ROI estimate for new Party SnG structure

[HelmetCrash]

Guys,

Assume for the moment that ROI is directly proportional to the average number of hands played in a SnG. This is certainly an approximation but not an unreasonable one. The more hands a winning player has to gain his edge, the higher his ROI will be.

Also keep in mind that the starting reference point is -9% ROI as it is the average ROI of everyone at the table due to house rake ($20+$2 SnG). A player that averages +15 ROI is really beating the table average ROI by 24%.

From a recent 2+2 post by Mark1978 and my own observations:
Avg # of hands in a 5 min speed SnG is about 20% less than the old 10 hands per round structure.
Avg # of hands in a 10 min speed SnG is about 40% more than the old 10 hands per round structure.

Using these assumptions, here is how the new structure will affect ROI:

ROI in old structure___________5min SnG ROI____________10min SnG ROI
+20%______________________+14%___________________+ 32%
+10%______________________+6%____________________+ 18%
+5%_______________________+2%____________________+ 11%
+2%_______________________+0%____________________+ 6%

Of course, hands early on in a tourney are not as important as bubble hands and beyond. Also, in a 5 min SnG there are like 6-8 hands in the first 3 levels, and like 12-15 hands at level 5 and 6. But, you will be around 95% of the time to play level 1, and only make it to level 6 like 25% of the time. This is a key reason why I consider the approximation I make here to be valid.


Whether or not you agree with what I say here, it is tough to deny that in a 5 min SnG good players will have to some degree fewer number of opportunities per SnG to overcome the house rake and -9% table avg ROI and attain a profitable ROI. This may be obvious, but for marginally winning players who are beating the –9% table average ROI but are not attaining a significantly positive ROI, the right choice is probably the 10 min SnG.

[Bigwig]

Eh, not really. The increase and decrease won't be a precise proportion. It would be more exponential. Curved, not a straight line.

[Relvin]

Nice post, so are you claiming that on average in the first 3 levels of the 10 min round sit and go's that each player will be able to play between 12 and 16 hands (per level). This sounds good, I thought less hands were typical. What did you base these numbers on? I have not had a chance to test it out myself much yet.

- thanks

[graydot]

I will be surprised if the ROI changes are anywhere more than +/- 3

[Freudian]

[ QUOTE ]

Assume for the moment that ROI is directly proportional to the average number of hands played in a SnG. This is certainly an approximation but not an unreasonable one.

[/ QUOTE ]

It is an unreasonable assumption. There might be some weak correlation but it certainly isn't linear.

[FlyWf]

I take significant issue with that analysis. Assuming that good players edge is primarily concentrated in the higher blind levels, you're getting >10 hands per level after level 4 or 5 of a speed. Anyway, I don't understand why everyone is so "OMG what ROI am I now?" about the change. Play some games and find out! There's no way to determine this rigorously.

[Jack Fate]

I'm not sure if I agree with your estimates, but I think it's an interesting idea. Could you go one more step and convert to $/hr using available estimates of average sng length for each? That will be the more interesting number I think. So we can see if it's worth it to take on a lower ROI due to being able to play more of them.

[HelmetCrash]

I think you are right that it won't be a "precise proportion", this is an approximation after all. But, there are a large number of things you can accurately model using linear approximations, and this is one of them. I truly believe that to a fairly good level of correlation ROI is directly proportional with avg # of hands played in a SnG.

[HelmetCrash]

Mark1978 posted his results comparing the 5 min and 10 min games a few days ago, do a search on him and you should be able to find the post I am talking about.

[HelmetCrash]

I myself believe it is much more than a "weak correlation", but don't think it is perfectly linear either. I DO think real ROI behavior vs avg # hands played is close enough to linear in behavior that the numbers I state in my post give relatively accurate insight into the ROI change one can expect in going from fixed to timed levels.