the final session of his career.
However, he couldn't completely quit poker. In April 2022, he started selling his 26-hour course for $33,000, and announced a new marathon: to win $5 million in 666 sessions.
This marathon was never meant to end. Roman posted screenshots of his progress for a while, but stopped updating the thread in May 2023.
Recently, Romeo had been increasingly spotted at stakes that seemed mismatched with his status as one of the world's top high rollers. Rumors began to circulate that he was struggling, but no one realized the true extent of the problem.
On March 10, Romeopro opened a thread on GipsyTeam, admitting that he had more than $3.5 million in debt.
Roman explained the impressive debts by the following reasons:
- Crazy level of infantilism and excessive self-confidence
- Underestimation of potential scales of dispersion
- Exorbitant living expenses. Money was constantly being taken out of the bankroll, lowering expectations.
- Incredibly bad run at high stakes ($5k+) with -30% ROI and almost $2 million lost over 1k entries.
Still, Roman does not doubt his skill:
I'm currently the biggest player at mid/high stakes and the best among those who play at reasonable distances. My belief in the multi-million dollar expectations I was working with hasn't changed, and the actual calculation is a simple 50k+ yearly total, ABI of 400-500, and ROI of 10-20%.
I'm confident I'll be able to navigate this and pay off my debts at a reasonable interest rate, but I understand my reputation can't be repaired. What can I do, I'll do everything I can to minimize the damage. My recent actions were related to my inability to provide for my family's basic needs.
Romeo showed the list of people he owes money to, in an anonymous form:
Mr. A: 1.7
Mr. B: 0.41
Mr. B: 0.3
Mr. G: 0.3
Mr. D: 0.23
Mr. E: 0.2
Mr. G: 0.106
Mr. Z: 0.103
Mr. E: 0.09
Mr. Y: 0.04
Mr. K: 0.033
Mr. L: 0.02
Roman has very ambitious plans:
– As for the marathon, by the end of the year, the plan is to play ~220 days out of 297, with an average of ~200 tournaments, in order to reach the goal of 50k+ tournaments played.
The short-term plan is to build up my bankroll to $50k by the start of GGPoker World Festival, win the Middle and High leaderboards this year, and take the Super next year.
Due to the author's fame and the record-breaking amount of debt, the thread almost immediately became one of the most popular in the Russian GipsyTeam forum's history. In two weeks, it has over a million views, even among plenty of high-profile stories in the Debt and Disputes section.
Romeo is successfully achieving his long-term goal of 200 tournaments per day. He has played 2,376 tournaments over 12 sessions, with a result of -2,496. His current bankroll is $7,504, down from $2,876 at his lowest point.
For some tournaments, Roman began selling shares in the GGPoker client.
The need to sell shares was suggested by a mathematical model :
"For the past three days, I've been talking to the math/BRM guy, ever since BitB, where he's been persistently trying to convince me to sell my MTT shares. He also pointed out that my strategy of grinding cheap MTTs is far from optimal, showing various calculations and demonstrating the insane variance and its impact on long-term bankroll growth due to drawdowns and rollbacks, and the endless time needed to realize EV in mega-fields. Yesterday, he even used AI to write an essay/article in his substack about our discussions and the fundamental mistakes that almost no MTT player understands (some people simply act better intuitively, while others have a worse intuitive approach to BRM – I definitely fell into the latter category in the past).
Ultimately, he managed to convince me otherwise – especially given the urgency of boosting my BR from 7.5k to 50k by the start of GGWF, where I'll take both mid-table leaderboards (unless I get seriously ill again like last year), his task became much simpler. And with a BR of 200k+ for 1k-level tournaments (except for 500-1k mega-fields), selling shares at a loss will disappear due to the sufficient bankroll depth. In the current situation (a screenshot of one of the model calculations will also be below), playing in mega-fields with <100/200 BI for building a long-term bankroll, for example, is strictly unprofitable in principle.
