Tuesday, November 10, 2020

Real Time Assistance (RTA) - A confession

Real time assistance is the current hot topic in poker. I’ve debated whether to write this blog or not but in the end decided it’s time to finally come clean. In this blog I’ll explain explain what RTA is and why it’s cheating, but first let me rewind almost 50 years to my schooldays.

Back then, computers weren’t a thing. Calculators were, but as is often the case when it comes to education, the educational system lagged technology with teachers continuing to teach certain skills well past the point they were rendered obsolescent. This is why young school kids were still taught multiplication tables. They came in a small self contained book, each page a different number between 2 and 10, with a list of the ten digits multiplied by that number. 2 times 1 is 2, 2 times 2 is 4, 3 times 2 is 6....you get the idea. 

We were expected to learn these by heart at home, to be quizzed on them in class. We got the book at the start of the year, and before we were ever assigned a table as homework, I had devoured it cover to cover. Numbers fascinated me from an early age. I was not interested in learning a few dozen calculations by rote: what interested me were patterns. Doubling was so easy I had no need to learn the 2 times table as I knew it already. That meant I didn’t need to learn the 4 times table either (just double twice) or the 8 times table (three doubles). The 3 times table seemed trivial too, being just a double and an addition. The 6 times table was just that doubled. Ten was the simplest of all, just add a zero. 5 times table was just 10 times halved. The first method I used for the 9 times table was to add a zero then subtract the number, but then I noticed another pattern: when you multiply a number by 9, the digits of the result always sum to 9, and the first digit is just one less. That left just the 7 times table to be learned, but then I realised since I knew all the others, and 8 times 7 is the same as 7 times 8, the only one I actually needed to learn by heart was 7 times 7.

Every Friday, our teacher ran a contest where we all stood up, he presented us with a random multiplication, and the first to raise their hand and answer won a chocolate bar. It turned out my pattern method was faster than the memory retrieval method all my classmates were using, so for ten or eleven weeks I went home every Friday with a chocolate bar in my bag. This did nothing for my popularity, and even the teacher started getting frustrated that the same kid kept winning. He took the reasonable decision of pretending not to see my hand when it shot into the air. It took me a few weeks to realize I’d effectively been banned for being too good in the interests of class morale. Once I realized this, I came up with a new strategy. I enlisted the help of the kids around me. The plan was simple, as soon as the question was posed, their arms would shoot into the air, and I would signal the answer with the digits on my unraised hands. Teacher would pick one of the kids, and we’d split the chocolate bar later. 

At the time I saw nothing wrong with what I was doing. Looking back I realize it was cheating. I effectively provided real time assistance to other kids allowing them to win, and benefited from so doing. I also realize I’m disappointed that 8 year old Doke wasn’t sufficiently au fait with game theory to have worked out that always being the first with the answer wasn’t GTO. Had I held back some percentage of time and let another kid win, my long term expectation could have remained above the half a chocolate bar it became after I was effectively banned for winning too much.

Flashing forward to GG poker

GG are very much the new kid on the online poker block. They have a bad boy swagger to them that suggests they might not fully play by the rules, and deep down what poker player doesn’t appreciate a little rule bending (especially dumb ones like you can’t play online if you live in a repressive country)? They recently sent Pads (Patrick Leonard) into a tizzy when they responded to some criticisms he leveled at them with the tactic you often see used in the schoolyard (“No, you’re stupid!”). It also needs to be said they’re doing a lot of things very well, in particular one the poker industry has struggled with in recent years: player acquisition. Recreationals love the software, and many of the features. The staking feature is particularly genius. It’s been an unspoken (although Lappin broke ranks and spoke it on his recent appearance on The Orbit) dirty industry secret for years that while Twitch is very good at attracting people who like to watch (poker or any other game), this doesn’t really translate to new player acquisition. The vast majority of watchers stay that way, and don’t contribute anything to the online ecosystem. That’s fine for other games that monetize their audience, and it’s even fine for poker twitchers making their living from donations and subs, but doesn’t translate into bums on seats or online sites.

GG have found a way to monetize people who like to watch rather than play. They can now buy a piece of any player (who is offering shares through the client) at the click of a button, without having to even contact the player in question. The client automates the entire process: there’s no having to chase players if they cash. As ever, there are rumblings that people are charging too much markup, but this is entirely missing the point. Traditionally, players have sold to other players, and there’s been a sort of gentleman’s agreement that the markup will effectively share perceived edge between the player and the stakers. Whenever I’ve sold in the past, I’ve estimated my edge and translated it to an expected ROI, then halved it, so if I think my long term ROI in a particular tournament I’m playing is, say, 60%, then I sell at 1.3. 

