Let me start this extremely long post by answering a few questions.
[ QUOTE ]
How much of a difference do you think it would make if they just made a few simple adjustments, something like freezing the tournament clock immediately as soon as there is an all-in and a call?
[/ QUOTE ]
Harrah's has specifically implemented that this year, and I have witnessed this first-hand at the ESPN final tables. When there is an all in and a call, the TD stops the clock while ESPN plays out the moment for drama. Sometimes there is an additional delay while they interview a busted player, but the clock doesn't resume until the cards are shuffled for the next hand.
It's a welcome change, but one that only affects ESPN events. In all other events (even the webcast ones), the all-in hands aren't slowed down at all for dramatic purposes -- trust me, I'm *very* aware of this as I try to keep up the live hand-by-hand coverage. So score one for Harrah's on this point, compared to the last few years.
[ QUOTE ]
Last year I'm pretty sure they increased the time for each level at the final table. Can someone confirm/deny?
[/ QUOTE ]
While I can't find the official structure sheets from last year on WorldSeriesOfPoker.com, Card Player has them embedded in their coverage for 2006. And carefully checking last year's live updates confirms that at the final table, the levels were increased to 90 minutes. This year, levels are 60 minutes all the way to the end.
How did the Player's Advisory Council allow this significant change? Probably because it's buried in the fine print at the bottom of the structure sheets, and the Advisory Council probably never noticed. It's doubtful that anyone told them, "By the way, the levels at the final table are decreasing from 90 minutes to 60 minutes this year." But when the player's signed off on the structure sheets (whether it was literal or figurative), they unknowingly approved the change.
[ QUOTE ]
Also, didn't they have different structures for 5K events and 1500 events in 2006?
[/ QUOTE ]
According to CardPlayer.com's 2006 coverage, the structures were slightly different. Comparing the two structures side-by-side, the blinds were slightly higher in the 5,000 events than the 1,500 events.
2006 $1,500 NLHE Structure -
http://www.cardplayer.com/tournaments/blinds/3191
2006 $5,000 NLHE Structure -
http://www.cardplayer.com/tournaments/blinds/3198
[ QUOTE ]
I ran the numbers on all of the structures as soon as they were released. Results showed that the tourneys were going to speed up v. fast in the late stages.
[/ QUOTE ]
I commend jsmith5 for bringing this issue to light with such a thorough, well-written blog entry at Pokerwire and here on 2+2. It's sparked a lot of much-needed analysis. (Admo, who I quoted above this paragraph, seemed to understand the core issue before most of us.)
Before I do a side-by-side comparison of last year's structure and this year's structure, I'd like to address a few misconceptions.
THE SPEED OF A FINAL TABLE
The speed of a final table (whether measured by time or number of hands) is a symptom of a problem, not proof of one. There are a lot of factors that determine the speed of a final table.
Example: At a WSOP Circuit Championship in Atlantic City in December, 2005, the first seven players busted in just 75 hands before John Juanda and Chris Reslock battled heads-up for another 225 hands -- three times as many hands as it took to reach heads-up play. (Coincidence: Nick Schulman finished 4th that day, and he was also at the 48-hand massacre in Event #22 a few days ago.) For what it's worth, I was the one covering that event for Card Player. Link:
http://www.cardplayer.com/tournament...?view_all=true
FINAL TABLE STARTING STACKS RELATIVE TO BLINDS
Final table starting stacks relative to the blinds is not a reliable indicator either, for the same reasons. Sometimes the final table is reached more quickly than others. That 10th-place finisher can bust in a few hands, or a few hours (and levels). The average starting stack would be the same in both cases, but the comparison to the blinds would obviously be much different.
NUMBER OF ENTRANTS
The number of entrants is also irrelevant. If more players enter an event, there will be more chips in play, but it will usually take longer to reach the final nine players -- and those differences cancel each other out. I can't think of a single case where a structure sheet was changed because of the number of entrants -- even in 2004, when the 2,576 entrants in the Main Event represented a 307% increase over the previous year and completely overwhelmed expectations at Binion's. A good tournament structure is good regardless of the number of entrants.
SIDE-BY-SIDE COMPARISON: 2006-2007
Harrah's has made a big deal out of the fact that all events start with double the number of chips this year. The blinds increased as well, but Harrah's claimed the extra chips gave them flexibility to slow the structure in the early and middle rounds.
However, in this thread, anecdotal evidence points to the possibility that the blinds are actually *higher* than last year.
So which is it?
This is a relatively simple comparison. Since players start with twice as many chips this year, just divide all the blinds and antes in half to compare them with last year. That's it -- divide by two.
For example, if the starting stacks were doubled, and all the blind levels were exactly double of what they were last year, the net effect would be zero. You could triple the chips or quadruple the chips, and as long as the blinds and antes increased proportionally, the net effect would always be zero. (Psychologically there might be an effect, but that's outside this discussion -- if players trick themselves into playing faster, it's their own fault.)