In short, and greatly simplifying, the bottom line is that his model suggests selling shares even at half their ROI due to the insane variance in tournaments that don't have at least 300 buy-ins in short fields and 600 buy-ins in kilofields. I won't be selling kilofields for $22-33, however, as I'll offset them with small fields at PS for $5-55 with low variance. My preliminary plan is to sell shares worth $44-440 on Wednesday at GG (when I finish resolving "poker-related issues"), and then sell them for $500-1k+ when the BR reaches at least 20k+. I'm still considering shares for PS/VPN/Coin tournaments, as the process there is much less smooth, and I've always considered time to be the most valuable resource, and I still do. Felix is also in the process of finalizing the calculations on the comparative growth rate of BRM with and without sales, and I'll send you the results in a few days."
Forum users were skeptical of the calculations.
Roman's ability to play such volumes in the black raises serious doubts among many.
A forum user posted a chart of Romeopro chips with EV win rates over a 200,000-hand period.
– The general conclusion that emerges from the overall graph is that Hero isn't beating the game or is breaking even in EV. Yes, you might say that 2 EVbb is a positive win rate, but suddenly there's rake in tournaments. To beat him, you need to have an edge. By the way, the top win rate is currently considered to be 12-15 EV bb/100 and higher, depending on the ABI.
Some rather interesting conclusions follow from all of this:
By any sample, the EV there should also be low. Romeopro couldn't have been unaware of its EV. We're not going to seriously consider the fund owner's inability to upload hands to a tracker and open a chart tab, are we? And he once wrote about win rates on GipsyTeam many years ago... Which means:
1. Hero was deceiving backers, misleading them about how great a player he was (I wouldn't be surprised if he deleted hands from the database, for example, but this is purely speculation).
2. Hero deceived creditors by misleading them about what a cool player he was.
3. Hero deceived his students at the foundation.
4. Hero was deceiving himself.
5. Hero lied to you on this topic.
6. Creditors and shareholders also, I don’t know how much money they gave without looking at the figures, and in general, how could you not look at who you’re backing?
A few days later, Romeopro denied the allegations:
Guys, maintain at least some minimal level of common sense. I understand this topic doesn't exactly encourage it, but still.
I don't read it myself, but I hear some absurd stories from various corners of the internet, like Romeo playing at -10EVbb/100 or 250k hands with a win rate of 1 (I don't know how that could have happened, maybe it's just bad mining – in my 15+ year career over 250k hands, I've only seen a VR of 1 or less maybe once). In the last few years I've been coaching, I can't recall seeing a single 250k+ hand analysis where my win rate wasn't in the 6-11bb/100 range. The adaptation period in 2025, due to the strategy restructuring to increase the overall annual expectation, was quite difficult; there were a lot of sitouts and the actual win rate was shown to be ~2EVbb higher (between 8 and 9bb – if we removed folds to 5-bet pushes with KK-AA, folds with nuts to all-ins in us on the river, etc.).
Winrates for 28 tables (I downloaded the database from PS and GG – I haven't played much on the other rooms yet – it's funny how the shortfall turned out to be so huge (more missing than gained :D) and for both overall and late-game winrates – that is, I knew the early game was very bad, but I didn't realize it was that bad) overall and in the mid-late game:
Win rates for the transition period from a potential expectation of $1-2kk to an expectation of $4-5kk (as I saw it and as I see it and as you, I hope, will see it in the coming years):
One user doubted the veracity of these graphs and posted a selection of Romeopro hands with questionable preflop opens:
I haven't played poker on 20 servers, so I can't judge anything. I'm just stating that I've been watching most of what's been posted on YouTube for the last 5 years, and I haven't seen anything even close to a similar preflop from any other high-stakes reg.
Three people from the list of "Mr." revealed their names.
Voffkin4 was the first to respond and offered to buy out part of the debt at a discount:
– If anyone's interested in supporting TS, I can sell a portion of the debt, starting at ~40k, at a 5% discount (10% discount on 10k). We've agreed on 1% per week, so the debt will grow.
After him, Gnat admitted that he was hiding under the letter G (debt 106k) and suggested asking questions.
He explained how such a debt arose:
The money was borrowed to cover liquidity. Roman told me about a stable where they all piled up his money, and I figured that was theoretically possible, plus his impeccable reputation. My brain still refuses to believe it. He borrowed a hundred thousand rubles in April 2025 (about $1,200) and paid back 105,000 rubles (about $1,280) two months later, as agreed. Although I asked if he had any other debts, he said 400,000 rubles (about $4,900).
