How did you learn to think in probabilities?

[firlandsfarm]

Given hard facts/numbers I can calculate probabilities then I can think but not the other way round! I can't look at a football match and think that TeamA has a 40% chance of winning as opposed to a 35% chance or a 45% chance. Likewise, I can't build a book for a horse race just by looking at the runners and their 'contributing factors'. I can look at a football odd and think "that looks about right" but only 'about right', not with sufficient accuracy to say "that's wrong, there's value to be had there"! So when in play with odds moving in and out at speed I have no chance of spotting value.

[stueytrader]

There's also a question of what 'probability' are we focused upon?

The probablility that a price moves (many strict traders view)

The probability that a horse wins, or falls, or loses.

All may be probabilities, but have different roles to play in our thinking.

[ShaunWhite]

and think "that looks about right" but only 'about right', not with sufficient accuracy to say "that's wrong, there's value to be had there"!

Same, but not suprising really because unless you're talking coin flips then probablity is often a range not an exact number. And we can never really now know on a match by match basis what it should have been on the day.

I think saying "they're somewhere between a 35 and 45% chance" would be a reasonable thing to say, and if they're priced at 42% then I suppose there's some sort of maths you could use to stake it with it being x% over what you had in the middle of your guess ? I dunno, not my field really.

And then you've got the other problem. eg
I think Team A has an 80% chance of winning.
... and I'm 75% sure about that.
....and I'm 60% sure about being right 75% of the time.
...probably.

[firlandsfarm]

And then you've got the other problem. eg
I think Team A has an 80% chance of winning.
... and I'm 75% sure about that.
....and I'm 60% sure about being right 75% of the time.
...probably.

very true

[stueytrader]

The world (and trading in it) is a stochastic beast, in many different ways.

i.e. we model our best guesses at all times, despite the degrees of randomness thrown in by numerous factors.

[arbitrage16]

There's also a question of what 'probability' are we focused upon?

The probablility that a price moves (many strict traders view)

The probability that a horse wins, or falls, or loses.

All may be probabilities, but have different roles to play in our thinking.

I’d say some in this thread have misinterpreted what is meant by thinking in probabilities. It isn’t about calculating outcomes of an event, thinking in probabilities is about understanding and fully embracing the risk that comes with having your edge play out over a series of trades.

In part it’s about accepting you may be ‘right’ on a trade and still lose money, for example.

[stueytrader]

There's also a question of what 'probability' are we focused upon?

The probablility that a price moves (many strict traders view)

The probability that a horse wins, or falls, or loses.

All may be probabilities, but have different roles to play in our thinking.

I’d say some in this thread have misinterpreted what is meant by thinking in probabilities. It isn’t about calculating outcomes of an event, thinking in probabilities is about understanding and fully embracing the risk that comes with having your edge play out over a series of trades.

In part it’s about accepting you may be ‘right’ on a trade and still lose money, for example.

Yes, there is the 'global' view of probability as well, like you say.

[stueytrader]

I suppose a simpler way to state that type of probability, is just to say 'long-term probability' rather than the more short term forms.

Then, there is the psychology aspect of how much we accept that 'long-term' view of probability.

[arbitrage16]

I suppose a simpler way to state that type of probability, is just to say 'long-term probability' rather than the more short term forms.

Then, there is the psychology aspect of how much we accept that 'long-term' view of probability.

Believe the distinction is conveyed in the terms probability and probabilities, the latter usually preceded by 'thinking in'

[stueytrader]

It is certainly a challenge (and one we all face) to accept longer term and more numerous examples of probability. In other words, to accept that we can lose (and badly) at times, despite doing things right.

It's similar to what Poker players often face - making good choices, with good probability considered, but having to accept it might not work short term. To accept that our bigger run has a positive probability of success.

[arbitrage16]

It's similar to what Poker players often face - making good choices, with good probability considered, but having to accept it might not work short term. To accept that our bigger run has a positive probability of success.

Yep absolutely, good analogy. A willingness to let markets/hands pass you by. In poker it may be due to unfavourable cards or running on tilt, whilst with trading tilt applies, and it would also pertain to there not being an occurrence of a favourable setup...but even then, everything can appear favourable, but no market is ever the same, so we can't 'know' what will happen, and have to coach ourselves to accept that. Not easy for humans to do.

