The Lucky Donut

Nope, nothing really interesting happened.  Still, it’s less than ten weeks before I’ll be in Vegas.. bear with me during the dark weeks 🙂

7. Pocket pairs

I play 93.6% of pocket pairs, and raise them preflop 43.3%.  I do have negative lines for 77, 66, 44 and 22 though.  I don’t see any major problems, just a few big pots I lost with a set against a rivered flush, or a full house against quads, etc.  I do still think I might need to dump small pairs in early position because I often struggle to get a big enough payoff when it gets raised ahead of me and I do flop a set.

8. Suited connectors

Profitable to the tune of 9BB/100.  I cold called with a suited connector only 12 times, mostly KQs and QJs.  Overall I lost, but it’s a pretty small sample size.  The hands I played were almost all only for a tiny raise and in position.  Doesn’t feel like there’s much wrong here.

9. Unsuited connectors

Leak.  Only a small one, but I’m losing 1BB/100 with unsuited connectors.  This includes AKo, which I’m nearly $90 down on.  That’s because of some big pots I lost with top pair vs a set or two pair.  I’m getting more confident playing AK now, and feel like I lose less when I’m beaten, so I expect this figure to improve.  I’ve cold-called unsuited connectors 11 times and won once.  Seven times it was with JTo and the one pot I did win was very small.  Mental note made.

10. Postflop aggression

Wow, I c-bet a lot.  After a preflop raise, I bet or raise the flop 77.8% of the time.  Is that too high, if the guide I’m referring to says "at least 40%"?  Well I don’t think so, because filtering my hands on just these situations, I’ve made 122BB/100 when I continuation bet and when they lead out and I get to raise, the win rate soars to a phenominal 191BB/100.  I can live with that.

My overall postflop aggression factor is 2.83.  Plenty.  However this is mostly from my flop play: 4.12 on the flop, 1.83 on the turn and 1.90 on the river.  It’s not exactly passive on the later streets, but I wonder if this shows that I slow down a little too quickly and could be betting and raising a little more.

11. Check-raising

I’ve check-raised 111 times, 1.11% in total.  Every one of them, naturally, felt great.  I’m not doing it excessively, and overall the hands where I’ve check-raised show my biggest win rate: 449BB/100.  Of course this figure is pretty meaningless, because I’m usually only check-raising with my strongest hands and by its nature a check-raise builds nice big pots.  Still, it looks good just to finish with such an impressive figure!

What better way to chill out at night when working away from home than to analyze your Poker Tracker stats?  I’d even set up GoToMyPc especially so I could get at PT and drill down whenever I felt it necessary to get into the nitty gritty of my play.

Well I thought it would be a good idea, but I crashed out after a heavy night of partying.  And by partying, I actually mean lugging server hardware from one floor of a datacenter to another.

Here’s the first part anyway.  I based the analysis on this guide, and answered each point in turn.  I’ll do the same thing after another 25k hands.

1. Do I have sufficient preflop aggression?

Apparantly not.  On the button and two seats behind I’m raising about half the hands I play, but everywhere else it’s lower.  I already think I play too many weak hands out of position so eliminating them would improve this figure – not by raising more, but by calling less.

2. Am I positionally aware?

Indeed I am.  My VP$IP figure on the button is 26.21, whereas under the gun it’s 13.13.  So I’m very close indeed to playing twice as many hands in the best position as in the worst.

3. How’s my stealing?

My attempts to steal the blinds figure is 33.55%.  When filtering my hands only for steal attempts, my win rate is $59 per 100 hands.  Not too shabby, and much more than double my overall win rate – in fact more than ten times bigger!  I think there are other areas I can look to improve ahead of this.

4. Defending the blinds

This is good and bad.  The magic number, so they say, is  a loss of 0.375BB/100.  Anything worse than this and you may as well just check/fold every blind.  My number is is virtually the same figure, but positive: 0.37BB/100.  It’s reassuring.  I don’t make money from the blinds overall (nobody does) but when I decide to pay to play from these positions, it’s been profitbale.  But it could be more profitable.  I’m at -0.46BB/100 when facing a blind steal.  Defend less, then.

5. Heads-up play

This was a pleasant surprise: I’ve won 7BB/100 when heads up going to the flop.  I certainly didn’t expect this figure to any higher than my overall win rate.  Continuation bets apparently do rock.  However, when I filter only the hands where I didn’t raise, it’s a loss of 6BB/100.  I’m not profitable when limping with weak hands.  Most likely it’s small pocket pairs and suited aces when I’m out of position (given that I already know I raise a wide range in late position) and I suspected this.

6. Multiway pots

I’m winning 5BB/100 in multiway pots overall, and 3BB/100 when limping.  Guess there’s nothing wrong with that.

This is already long enough, so I’ve split the analysis into two parts.  Two blog entries for the price of one – bargain!  Some may say it’s a cheap trick to make readers come back for more, but given the content is really all for my benefit and not at all interesting to anyone else, that’s hardly the case.  I’m just getting bogged down in all those lovely figures.