I've compiled a side-by-side comparison (adjusted for starting stacks) of the blinds for the $5,000 no-limit hold'em event that Jeff Cabanillas won last year in roughly 8 1/2 hours (field: 622 players), and James Mackey won this year in 48 hands (field: 640 players).
We don't know how many hands were played in the event last year, because Card Player didn't do hand-by-hand coverage. I offered to freelance for them, but they turned me down so quickly it made me dizzy.
2006 Structure:
http://www.cardplayer.com/tournaments/blinds/3198
2007 Structure:
http://www.worldseriesofpoker.com/to...mp;groupid=309
Level 1
2006: 25-50 (0)
2007: 12.5-25 (0)
Level 2
2006: 50-100 (0)
2007: 25-50 (0)
Level 3
2006: 100-200 (0)
2007: 50-100 (0)
Level 4
2006: 100-200 (25)
2007: 50-100 (12.5)
Level 5
2006: 150-300 (25)
2007: 75-150 (12.5)
Level 6
2006: 200-400 (50)
2007: 100-200 (25)
Level 7
2006: 300-600 (75)
2007: 150-300 (37.5)
Level 8
2006: 400-800 (100)
2007: 200-400 (50)
Level 9
2006: 600-1,200 (200)
2007: 300-600 (50)
Level 10
2006: 800-1,600 (200)
2007: 400-800 (100)
Level 11
2006: 1,200-2,400 (300)
2007: 500-1,000 (150)
Level 12
2006: 1,500-3,000 (400)
2007: 750-1,500 (200)
Level 13
2006: 2,000-4,000 (500)
2007: 1,000-2,000 (250)
Level 14
2006: 2,500-5,000 (500)
2007: 1,500-3,000 (250)
Level 15
2006: 3,000-6,000 (1,000)
2007: 2,000-4,000 (500)
Level 16
2006: 4,000-8,000 (1,000)
2007: 3,000-6,000 (500)
Level 17
2006: 6,000-12,000 (2,000)
2007: 4,000-8,000 (1,000)
Level 18
2006: 8,000-16,000 (2,000)
2007: 5,000-10,000 (1,500)
Level 19
2006: 10,000-20,000 (3,000)
2007: 7,500-15,000 (2,000)
Level 20
2006: 12,000-24,000 (4,000)
2007: 10,000-20,000 (2,500)
Level 21
2006: 15,000-30,000 (5,000)
2007: 15,000-30,000 (2,500)
Level 22
2006: 20,000-40,000 (5,000)
2007: 20,000-40,000 (5,000)
Level 23
2006: 25,000-50,000 (5,000)
2007: 30,000-60,000 (7,500)
Note: Play ended during Level 23 in 2006 and during Level 21 in 2007.
As you can see, the blinds and antes were lower during the entire tournament in 2007 (through Level 21, when it ended). In that sense, the 2007 structure is "better" -- there *is* more play, as promised.
However ...
While the blinds and antes are lower, they accelerate much faster on Days Two and Three. Take another look at the side-by-side comparions -- for the first 13 levels, the blinds were about 50% of what they were last year. Statistically speaking, a higher percentage of the starting field should survive Day One.
But from Level 14 to Level 21, the blinds quickly catch up to where they were last year. In just seven levels (mostly on Day Two), the blinds go from 50% of last year to 100% of last year. That's a steep acceleration curve that seasoned tournament players aren't used to. Statistically, that will increase the chip swings and the rate of bustouts during that period of acceleration.
It's a weird juxtaposition. While this year's structure definitely offers more "play," it's a much "faster" structure. Traditionally, poker players have assumed that these two variables were one and the same.
They're not.
Harrah's can't be blamed too much. In past years, everyone complained that there wasn't enough play at the start of the lower buy-in events ($2,000 and less). Harrah's has definitely corrected that. And since there is mathematically more "play" than last year, it was difficult to argue with the new structure.
But the law of unintended consequences has kicked in, and most people didn't see this coming -- including the Players' Advisory Council and the poker media. (Kudos to Admo for figuring it out before the rest of us.)
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
CONCLUSIONS
1. Blind levels at the final tables should be 90 minutes, like they were last year. The levels at final tables are changing 33% more often than they were last year.
2. The 2007 blind structure offers more "play" throughout the tournament than 2006, but is much "faster" on Day Two, when it matters more than Day One.
3. I have too much time on my hands, and I'm an idiot for not writing this up as a PokerNews article and getting paid for all this research/writing. But it's late, and I'm too tired to reformat this as an article.
4. Everyone from the Players' Advisory Council on down should lobby for 90-minute levels at the final table. I think Harrah's would be willing to make this change, but I don't know how much influence ESPN has on this decision -- and Harrah's would never open themselves to WPT-style criticism by blatantly changing the final table structures solely for ESPN events.
5. When the structures are evaluated for next year, they need to be compared more directly and more closely to prior years. The speed with which the blinds increase (and during which stage of the tournament) is just as important as the size of the blinds themselves.