On Roman's spending and lifestyle:
"I don't see the point in discussing a debt repayment plan. Roman is struggling to survive right now, not to repay his debts. He still has expenses for a family of five, and I think it's at least 15,000 a month. And to do that, he only needs 180,000 a year in expenses. Meanwhile, he still has to ramp up his BR and not go crazy with the amount of work at 35."
"I'm still smiling about this situation because I can't get over how ridiculous it is. 😂 I call Roman and ask about the ethics of certain online statements. What can and can't be advertised. Roman calls me, borrows money to provide his staff with good living conditions. Yes, Roman rented a separate apartment for his nannies/housekeepers.
"He was spending 60-80k a month, the current budget is already cut, and further cuts will take time. So, the bulk of it was spent at a rate of 700-800k a year. My ChatGPT can't even imagine spending that much😁. And I live in Miami."
"As far as I know, there aren't many assets there. Literally up to a hundred or two. My wife didn't really dabble in jewelry, and Roman didn't buy any expensive watches either. They lived on the bare minimum. So yeah, hedging their bets by selling the apartment or bitcoins isn't an option."
Other people Roman owes looked for debt buyers:
Hello everyone! We didn't manage to take first place in Roma's Mister Leaderboard, but we're far from the bottom of the standings...
If anyone's interested, I'd be interested in selling Roma's debt at a significant discount (from $5,000-$10,000). I check in here once every three years. You can find my contact information through mutual friends or by direct message at t.me/svinopigvladimir.
But some condemned the debt sale:
You and everyone else trying to sell Roman's debt are essentially even bigger scammers than Roman, because you know the real state of affairs, and some of you are HS MTT players who have known for 4 years that Roman is playing completely wrong. You know globally that the debt is worth $0, but you're trying to sell this nonsense to hamsters.
The debt seller disagreed:
I've been fortunate to discuss poker with a variety of people over the past few years, including those I'd rank in my top 5 MTT players right now. In terms of his understanding of poker and his ability to break down hands, come up with lines and adjustments, Rom has always been top 1 for me by a wide margin. I believe his coaching has made a HUGE contribution to my rise to high stakes. Yes, he overdoes it with exploits and plays some really bad hands. Perhaps he needs to retool a couple of his strategies. But the HS regs you mentioned always ranked him in the top 3 before COVID (before he retired from poker). And even when he first returned, he played quite well. Personally, he's always been a tough opponent for me.
With the right approach, he has a chance to turn things around. But it would have to be a long marathon, not the sprint Roman is currently attempting. I could spend a lot of time describing what I would do, but I doubt anyone would listen.
The topic has awakened many forum legends from their dormancy.
In his characteristically flamboyant style, a 16-year forum user reminded young readers that a dollar an hour isn't everything:
"And when they tell you that time is the most important thing, don't listen. There's no point in saving what you don't have. Every person (even a poker player) has only one truly important resource—their head. Or their psyche, if you prefer academic jargon."
If you're playing NL500 (or NL100) and feeling great, but you're frustrated because you "should" be playing higher for higher expectations, while experiencing extreme stress—don't do it. So many regs have burned out trying to outperform. And what's more, it turns out even those who managed to outperform the stars burn out.
Value stress-free living more than the hourly rate. You can't earn all the money you've earned. And even those who have earned it all are often unhappy.
Stefan11222 didn't like some of the words in this post:
"Uncle, your talk about 'valuing your head and avoiding stress' sounds like a clever attempt to justify your lifelong paralysis. It's probably convenient? You take a personal fear of life, juggle terms and concepts, and present it as wisdom for the younger generation. You probably don't understand how ridiculous this looks from the outside, as ridiculous as a parking lot guard lecturing a pilot who crashed a plane.
Feruell also disagreed with the theses voiced:
"A head isn't a resource, but a tool for increasing the only thing that matters to a poker player—your dollar/hour. There's no point in saving your head if your dollar/hour is low, because if your head had real value, your dollar/hour wouldn't be low."
