The reason why I posted all the relevant judgments is because it's clear people posting ITT need to do some homework.

Received my 22 thousand dollar refund a month ago. Used a lawyer to right a letter citing the relevant case history, and how it applies to me specifically.

The letter cost ~2.2k, so I was getting 10:1, and it looked like a good bet

Would you mind elaborating on this?

Do you like apples? Very good news in the latest CRA interpretation folio for poker gamblers with personal winnings:

http://www.cra-arc.gc.ca/tx/tchncl/n...f9-c1-eng.html

I've read it twice. I don't see anything that is new or changes anything from where we were before this interpretation bulletin.

If anything I see this as a negative for poker players. No news I would interpret as a lack of interest. Issuing a restatement of the current situation signals consideration and interest.

http://taxinterpretations.com/?p=32348

I don't understand it either but I'm probably not meant to; I'm just a grinder:P

How the hell is this good? Sounds horrible to me.

This was already established so it is not new. I actually see this as a negative.

This is the argument CRA has used in the past and they have lost with this argument over and over again. Nobody expects that CRA will ever be successful with this argument. The best scenario would have been for CRA to continue to advance a losing argument and increase the series of unsuccessful trials.

That CRA has come to terms with the fact that this argument is a loser means they will not do that. To see this as a positive one has to assume that other alternative and better arguments do not exist -- which is incorrect. The Court has made it very clear that there is a path to a winning argument it just involves a lot of work. If we were to use weight loss as an analogy CRA has been wasting their time on magic pills and they have finally accepted that magic pills will not work. They still know that good diet and physical exercise will work but that is hard. This is only good news for poker players if we make the assumption that upon this realization CRA will be too lazy to do what is needed. That may be the case but it is just as likely that they will now do what they should have done in the first place. For poker players the best outcome is to have CRA remain delusional and continue to repeat their previous mistakes and that is now something they have announced they will not do.

Between this and the announcement that apparently players who were paying before are now requesting and getting reassessments I am much less optimistic that this is going to remain dormant.

I honestly struggle to imagine the government convincing a judge that personal gambling winnings are includible in income given all the case law. Judges don't want to go out on a limb and try to reinvent / run counter to an entire corpus of case law. Even if the CRA and the DOJ get brave and try to fight yet another case before the TCC, I expect that the government will lose.

The reason is this: the government would need to be able to show that someone is a professional irrespective of their results, their intensity of play, and their intention to profit.

What is left? Showing that the player has the right state of mind to be sufficiently profitable to overcome the rake. At best, it will be a battle of experts.

Given the ambiguity and the fact that the case law would have to be countered, a judge of the TCC would be hard-pressed to reach the conclusion that any given poker player is taxable.

I suppose the best case for the CRA would be finding a player who is known as a pro, holds himself out as a pro, gives lessons, attracts appearance fees, etc.

What is left is how they play.

An easy example would be a short-stacker. It is very easy to demonstrate that they have a specific way of playing that they employ.

A second example we could use Leblanc. CRA lost because they did not understand their system. It would have been very easy to win had they explained to a judge that Terry and Brian were exploiting the static lines. Basically that they would wait for something to happen that had a significant impact on the outcome of a game and that they would then round robin that +EV game with a series of -120 near locks.

Both of these are easy examples. With a normal poker player it would be more difficult but not that difficult -- you just need the documentation of the player's behaviour while playing, documentation of the behaviour of a sample of recreational players, and someone who knows stats.

---------

With respect to going against case law it wouldn't be. The case law is very clear that if the poker player employs a system then he will be found to be a professional player. CRA has lost every case because they are using the outcome of employing the system as the only evidence that the system exists. The Court has been very clear in saying that is not enough but at the same time in every decision they also say that if you can show me the system then that will be enough. If CRA does that then finding a player is taxable would not run counter to the case law -- CRA would just finally be doing what the Court has told them they had to do.

If I was going to advance a poker case I would look for a short stacker (assuming that is still something people do). It is by far the easiest system of play to document and to explain to a judge.

An alternative option would be someone who only plays tournaments with overlays. There is also probably some pretty simple methods of play that could be documented for 50/50 tournaments. Another possibility for a simpler situation might even be rakeback grinders but I don't know enough about how they play and how their play is different other types of players.

