Discuss: Evaluating Internet Poker Tournaments

[Dynasty]

Evaluating Internet Poker Tournaments by Dene Tribe


To give our author feedback and to encourage discussion, I'm creating this thread to discuss the article linked above.

[burningyen]

Wondering if the author saw my old post on structures in the MTT forum:

http://forumserver.twoplustwo.com/sh...n=0&page=1

There are differences between 9- and 10-handed tables, and the late-tournament structures matter just as much as the 1st hour, IMHO.

[tribet]

Great Chart. Lots of good information. How many hands dealt per hour did you use?

Since I used 70 per hour, I calculated the UB1 tourney for the first hour to use up 9.6%

The data I used to establish the 70 hands per hour was from some of my own play. I also used overall stats from pokertracker which i found on www.bonuswhores.com. All sites generate around 70 hands per hour (9 and 10 handed). This is probably due to individual speeds of the sites but it does equal out very well for comparison purposes.

I agree that the structure of the entire tournament is important. By focusing on the initial structure for the first few levels we can get a good indication of the speed. By evaluating the M at the start of the tournament and determining how the M will change if no chips are accumulated in the first few levels a good plan can be established.

For the Poker Tournament Evaluation System to be completely reliable the blinds need to escalate steadily. As far as I know this is the case at all major internet sites. The chart you produced proves this, Tournaments consuming a lower percentage of the chip stack in the first hour are also the ones with the lower percentage at the 6 hour point.

Dene Tribe

[burningyen]

Unfortunately I didn't have the time to do a comparison of the different hands/hr of the different sites. I mentioned in my post that someone should take a look at those differences. I'm pretty sure that those differences would make a pretty significant difference in the effective tournament structures. For example, my guess is that UB has the fastest software/deal speed, which allows you to play more hands/level.

[tribet]

Actually the data shows each site around 70 hands per hour but that is just the data. That data is from ring games and from my experience the tourneys run a little slower due to players using the clock to consider decisions. In my calculations I posted the blinds every 10 minutes to show 6 full reveloutions per hour. If anyone has experience with a site which is significantly faster on a regular bases I can alter those calculations.

When you generated your chart how many times an hour did you post the blinds to generate your calculations?

[burningyen]

I think you misunderstand how I did the calculations. I calculated the cost per orbit as a percentage of the starting stack, not as a percentage of a dwindling blinded-off stack. So I didn’t have to worry about calculating how fast the starting stack would blind off. And in order to account for the fact that you will have to pay the blinds 11% more frequently per level at a 9-handed table vs. 10-handed, I adjusted the cost per orbit for the 9-handed sites by multiplying the blinds by 10/9. (It’s been so long since I looked at this stuff that I should go back and make sure I didn’t make a math error by multiplying the blinds+antes rather than just the blinds. You only see the blinds more frequently; you pay the antes every hand. I’m pretty sure I did it correctly.) Of course, all of this depends on the speed of play of each site being the same. I’m surprised at and a little skeptical of the bonuswhores.com stats on hands/hr (which are really pokersitescout.com stats). I wonder how the stats were collected.

Re only examining the 1st hour of these structures: back when I started that thread there was some debate about Party, where some pros were saying that even though the Party structure started out bad in the 1st hour, the late game structure was actually slower and superior to other sites’. You can see when comparing the “UB1” and “PT1” structures that the Party structure is worse in the 1st hour but becomes somewhat better towards the end. All the structures have changed since that thread, so I have to go back and re-visit the chart.

[tribet]

Actually i thought the stats came from pokertracker but i will check with pokerscout on their stat collection procedures.

As far as structure is concerned if you find an example of when it is better to look further than 1 or 2 hours to determine speed let me know. The only cases i have ran into so far are in brick and mortar games and this happens only rarely. The cardroom will actually remove levels from the structure to speed up play and this causes large jumps in the cost per round. This is an unfair practice and players should not participate in these events. I ran into this in Caesars Indiana earlier this year and they actually changed the structure just prior to the start of the tournament.

Another reason I think the early structure is so important is that if you emply the correct game plan and accumulate a decent chip stack by the time the middle portion of the tournament comes around you are happy to see the blinds escalate and it is the small stacks that are being killed off by the escalating blinds. I know, easier said than done to get that large chipstack