Why SNGs have MORE variance than ring games

[newhizzle]

quote from mason malmuth's poker essays refering to small(15-20 person) tournaments:

[ QUOTE ]
There is no question that tournaments, whether large or small, offer a significant return on you money if you should happen to get lucky and win. This is the reason why they have become so popular. But the statement i quoted also seems to indicate that small tournaments are a good investment for those on a small bankroll. Is this true? If you are virtually broke, are you better off in these events as opposed to playing in the little games? . . . Does a tournament fit this criteria? In other words, can you have a high win rate and a small standard deviation in these events? Are they a good investment if you currently are struggling to come up with the buy-in? The answer is, probably not.

[/ QUOTE ]

Thoughts, Comments?

My 2 cents:

If we have a player who plays well enough to beat both Sng's and limit ring for an average win rate then he "should" have lower variance per SnG, than ring.

I'll assume 1/2 ring is the same skill level as the $11s.

At the ring tables, the player will experience a downswing of 300BB or $600.00

At the Sng's, the player will experience a downswing of 30 Buyins or $330.00

So that means less variance at the SnGs. I think Aleo used different statistics to come up with his conclusions (I saw lots of SD numbers in there [img]/images/graemlins/confused.gif[/img]) but I dont think they are all that relevent.

If I deposit $400.00 to my account, Id want to play SnGs.

[A_PLUS]

Yeah, I dont know if Mason knew about 8-tabling party SNGs

Thanks Brad for the thoughtful examples. I have a couple simple principled reasons which indicate SNG will generally produce higer variance at a comparable skill/dollar level (e.g., 22's sng V. 25$NL ring:

1. in the ring game you are making many, relavtively, small gambles (hands), say 60/hr; 1-25$ each, with most less than 5$). Compare that to a SNG with ~1-2 gambles (tournys) per hour for 22$ each.

2. In a ring game the probability of winning each gamble is ~0.5. In an SNG it is ~0.3. The opportunity for a series of losses (i.e., running bad) is proportional the size of the probability of lossing each gamble.

Example: loosing 5 gambles in a row

a) ring game: p=.5^5 = 0.031
b sng p = 0.7^5 = 0.17

Over 5x greater for the sng in this example (not a linear relationship).

Further, b/c the size of the gambles are greater for sng the amount of money lost when running bad (or won when running good) is much greater:

5 x ~3$ (a guess as to avg loss in a ring game hand) =15$

5 x 22$ = 110$

I think that generally for a comparable skill level (i.e., assuming you don't play 5+1 sng and 1000$NL ring games) the sng will have far greater variance. This is certainly my personal experience. I have approximately the same hourly income in both but my swings in SNG is ~10-fold greater.

Brett

[The Don]

[ QUOTE ]
[ QUOTE ]
The optimal strategy for those games is to advertise insanely for the first couple of orbits then wait for big hands.

[/ QUOTE ]

This just simply can't be true. The players down there are really really really bad and unobservant. So why waste the money on advertizing that they're not going to pay any attention to anyway, when you could just skip to "nutpeddle?"

citanul

[/ QUOTE ]


Have you ever sat at an NL50 table with $575 before after getting a medicocre run of cards? I think you are underestimating the ability of the low limit player to use basic logic...

[A_PLUS]

[ QUOTE ]
My 2 cents:

If we have a player who plays well enough to beat both Sng's and limit ring for an average win rate then he "should" have lower variance per SnG, than ring.

I'll assume 1/2 ring is the same skill level as the $11s.

At the ring tables, the player will experience a downswing of 300BB or $600.00

At the Sng's, the player will experience a downswing of 30 Buyins or $330.00

So that means less variance at the SnGs. I think Aleo used different statistics to come up with his conclusions (I saw lots of SD numbers in there [img]/images/graemlins/confused.gif[/img]) but I dont think they are all that relevent.

If I deposit $400.00 to my account, Id want to play SnGs.

[/ QUOTE ]

Limit is high variance by nature. What about NL ring 25$max ~~11s.

I dont know, Ive never been any good at NL ring, I have no basis for comparison.

The OP's subject didnt specify which type of ring game, I just used one that could disprove his theory.

[SuitedSixes]

Game A (Sample size 1175 events)

$/event $0.95
SD$/event $17.90
Bankroll required for .5% ROR = $1388
Multitable earn = $23.69/hr

Game B (Sample size = 930 events)
$/event = $6.83
SD$/event = $83.41
Bankroll required for 0.5% ROR = $2699.18
Multitable earn = $24.24/hr


Both of these allow me to make my weekly nut (if I play the required number of hours). If my concerns are (in order):

Lowest ROR
Smaller bankroll swings
Profitability

Which game should I play?

[AleoMagus]

[ QUOTE ]
If I deposit $400.00 to my account, Id want to play SnGs.

[/ QUOTE ]

That may be true for you, but I think that it misses the point of this thread

It's not NECESSARILY true for all people, because your skill levels can vary quite a bit from one form of poker to the next.

Regards
Brad S

[ QUOTE ]
[ QUOTE ]
If I deposit $400.00 to my account, Id want to play SnGs.

[/ QUOTE ]

That may be true for you, but I think that it misses the point of this thread

It's not NECESSARILY true for all people, because your skill levels can vary quite a bit from one form of poker to the next.

Regards
Brad S

[/ QUOTE ]

I agree. If a player could beat ring tables more than he could beat SnGs, he should be playing them. I dont think this is relevant though, if we want an accurate estimate of variance, we should assume identical ability in both.