I have no doubt it can be done for regular ring players as well but in situations like that the complexity is greater. I'd want a simpler situation just to break the streak.

Don't forget that all of this supposed +EV behaviour needs to overcome the structure of the game. Showing an edge is not enough. It has to be a sufficient edge, and has to be sufficiently organized and commercial. I see it most clearly with Leblanc, and we all know how that case turned out. I don't think any poker player is anywhere near as organized and systematized as the Leblancs were, not least because the structure of poker is much more difficult than static sports lotteries back in the day.

Leblanc is the correct decision based on the facts presented in court but the facts presented and accepted by the court were wrong -- they were not an accurate representation of what Terry and Brian were doing to beat Pro-line. Had the court had an accurate account of the facts CRA would have been successful. This was a really mismanaged trial so much so that CRA allowed incorrect expert testimony to go unchallenged. That expert testimony was the deciding factor in the Court's decision.

When it comes to poker it is just behaviour. All behaviour that is informed by a strategy can be discovered through statistical analysis. I don't think it is easy but much more complex versions of this are being used for modeling shopping behaviour, predicting criminal behaviour, and anti-terrorism.

No, it's quite the contrary. Most recreational short stackers that "play well" will never be good enough to show profit. Even ones with very expensive and well tested "systems" or "strategies" will be unlikely to show profit, let alone enough to be a professional.

Your logic suggests that nearly every poker player is a professional, as most recreational players have some sort of system. Surely you don't believe that this is the case?

Henry, you sound much like the typical recreational player. You make it sound like you think poker is easy and simple to win at, and that it doesn't take much to cross the threshold from a "hopeful recreational player" to a "full time pro".

The things you mention to be simple and easy to show are extraordinarily complex.

I would expect that the difficulty in predicting with accuracy whether or not someone was a winning player through statistical analysis would be at least an order of magnitude greater than the difficulty of writing a winning poker bot for the stakes in question.

At the very least, doing such analysis would not be valid if those doing the analysis weren't able to beat the games that they were analyzing.

My understanding is that there was short stackers making steady money -- I remember one of them had a cartoon train as a icon on 2p2.

Anyone who puts thought into any activity has a system. There is nothing special about poker in this regard.

You are correct that both recreational and professional players have systems but that is irrelevant to this discussion. For CRA to be successful they do not need to demonstrate that recreational players do not employ strategy in playing poker. All that is required is that CRA demonstrate that the player they are pursuing employed a system.

That you are a successful poker player does increase the argument that poker can't be that hard.

Actually it isn't. The problem is you don't understand statistics so you think it is.

Not even close. A bot involves apply probability to solve a problem of imperfect information. Looking for a system in a player's behaviour is just simple retroactive pattern recognition. A much better comparison would be to compare what CRA has to do to what someone would do in auditing a site for fairness using hand histories. CRA's task is slightly harder because they don't know where the road ends before starting but except for that the procedure is pretty much the same.

lol again no. You constantly jump into this conversation despite lacking the understanding of what a system is. I've tried to explain what a system is on multiple occasions and you are way too dense to understand. A system is not a winning system in the way Hi/Lo is for blackjack. You keep returning to that meaning of the word "system" despite having been corrected multiple times. All a system means in this context is organized meaningful behaviour. There is absolutely nothing difficult about apply statistical analysis to this and you don't need to be good at poker to do that.

I also get the strong impression that you think that this topic will somehow impact the eventual determination of whether poker winnings are taxable or not -- it won't. The only way this topic will have any impact on that is if you piss me off enough that I volunteer to consult for CRA. You haven't studied tax law, you haven't read the case law, and you wouldn't understand it if you did. You have nothing to contribute here.

The fact that there are short stackers making money doesn't mean that it would be easy to understand why most aren't successful with that strategy.

Now this is what I don't get. When I asked about this before, I think you said something about there needing to be a reasonable expectation of profit.

Are you saying that a player's system doesn't need to have a reasonable expectation of profit for taxation?

Sigh. Ok, we got off on the wrong foot. Truce?

I would guess that the difference in how we view this is that I believe that the system needs to have a reasonable expectation of profit, but that you don't?

