Sunday, November 08, 2009

Graph of a tournament

When I'm playing in a live poker tournament, I count my chips often. Stack size is virtually always one of the most important factors in every decision, so I want to be sure I know where I stand. But I have a problem: my short-term memory doesn't hold the result very well. So long ago I came up with the obvious solution: Write it down. I always have scratch paper and a pen in my shirt pocket anyway. That way, when I'm in a big hand and need to know how many chips I have, I don't have to be distracted from what I want to be focusing on to try to either recall how many chips I had at the start of the hand, or try to count them when the spotlight and pressure are on me. (I have made serious errors counting under those circumstances.) I just glance at my little sheet of paper.

I also usually jot down the time for each chip check, so that (1) I know if I've neglected to update it recently, and (2) I stay aware of the time of day.

I have long thought that it might be cool and interesting to take an entire tournament's worth of chip counting and graph it out. But the problem is that I usually don't do well in tournaments. I bust out before making it to the money, so the graph falls off of a cliff.

Last night was the first time since I have been undertaking this practice that I made it all the way to the end, and can thus make a graphic display of how my stack evolved over the course of the tournament. The other four who ended with approximately the same number of chips (all of us starting with the same 10,000, of couse) would, I'm sure, have very, very different lines, though all with the same starting and end points.

I used an Excel spreadsheet to enter the times, which it then converted into number of hours from the start of the tournament, shown along the x-axis. My goal was to do a count every ten minutes, plus after every major swing up or down. I didn't always succeed, but I came pretty close. You can see the result above.

I started out like gangbusters. Half an hour into the thing I had nearly tripled my stack, having busted two players. That all came in three big hands, with JJ holding up and AA holding up twice for the two eliminations.

As you can see, though, that was followd by two and a half hours of basically no net forward progress. Then shortly after I was moved to a new table I had the hand of the day: 7-10 of clubs on the button. I flopped top two pair, and it held despite aggressive, scary betting from the other two players contesting the hand. That took me to 61,750, and as I looked around the other tables at the time, I didn't see anybody else with more than about 30,000, so I was not only the tournament chip leader, but massively so.

It's a good thing, too, because I then hit the Death Valley of card deadness, the Dead Sea of card deadness, the Arlington Cemetary of card deadness. This lasted for the next couple of hours. It was disconcerting to see that enormous lead dwindle as the tournament clock showed the average chip stack number growing ever closer to what I had. But I was sure glad that I had entered that horrible spell with a stack sufficiently large that I could coast for the long, long time that I needed to.

The next interesting section of the graph is between 5 and 6 hours in. That's when we hit that threshold where there are lots of short stacks that have only one move, and the bigger stacks have to decide whether to call the all-in or not. That's always a high-variance interval in a tournament, and that fact shows nicely in the jagged up-and-down line; obviously I won some of those and lost some. I think that last uptick right at the 6-hour mark was when I busted the bubble boy with a bad beat (I was about a 3:1 underdog when the money went in) and moved to the final table.

It looked to me that I was probably fourth in chips when the final table of ten began, with about 10,000 more than the mean. I made some initial careful progress, but then had a big downturn when I overplayed a suited A-K that missed the flop. I would have had a substantial loss there no matter what, but I lost about 20,000 more than I should have because of being too stubborn.

I then turned up the aggression, and became one of the most frequent blind stealers. I was determined not to try to creep up the money ladder one little step at a time, meekly hoping that others would bust out first, but instead go for the gusto. I had apparently developed enough of a tight table image that I could get away with it, because I was open-raising all-in for 10-18 big blinds with 5-6 suited or 10-8 offsuit, and getting the players in the big blind to reluctantly fold much better hands (like suited K-Q, medium pocket pairs, or small aces) face up.

The last move up the chart came from busting out the last player to go, and shortly after that we negotiated the tournament-ending deal.

So there you have it. That is one possible route to success in a tournament. There are infinitely many possible paths that get to the same place, but this was mine. I'm feeling rather fond of it, and glad that I now have it memorialized this way.


Luke said...

Hi Grump, I am DayTripping, the one you busted out on the chop bubble last night. I did notice your frequent shoving preflop near the end there and I was impressed with it. Congrats on your cash!

Chris said...

Interesting to see someone else do this.

Here's one I made earlier, but with a different approach:

I wrote this web site a few years ago so I could report chip counts in real time when I had backers for a tournament, by sending a text message from my phone with the number of chips to a predefined number.

I tried to open this up to other players, but response to it was poor and It didn't take long before everyone was using Twitter instead anyway.

But I'm sure there are more people than myself who just love graphs :)

I've toyed with the idea of developing this a bit more; not only making the graphs a little prettier but allowing updates from a mobile web page or iPhone app (which could even remind you every ten minutes that it's time for an update), or even looking for tags in a twitter feed. When I first made it, mobile internet was nowhere near as available as it is now so SMS seemed like the best way to throw data at it.

Would love some feedback on whether you think this is a good idea and something you'd use, rather than writing the counts down on a bit of paper. Or whether you prefer to keep your trainwreck graphs a closely guarded secret, and just cherry pick the best ones :)