You never know, something interesting may happen that I just have to write about in the meantime.  Doubtful though.

A month or so ago, I came across a security hole in an online poker network.

I’d thought about writing something about it after it had been fixed but time passed and I’d forgotten quite how major it was until I just mentioned it to someone who works for a network operator.  His reaction was similar to mine when it first came to light: holy shit.

Firstly I need to say that this has definitely been fixed now and it was, rightly, treated with some urgency by the developers.  However, even they didn’t know this was an issue.  Nor did any of the 40+ operators on the Microgaming network (formerly Prima Poker), which includes high profile UK names like Stan James, Ladbrokes and Bet365.

All of their players were at risk.

We can only hope that because so many different technical teams had failed to spot this, fraudsters hadn’t noticed it either.  After all, it’s such a fundamental security flaw, you probably wouldn’t even think to look.

OK that’s plenty of hype.  If you have the geek gene, this picture may frighten you immediately.  If not, read on and I’ll explain what it means:

This is the text view output from an HTTP traffic debugger.  It shows that player information was being sent over the network in the clear.  I’ve highlighted the key parts: if you read between the ampersands, you can see my username and password (obviously this isn’t my password, I just changed it for the screen grab) and my real money balance.

Yes, at the time I did have over $14,000 in my account.  I don’t any more.  It was never really mine, just a fallout from testing new deposit methods.  But I did sit down at a $1/$2 limit table with a five figure roll once. 🙂

From a crook’s point of view, being able to see the real money balance is a luxury that would not normally be afforded to them by using keyloggers or the less subtle approach of watching people as they key in their username and password.  They would not even need to attempt to access a stolen account to know whether there is enough money in there to make it worth their while trying to run off with it.

Although I ran this traffic sniffer on my own PC, software does exist to read such traffic over a network.  It’s the reason you have to look for the padlock in Internet Explorer when you’re entering credit card information.  Then you know the details are encrypted before they are sent in such a way that only the web server can understand them, and not anyone listening to the network along the way.

A point I failed to get across recently when I had to pay import duty when collecting from a Parcel Force depot and they took me into the back office to enter my credit card information in an insecure web page.  Seemingly, nobody had ever challenged this before.

There’s no padlock in a poker program, you just have to trust it.

The Microgaming client was in fact using SSL to send encrypted requests to the server.  However along the way, they were being redirected and ended up unencrypted in the process.  The redirector idea is great in theory, meaning that if an operator wants to change the location of a page or a script that’s used for their site, their players don’t need to download a whole new client for that simple change.

It’s the sheer number of requests that sent this information that was really worrying.  Simply logging in to the client was secure, no username or password was visible.  Accessing the banking pages was similarly secure.  Whew, you may think.

However, accessing "My Page" sent all this information in the clear.  This page typically shows players their loyalty point status, allows them to change their contact information and also allows access to banking anyway.  It could often be the button that players press to make a deposit or withdrawal instead of "Bank".

More worrying, as soon as you were logged in, a promotions page appeared in a popup.  This is just a web page designed to appear in a window inside the poker program.  So why would it even need to be passed information about the user?  Similarly, the promotions banner that is displayed in the lobby was being passed all this information.  There’s really no need.

These screenshots (click to enlarge) shows all the sensitive data being passed out and then back again.  At least on the way out (the bottom right pane shows an HTML form that submits these values to the actual location of the pages) it uses a secure connection.

The operator’s news page was accessed in a similar way.  There’s no need at all to send a password to this page – is the latest news really only available to registered players?  The responsible gaming information and support pages also received way more information than they needed.  No password here, but all the other information is present.  Someone must have made the decision to omit the password but to still transmit username and balance to these pages.  I just can’t think what the reasoning would be for that decision.

So that’s six different places that a user’s account information was being transmitted.  In case that’s not enough, there was one other instance that made sure anyone who was listening in for account information would not be disappointed.

In the program’s lobby – the screen where you search for which poker tables you want to play at – there is a scrolling message that is set by the operator.  Because this message is intended to be updated frequently – for instance, it might say "check out our money added tournament starting at 8pm" – it refreshes every three minutes.

As this process was also transmitting all the sensitive information in the clear, all a wannabe thief would have needed to do is set up a traffic sniffing program for a few minutes and wait to be furnished with the account details.  Minimum effort, maximum loot.

Marquee text

The screenshots I’ve posted are from Gutshot Poker.  I need to stress that this operator was not at risk from this flaw – it was spotted and fixed before the player base was migrated from their previous operator.  This hole would have been catastrophic to Gutshot, who operate an internet cafe where every PC is used to play online poker.  A scammers paradise.

So I’ll say it again: this has now been fixed.  There’s no need to kneejerk and cashout from whatever site you play at.  Don’t bother the site’s support asking whether your money is safe.  It finally actually is.

But if you’ve ever played on a Microgaming site in the past you really should change your password right away.