Mark Douglas wrote fairly in-depth on the topic of probabilities.

[sniffer66]

It is certainly a challenge (and one we all face) to accept longer term and more numerous examples of probability. In other words, to accept that we can lose (and badly) at times, despite doing things right.

It's similar to what Poker players often face - making good choices, with good probability considered, but having to accept it might not work short term. To accept that our bigger run has a positive probability of success.

Interesting point. It was poker botting that taught me to think in terms of probability. I harvested or bought millions of real money hands and used to run a very simple push or fold strat based on the results and my 2 hole cards playing Texas Holdem. Very easy to go on tilt shoving with the top 13 best hands and get beaten by some muppet calling you with 72 off and winning. Armed with the data you just smile and know you'll take the money in the long run

[wearthefoxhat]

It is certainly a challenge (and one we all face) to accept longer term and more numerous examples of probability. In other words, to accept that we can lose (and badly) at times, despite doing things right.

It's similar to what Poker players often face - making good choices, with good probability considered, but having to accept it might not work short term. To accept that our bigger run has a positive probability of success.

Interesting point. It was poker botting that taught me to think in terms of probability. I harvested or bought millions of real money hands and used to run a very simple push or fold strat based on the results and my 2 hole cards playing Texas Holdem. Very easy to go on tilt shoving with the top 13 best hands and get beaten by some muppet calling you with 72 off and winning. Armed with the data you just smile and know you'll take the money in the long run

As a side issue, I'd like to see a robot play in a live poker competiton, in a similar way that Deep Blue played chess against Gary Kasparov. (and won)

A human would be needed to input key data information, like the hand received, stack sizes, table position, number of players at the table. The decision of the robot then made in the form of fold, check, bet x amount, raise, all-in. Each action made by each round of betting, including the flop/turn/river. In time, heat sensitive receptors could pick up body heat, facial expressions, talking tells..etc

Even a robot tournament of sorts which would then depend on the best strategy programmed into each one. A sort of Robot Wars for poker. A robot that trash talks would be fun. Could base one on Phil Hellmuth..."Honey he called me with 7-2 offsuit, idiot from tin can alley."

[sniffer66]

It is certainly a challenge (and one we all face) to accept longer term and more numerous examples of probability. In other words, to accept that we can lose (and badly) at times, despite doing things right.

It's similar to what Poker players often face - making good choices, with good probability considered, but having to accept it might not work short term. To accept that our bigger run has a positive probability of success.

Interesting point. It was poker botting that taught me to think in terms of probability. I harvested or bought millions of real money hands and used to run a very simple push or fold strat based on the results and my 2 hole cards playing Texas Holdem. Very easy to go on tilt shoving with the top 13 best hands and get beaten by some muppet calling you with 72 off and winning. Armed with the data you just smile and know you'll take the money in the long run

As a side issue, I'd like to see a robot play in a live poker competiton, in a similar way that Deep Blue played chess against Gary Kasparov. (and won)

A human would be needed to input key data information, like the hand received, stack sizes, table position, number of players at the table. The decision of the robot then made in the form of fold, check, bet x amount, raise, all-in. Each action made by each round of betting, including the flop/turn/river. In time, heat sensitive receptors could pick up body heat, facial expressions, talking tells..etc

Even a robot tournament of sorts which would then depend on the best strategy programmed into each one. A sort of Robot Wars for poker. A robot that trash talks would be fun. Could base one on Phil Hellmuth..."Honey he called me with 7-2 offsuit, idiot from tin can alley."

It's going back 15-20 years or so but I used to use this:

https://documentation.help/OpenHoldem/

There are some articles out there re the latest, university created bots beating current top players. Poker sites and real players hate them and I used to get banned and bank confiscated regularly (against T&C's). Uni ones obviously aren't for public release, I guess it's a good way to test out an AI when there are unknown factors like bluffing, which you don't get in games like chess

[sniffer66]

https://www.npr.org/2019/07/11/74066147 ... 8434809726