Again, I think the misunderstanding is that I believe that the analysis needs to show a valid/working/profitable system, while you seem to be saying that any ol' system will do.

Ahh, ok. Before you had said that I should consider system and strategy to be synonymous in this context.

"Organized meaningful behaviour" sounds weird here.

What I mean by that is that couldn't anyone trying to play well at poker be shown to have organized meaningful behaviour?

I'm missing a distinction in what makes a "recreational player", or more specifically a player not engaged in business different from one who is.

You keep bringing up the idea of "all that needs to be shown is that the player has a system". The reason this seems deficient to me is that I don't see this argument making any distinction between a business-like taxable player and a non-business like non-taxable player. Unless your argument is that virtually all poker players fall into the taxation category?

Yes, you understand law better than me. And I understand poker better than you.

Since we both have understandings of things that the other is deficient in, this can be a mutually beneficial conversation.

The CRA has decided in my case that I'm not taxable, so this is really just a topic of interest for me at this point.

You don't need to know anything about poker. If I was looking for assistance on a project like this I'd take a Palantir employee who only has the basic rules of poker on a piece of paper and that is the first time he has heard of hold'em over the help of the top 10 poker players in the world. This is a data problem not a poker problem

There is absolutely no doubt that it is possible to meet the requirements the Court has set out for poker winnings to be taxable for online players with the cooperation of poker sites supplying data.

There is a lot of doubt over CRA managing this but that is because they are lazy and stupid which is different than it not being possible. If Canada's tax system was administered by the IRS I'd expect poker players would be ****ed but we prioritize being a francophone over competency so I wouldn't worry much.

The structure of poker as a game is that it is symmetrical and -EV for the players. A successful demonstration of a system would require a demonstration, without reference to a player's results, that that player is +EV when he or she next sits down at a table to play.

The demonstration must be prospective, not based on results (only strategy, behaviour, approach, etc), and sufficient to convince a judge to give up a reasonably good bright line rule (no deduction of losses, no income from personal gambling) and substitute some judicially legislated criteria for sniffing out +EV players.

The two biggest hurdles are (1) not being able to make reference to results (making much of the statistical analysis being proposed impossible); and (2) the deep and symmetrical -EV nature of the game (why is this player +EV and everyone else even more -EV than you would otherwise think?).

Judges like to decide cases in a way that is replicable, not going to lead to unexpected consequences, and that does not to reversal by higher courts. Remember that we have a system of taxation based on self-assessment. The easiest path for a judge here is to find that personal gambling winnings and losses are beyond the reach of the income tax. That's where I would place my chips.

Agreed. That would not be difficult to do for a a online player as long as we assume CRA has the cooperation of the poker site.

The bolded part is not correct.

In Epel the law is clear. Poker winnings are taxable income if you are like John Turmel. Poker winnings are not considered income from a taxable source if you are like Tony Epel. The law is 100% clear. The only issue is a question of fact over if a particular player is like Turmel or like Epel.

Because Epel was a B&M situation the Court relied exclusively on the opinion of Turmel who said that Epel was not a professional player. Had Turmel said the opposite then Epel would have resulted in a successful outcome for CRA. It is very clear in the reasoning that the judge abdicated the decision of Epel status of recreational player or professional player to the subjective assessment of Turmel (a mentally unstable loon that somehow was accepted as an expert witness). The outcome could have easily gone the other way if Turmel had just had a different opinion of Epel's play. It could have gone a different way if CRA had discovered their own experts who were also witnesses to Epel's play and offered more convincing evidence than Turmel.

There is no law established since Epel being a recreational or professional player is a question of fact. That the Court repeats multiple times in the reasoning that if someone is like Turmel then their poker income is taxable removes any doubt about the willingness of the Court to return a favourable outcome for CRA. All they have to do is show that a player is like Turmel. For B&M players that would be difficult (almost impossible) but for online players where every action is recorded and available as data that can be easily manipulated it is not difficult.

