Sunday 10 June 2018

Leveraging Changing Probability

My recent post on Scientia Trading garnered some attention and debate, which is always a good thing.

I mentioned in the last post my issue with stopping after four bets in any one day and thanks to Matthew, who pinpointed for me where I might find Mel's thoughts on this, I was able to hear them without having to sit through almost 8 and a half hours of fluff.

I did start to watch Episode One, but gave up after ten minutes. It wasn't the most motivating of talks. 

Apparently the thinking behind the four bets limit is to avoid burnout, with six spells in a day of watching twenty minutes of a football match too much for him. 

Mel speaks of this 20 minutes as being in a 'high stress situation'. Our ideas on what life events can be considered 'high stress' appear to differ. If I'm stressing over a bet, it's because my stake is too high. Reduce the stake size! Save stressing for the things in life that are actually important. 

Mel's struggles in this area certainly boost my argument that outright punts are the way to go. If two hours of 'work' a day is too much, I hope Mel never gets a job with some real responsibility. Not that he'll ever need a job with a huge edge like this. 

In an attempt to justify this nonsensical four matches a day limit, Mel says:
"When I'm trading, I can't just go out there on a daily basis trying to trade as many times as I can thinking that probability will help me through. Rather, I need to leverage probability as a window of opportunity in which to extract the profits that I want because I know that probability can change and what used to happen in the past was that I would get myself into a situation where I was up, had some profit for the day, and I would just trade and trade and trade because I'm thinking to myself well if I can trade 20 to 30 times a day, look at the profit potential and then what would happen? I would be up £300 or £400, I would be back down to zero and then £300, £400 down in the red and then I would continue to trade because now I'm trying to get myself back to break even and before you know it I'm mentally fatigued and I end up with a loss for the day or break even or just a small profit, and I remember thinking to myself "I can't consistently trade this way on a day by day basis" 'cos of the fact that if I was living on my own, then maybe but I need a much more precise way to trade."
Where do I begin?  First of all, it doesn't appear that Mel is actually trading - rather, he is watching a match "trying to develop a justification" (his words) to place a bet (not a trade, since I haven't seen or heard any mention of exiting the open position). 

"Trying to develop a justification" sounds about as un-scientifica as you can get. Guessworkentia Betting would be a more appropriate name for this endeavour. 

Of course probability changes. The probability of a goal after 83 minutes is one thing. The probability of a goal after say 86 minutes is another. 

What "leveraging probability as a window of opportunity in which to extract profits" means is anyone's guess. Sounds like meaningless mumbo-jumbo nonsense to me.

What Mel seems to be missing is that if he seriously thinks he can hit winners at 2.8 50%-60% of the time, he is sitting on an absolute goldmine, although yesterday he did admit that this is now closer to 50%. It would not be a huge surprise if this ultimately drops to approximately 35.714285%. 

With an edge that huge, (a win rate of 50% to 60% at 2.8) full Kelly suggests staking 36.6% of the bank, and anyone who has been betting for any length of time will tell you that it is palpable nonsense.

Denzillion @bettingMayhem tweeted that:
I haven't seen any reference to 'investors' myself, but presumably these are people contributing to his starting bank rather than subscribers to any service he might be offering.

I'm not sure how an in-play service could work for this situation in any case. This would also beg the question, why would Mel need investors with such a huge edge? Why would you not start with a £100 bank and build it up for yourself, beholden to no one and perhaps reducing the stress he feels. 

Another error Mel made in the few minutes I listened to the webinar was when he was talking about having a "Premium Hand" which is a poker term and totally irrelevant for a discussion about betting on football.  

Mel plays some audio from a trader who basically talks about the importance of maximising your profits when you are trading well, "letting winners run", but this is in the context of trading in one specific market. 

Mel tries to justify quitting after three losses in terms of "it's not my day", or "the market Gods are against me", completely missing the point that each match is an individual event, and that his huge edge applies to each match, regardless of previous and unrelated results. 

If you've found a casino that pays 50-1 on any individual number, would you arbitrarily limit the number of spins you play based on whether the first three spins are winners or not? Of course not. If you have a positive expectancy, you maximize the number of events in which you are active.

How results are distributed should have no bearing on your betting. In fact, in Mel's intro, he writes:
You can only learn if you embrace the good the bad and the ugly of how trading results are distributed because for years I tried to fight probability and this hindered my progress. It was only by accepting uncertainty that I found peace with my trading.
Mel has much to learn, including from himself. You can't "fight" probability. Markets aren't conspiring against you. 

Phil @workingawayphil, who seems a pretty sharp chap, had a few comments on the likelihood of someone consistently being able to find such huge edges in football markets. Denzillion wrote:
From what I can gather he is assessing multiple pre match stats and in play indicators so has taken a view that from minute X, the win rate is X% hence any price X and over is value.
and Phil replied that:
He also clarified with:
Phil added:
I collected data from thousands of games, goal times, odds, shot locations penalty/ free kick etc. Seen so many people on twitter talk about edges but every time I’ve looked into them in real detail I’ve either found them to be wrong or a tiny sample size.
The problem with a fundamentally flawed approach of using pre-match stats, 'goal times' etc. is that not only are all matches independent events, but that as previously mentioned, far better resourced enterprises have already crunched this data and more, and extracted any value out of the market. One example:
Star Lizard Consulting, which was set up by Mr. Bloom’s associates to provide support for his betting syndicate, operates like a quantitative hedge fund. About 200 employees — traders, software engineers and analysts — focus on helping Mr. Bloom’s syndicate make data-driven bets on soccer and other sports.
The gambler in his one bedroom flat has little to no chance of beating this kind of opposition long-term.  

Denzillion suggested that maybe:
I dunno, does Tony [Bloom] care what Mel does in his bedroom in Iceland Division 2 with 30 spectators. Are you being too cynical here? Time will tell if his win rate stands up.   
Possibly not, but the problem here would be that liquidity is correspondingly low and when you back at 2.8 in an illiquid market, the true odds are probably 3.0 or greater. If the market is liquid, then no doubt Mr. Bloom would be very interested. 

If Mel has truly found a system that generates winners at even 40% of the time at his 2.8 target, then he has an impressive edge and really shouldn't be telling everyone about it. If anyone knows more about the 'investors' angle or where it is mentioned, please let me know. 

No comments: