More on ROI

[NotReady]

I did some fiddling with ROI, Win Rate and Standard Deviation with the goal of trying to evaluate ROI in terms of BBs per hour. The idea is to compare SnG results to limit results.

With a given ROI, SD ( I used $17 since that's what I get for $10+1, I'm not sure if it would be higher or lower for higher levels ) and RoR(I used .1 % because I hate to gamble). These numbers produce a dollar amount, say $300, as the RoR bankroll. Assume 300 BBs is a good BR for limit holdem, then $300 would equal a BR for a .5/1 game.

I got the following results :
ROI --- BBs Per Hour
16.7 --- 1
23.5 --- 2
29.0 --- 3
33.5 --- 4
37.5 --- 5
40 --- 5.785

Another way to state this is : If you are a 1 BB winner at a given limit with a 300 BB BR, you would need a 16.7 ROI to achive the same hourly WR.

This doesn't take into account the fact that SnG's only require 30-45 minutes to play, so your hourly would be somewhat higher than 1 BB.

I don't know if this is a valid comparison, caveat emptor, just trying to evaluate ROI and compare it to limit results. Not even sure any comparison is valid.

If it is even close to valid it may indicate why a 40% ROI may not be sustainable in the long run.

[LinusKS]

[ QUOTE ]
I did some fiddling with ROI, Win Rate and Standard Deviation with the goal of trying to evaluate ROI in terms of BBs per hour. The idea is to compare SnG results to limit results.

With a given ROI, SD ( I used $17 since that's what I get for $10+1, I'm not sure if it would be higher or lower for higher levels ) and RoR(I used .1 % because I hate to gamble). These numbers produce a dollar amount, say $300, as the RoR bankroll. Assume 300 BBs is a good BR for limit holdem, then $300 would equal a BR for a .5/1 game.

I got the following results :
ROI --- BBs Per Hour
16.7 --- 1
23.5 --- 2
29.0 --- 3
33.5 --- 4
37.5 --- 5
40 --- 5.785

Another way to state this is : If you are a 1 BB winner at a given limit with a 300 BB BR, you would need a 16.7 ROI to achive the same hourly WR.

This doesn't take into account the fact that SnG's only require 30-45 minutes to play, so your hourly would be somewhat higher than 1 BB.

I don't know if this is a valid comparison, caveat emptor, just trying to evaluate ROI and compare it to limit results. Not even sure any comparison is valid.

If it is even close to valid it may indicate why a 40% ROI may not be sustainable in the long run.

[/ QUOTE ]

Very interesting post. If I'm following you, and if your math is right, that would mean 100 sng's were roughly - very roughly - equivalent to 10,000 hands of limit.

Anybody know how many limit hands you need to get a good idea of your bb/hr?

[NotReady]

I think 1 SnG is equal to about 1/2 hour of limit play, figure about 60 hands per hour, so 100 SnG's would be about 50 hours, or 3000 hands. All very ball park.

True win rate is very elusive, and can take literally thousands of hours. It's possible for a good player to have 500-1000 hours playing behind, so that would translate to many, many SnG's. I think under normal circumstances, a decent win rate after 500 hours coupled with a solid understanding of poker theory, would indicate someone is at least a decent winner at a given limit.

One problem with all of this is, by the time you've played enough to get a fix on win rate, the games will have changed and the player will also have changed, usually for the better, but not necessarily. One of the things that keeps poker alive.

[dethgrind]

This is exactly the discussion I've been looking for.

[ QUOTE ]
ROI --- BBs Per Hour
16.7 --- 1

[/ QUOTE ]

Please explain your math here. The way I see it, if you're getting a return of 16.7% on an $11 investment, that's $1.84. Is this assuming something other than 10+1 and .5/1?

[ QUOTE ]
With a given ROI, SD ( I used $17 since that's what I get for $10+1, I'm not sure if it would be higher or lower for higher levels ) and RoR(I used .1 % because I hate to gamble). These numbers produce a dollar amount, say $300, as the RoR bankroll. Assume 300 BBs is a good BR for limit holdem, then $300 would equal a BR for a .5/1 game.

[/ QUOTE ]
I'm not sure I follow this exactly. Did you make up those bankroll figures? I've been playing with simulation stuff (http://forumserver.twoplustwo.com/showthreaded.php?Cat=&Board=singletable&Number=921 198&Forum=All_Forums&Words=&Searchpage=0&Limit=25& Main=914767&Search=true&where=bodysub&Name=10596&d aterange=1&newerval=1&newertype=w&olderval=&oldert ype=&bodyprev=#Post921198) and could produce more precise figures for SNG bankrolls. From what I understand, a normal approximation works fine for limit ring games.

Correct me if I'm wrong, but is the question you're asking this: given a bankroll and a RoR tolerance, which game is more profitable? That's the question I'm asking, anyway.

[LinusKS]

[ QUOTE ]
I think 1 SnG is equal to about 1/2 hour of limit play, figure about 60 hands per hour, so 100 SnG's would be about 50 hours, or 3000 hands. All very ball park.

True win rate is very elusive, and can take literally thousands of hours. It's possible for a good player to have 500-1000 hours playing behind, so that would translate to many, many SnG's. I think under normal circumstances, a decent win rate after 500 hours coupled with a solid understanding of poker theory, would indicate someone is at least a decent winner at a given limit.

One problem with all of this is, by the time you've played enough to get a fix on win rate, the games will have changed and the player will also have changed, usually for the better, but not necessarily. One of the things that keeps poker alive.

[/ QUOTE ]

Alright.

So you're saying 1000 sng's would be a minimum for estimating long-term ROI.

[NotReady]

I'm just guessing, but in limit poker terms, I think 1000 would give you a solid indication that you can win at that level. I don't know how accurate your ROI would be.