In LeBlanc the Court again addresses the law -- if LeBlanc was a poker player his winnings would be taxable is stated twice in that reasoning. The Court relies on the expert testimony of LeBlanc's to distinguish their behaviour from a poker players behaviour which would be taxable income. Again this does not establish law -- it is a question of fact that the Court got wrong because they accepted the factually incorrect testimony of an expert witness that CRA did not challenge or attempt to refute.

This is not difficult. Poker software companies create very basic software to help people do what CRA needs to establish. The software that is sold is basic because it needs to apply to a general player base and because it is designed to help people play. It is infinitely easier to go from data and work backwards and it is infinitely easier to find evidence when you are dealing with only one player.

When it comes to questions of law certainly but absolutely not when it comes to questions of fact. If we had replicable outcomes on complicated questions fact we wouldn't need trials or judges.

Except they have never done that. In every single decision the Court has taken the time to include in the reasoning that gambling winnings can be considered income from a taxable source but that in this particular situation they are not.

I think your opinion is way too strongly influenced by the outcomes of previous cases and you're ignoring the fact that has CRA completely mishandled those trials. In almost all of them CRA would have been successful if they weren't so incompetent. This is especially true in LeBlanc. If someone who understood sports betting or even just probability was involved CRA would have been successful. LeBlanc was the perfect case. It was a slam dunk and CRA messed it up.

That said we can't really rely on CRA mistakes to influence our opinion of how a court will decide something -- if the Crown incompetently fails to introduce the breathalyzer/toxicology report and loses a half dozen DUI trials we don't conclude that drinking and driving is legal but that is exactly what is happening here.

I see what you're saying. You're saying, if only the expert evidence had been produced and shared with the court in the right way, then based on what the courts say, the outcome would have been different. From my perspective, you may be placing excessive weight on obiter dicta. Judges can say all they want about "if the facts were different ... etc, etc, etc" but when the rubber hits the road, I'll take what the judges do over what they say. (A good example of this is Morden, where the court basically said that if Morden kept doing the same thing in future tax years he may be taxable.)

Leblanc was the perfect case and Chief Justice Bowman is no dummy. He reached the conclusion that provided a reasonably good bright line rule and is not going to lead to all hell breaking loose in a self-assessment taxation regime. I'm confident that other judges will do the same and follow the CJ of the Tax Court of Canada.

It is not a matter of intelligence -- it is a matter of the evidence introduced in court. The decision in LeBalanc is correct based on the facts that Court heard but those facts are not reality.

I'm not sure how much you know about sports betting but in this particular case LeBlanc had an expert witness Gary Smith and Smith's testimony was mathematically incorrect. CRA did not put up any experts to refute Smith's testimony so Justice Bowman accepted it and his decision is completely determined by what Smith testified -- but everything that Smith testified to was wrong. This isn't a matter of opinion we are talking Smith's testimony is mathematically incorrect. The judge isn't going to know that.

The reason we have expert witnesses is to inform the court on issues that are beyond the scope of what a trier of fact can reasonably be expected to know. The system works because both sides can consult their own experts and if someone tries to bull**** the court it is easy enough to refute the expert with your own expert -- CRA never did that. There is a Boston murder case where a professor killed his wife -- to refute the DNA evidence he hired an expert witness that testified the DNA could be by tertiary transfer. That is impossible but he still found an expert out of California that would testify to it. If the prosecution didn't introduce their own DNA expert to refute that claim but left it unchallenged the jury would really have no choice but to expect the defence expert witness.

So this is the first reason LeBlanc is the correct decision but the wrong outcome. CRA allowed an expert witness to lie to the court.

The second reason LeBlanc has the incorrect outcome is that CRA in making their argument for why LeBlanc should be considered as engaging in a business never once mentioned their system. The LeBlancs had a very simple system -- it was easy to detect, it was easy to document, it would have been easy for anyone who is even mildly articulate to explain how it worked in a way that a non-sports better would understand. CRA did none of this. I mean WTF? They spent a lot of time introducing mostly irrelevant evidence about the amount of organization in LeBlanc's operation with respect to how the physical purchasing of the lottery tickets happened but not once did they actually mention how the LeBlanc brothers were able to beat Pro-line.

Again this is not the Judge's fault. If CRA doesn't introduce the most important element of the case as evidence in court what is a judge to do? What CRA did here is the equivalent of prosecuting a murder trial where you have video of the accused committing the murder but the prosecutor never introduces or even mentions that in court and instead tries to get a conviction using circumstantial evidence.