Of course nobody ever knows their exact edge, and most of us are possibly guilty of over rather than underestimating, which tended to lead to one of the most tedious debates in poker between self appointed markup police (usually people with no actual interest in buying action) and sellers. In recent times we have seen much less of this as most people are happy to let the market decide. I think there’s also a growing acceptance that there’s an entertainment aspect to this, particularly with the Twitchers. Most people make minus Ev sports bets to make their viewing of sports more enjoyable. Similarly, rail birds will happily buy a % of their favourite Twitcher to make sweating them more fun. They’re not too bothered whether the long term Ev of that percent that costs $4 is $4.24 or $3.76. Conversely if the Twitcher can quickly sell out at $4 why should they sell lower?

This is great for the overall ecosystem as it brings fresh money into it and allows streamers to play higher (which is inherently more entertaining for their audience) than on their own dime. The idea is such a winner, I’m surprised other sites haven’t scrambled to copy it as fast as they copied PKOs from Winamax. 

Other GG innovations haven’t been met with universal approval. RTA (real time assistance) is the current hot topic in poker. It’s obviously not just a problem for GG, but the exposure of Fedor Kruse as using it to rise the stakes rapidly there moved it centre stage. As a former runner I have a particular reason to feel that it’s a little unfair to point fingers exclusively at GG because of Kruse and some other players there who have had their accounts suspended over allegations of RTA use. The fact that people are being caught (assuming they are guilty) is a good thing, not a bad thing. Running is one of the best if not the best sport at catching and punishing drug cheats: yet the public perception of this often translates to “they’re all juicing”. Meanwhile other sports which take a much more low-key approach to testing sail under the radar successfully presenting the image that they don’t have a big “drug problem”. We may have a similar problem in poker. The sites making the biggest efforts to catch cheats are the big regulated sites because they correctly recognize cheats as an existential threat to their business. That means they will catch more cheats than smaller unregulated sites or apps, but if the message recreationals take from this is they are safer on an app, they are very much mistaken. 

What is RTA?

The most hard line definition of RTA is it’s anything outside your own brain or memory that helps you make a poker decision in game. This would mean preflop charts like the ones in “PKO Poker Strategy” or “Poker Satellite Strategy” or apps like Snapshove. In theory, even sticking a post-it on your computer saying “Don’t call threebets with AJo” would be crossing the line. 

Most people and some sites don’t take this hard a line. Stars explicitly allow you preflop charts if it’s folded around to you. Most people would accept this is not cheating: the question is how much above and beyond that you can go. If a chart is ok, what about a spreadsheet where you type in stacks and blinds and it works out which chart you should use? Most people think we are getting very close to cheating now, and most would agree we cross the line for sure if we write a programme to scrape the screen for this information, retrieve the correct chart and tell us exactly what to do. 

Postflop is even more problematic. You might argue that it’s ok if you just happened to run a sim looking at AK4 flop to look back at the results if it arises in game, but it’s becoming easier every day to build (or even buy) libraries of solves. The real doomsday scenario is full blown RTA software that scrapes the screen for all the relevant data, and then consults a comprehensive list of charts or solves to tell you exactly what to do in every spot. And if you think that’s far fetched and futuristic I have bad news for you: the technology is not only here but has already been demonstrated on Twitch.

The solution 

So what can we do, Daddy?

The first thing to say here is the sites haven’t given up. Nor are they powerless to catch people using RTA. The easiest way to start protecting yourself from cheats is by sticking to sites that have displayed a will and ability to catch them. I regularly get small refunds from Stars and GG with a note saying they caught some cheater I played against, confiscates his funds and are redistributing to those of us who were directly affected by the cheating.

The next thing to think about is that different forms of the game are less susceptible to RTA cheats. Cash with its fixed stack sizes, predictable parameters and pure chipEv is the easiest to cheat. Spin and gos, headsup sit n goes, and any other winner takes all format are similar, given their lack of ICM. Tournaments introduce ICM and another level of strategy that RTA tools will struggle to adjust for accurately if at all. Best of all are PKOs which introduce another level of strategy in the form of bounties, which will probably never be fully solved for. 

For my own part, I’ve shifted the vast majority of my online volume to PKOs on legit fully regulated sites. PKOs actually address not just RTAs but a lot of the downsides of the online poker experience. They disincentivize stalling, they encourage looser more fun play, they create more multi-way pots which can’t be solved in PIO, and they make the action meaningful right from the start. And I’m not just saying that because I just wrote the book on them.



Twitter Delicious Facebook Digg Stumbleupon Favorites More