OLG knew what was up. By the time LeBlanc was on trial the rules to Pro-line had been changed so that you couldn't use their system. That actually would be the most charitable explanation for CRA missing it and Smith giving false testimony. If you looked at it using the new rules when the trial was happening then Smith's testimony is factually accurate and how the LeBlancs were able to win would not be obvious to CRA. That is obviously incorrect since what matters are the rules when the wagering was occurring but of course none of these are facts the Court is aware of. That OLG changed their rules to prevent this method of wagering wasn't even introduced as evidence.

If I had handled this case I would have with 100% certainty gotten a favourable outcome for CRA. All that was needed was;

1) An expert witness to refute Prof. Smith's as well as asking Smith the correct questions about his methodology to destroy his credibility.

2) Actually introducing as evidence the method by which the LeBlancs picked their parlays. This would be easy to do -- all you need is access to the OLG, historical lines for a major sportsbook, and to make it easier to communicate to non-gamblers archives of the sports section.

This was a slam dunk case. The only reason the LeBlancs avoided having to pay income tax was CRA's incompetence.

Now that we're all on the same page, and understand that the CRA would be required to show that a given strategy is +EV, I wanted to go a little deeper into what that would look like.

The requirements of this software are to accurately analyze the EV of a given player over a given timeframe without referencing results in any way.

First, the software would need to import all of the hand histories from a given player into some sort of internally readable format. At the bare minimum, this should be designed to handle 2 million hands per year.

It would be required to be able to import every hand from every game that the poker site offers.

This first part of the problem is deceptively difficult, though certainly doable with a sizable budget. Existing poker software shows that this is a viable approach, but they are all lacking in the variety of poker games that they can handle.

There’s already a fundamental flaw in this approach, which is that poker sites don’t record all information that a poker player has available to them when making a decision. The biggest important piece of information that is missing here is the amount of time someone takes to make their decision. This is critical. For now, I’ll overlook this part, though in a production version it would be something that needs to be resolved.

Ok, so we’ve imported all of a player’s hands into our database. Next up, we need to analyze them. Easy way would be to look at monetary results, but we can’t do that. What else is there?

We must look at every situation that the player finds themself in, and assign a value to it irrespective of results. The values assigned must be accurate enough to separate players with a positive EV from a negative one.

At first, this would be relatively straightforward, at least in terms of separating good decisions from bad ones. Raise AA? Good. Fold AA? Bad. Ok, but that doesn’t really help us, as all poker players are doing that. The important distinctions are the ones that separate the winning players from the losing ones.

In the past, this may have been a doable, or even simple task. In the “old days”, it would not be unreasonable to sit at a poker table and just know who the good players were relatively quickly.

This is much more difficult to asses with accuracy. Good players have incentive to look as “bad” as possible to other players. Bad players also have incentive to misrepresent their skill levels at the poker table.

The most important part of the software analysis is that of understanding what makes one situation different from another.

Raising AA has value x. Add value y because of position. Add and subtract other values based on anything and everything that could be important to a poker player when making a decision. Other players at the table. Time of day. Playing history. Timing. Flights of whimsy. Or, rather, “I’m supposed to deviate from the norm a certain percentage of the time, and I’m choosing this time to do so.”

Very quickly, it will become an astronomical feat to accurately assess the situations the poker players find themselves in. The values that you start assigning to various players will quickly be stabs in the dark.

Players? Why players? Weren’t we only looking at one player?

Yeah, so there’s a problem with looking at just one player. Poker is a game that is inherently –EV, which means that we cannot assign values to a player’s actions that push him into the realm of +EV without comparing him to the other players in the game. He can only be a winner if there is enough loss to compensate for the rake.

This is a brief overview of the process of writing such software. This software is extraordinarily difficult, and I wouldn’t be convinced that it’s possible unless I saw it for myself.

I've explained this is not correct too many times to bother reading anything you write.

You are wrong -- just constantly restating it will not change that.

You just agreed with TaxGuru about this, and then when I repeat it you say I'm wrong.

What am I wrong about exactly?