Why ROI?

[cagedman]

Was reading another post and it had me thinking...

Caveat: this post assumes that skill, not bankroll limits the number concurrent games that can be played. This is a fair assumption since bankroll requirements don't change much when you multitable (only so much as your std deviation changes with multitable play)

I'm kind of new to this forum.. but for the record...

ROI is irrelevant! Why is this discussed?

The only sensible metrics IMO are aggregate hourly rate of your play, at the limits you play, at the number of tables you play concurrently and the standard deviation of that rate per tournament (or better yet per hour).

What is this obsession with ROI? At least use $/tourney which is essentially the same statistic but removes that level of obscurity.

[AleoMagus]

The best explanation that I can give for this is just that many SNGs play very similarly. 10+1, 20+2, 30+3, for example. This is similar to the way in which Limit hold'em can have a similar style of game at many different stakes. You may have noticed that those guys like to talk about BB/hr for the same reasons.

Still, I get your point, but it might be too much trouble if everyone started saying things like 'I get a $12.20/hr return multitabling 2 at a time 20+2 SNGs which average 44 minutes each' in every SNG conversation on here.

Regards
Brad S

[Kopefire]

ROI is a valuable statistic because it's relevant.

you are investing x number of dollars for a possible return. The rate of return more or less defines if you're doing well or not.

[cagedman]

I can see A.M.'s point:

Provided ROI percentages at various levels equate to comparable skill vis-a-vis level being played, it allows you to give a rough estimate of 'how you are doing' without specifying the buyin of the game itself (although the buyin is usually given alongside this figure)

I'm not in agreement with Kopefire's response though...

[ QUOTE ]

The rate of return more or less defines if you're doing well or not.


[/ QUOTE ]

Of course it does.. as does any metric that allows you to backsolve for your actual profit. read: bottom line per time/effort unit.

[ QUOTE ]

you are investing x number of dollars for a possible return.


[/ QUOTE ]

How much you invest in order to obtain a fixed return is irrelevant, insofar as your capability to earn the return you like is not limited by your capital.

For instance (I hate to use analogies because it leads to all kinds of non-sequitirs and off topic convos): Considering all things exactly equal, you would happily play a buyin twice as big with 3/4 of the ROI because you'd simply make more in the same unit of time. The only exception to this is if you couldn't -afford- to play this game (which is where my caveat comes into play).


In other words, I accept A.M's explaination that ROI is a way to recenter win rate against buyin in order to provide an indicator of -skill- (provided ROIs across levels are compparable). I reject the idea of ROI as the best means of reporting 'earn rate'.

I guess what this boils down to is reporting of level success vs reporting of actual earnings.

[Kopefire]

But see .. no one cares about 'actual earnings' except the IRS /images/graemlins/smile.gif

When posting here, lots of people want to know things like "am i playing to tight?" or "am I not aggressive enough short-handed?"

When talking to ring players, you can get a lot of information about their ability by looking at their BB/hr rate -- and that information is transferable across limits. It's relevant information about their performance.

Likewise with ROI. We're not hear to compare our bankrolls to each other. I don't care if you're earning $5 an hour if they way you're doing it is making .001 ROI on $5000 SnG's. Yippee that you have a big bankroll . . but frankly you can't teach me much about poker if that's the best you can do.

Likewise, if someone is earning 45% ROI long term . . I don't care if they only play $5+.50 tables .. there is probably something I can learn from them.

ROI in the SnG world conveys something about the mastery of the tables over time. That information conveys what people here are interested in conveying with respect to poker play in a nice, simple number.

Now, when we're talking about who's buying the beer for the next get together.. then $/hr is important. But that isn't what we're doing here.

[rachelwxm]

Since I am not going to give up my full time job for poker anytime soon, I think ROI is more important to me. /images/graemlins/smile.gif

[PrayingMantis]

The "unit" of game we discuss here is one SNG. In some sense, it is equivalent to a hand in a ring game. Not quite the same, but pretty much, since in a ring game most of the information you need to know about the relevant $EV can be found in the specific (posted) hand (not all of it, of course), while in an SNG, as in any tournament, big portion of your $EV, with regard to decisions you make, is a function of how you play a long string of hands.

For instance: the _way_ you play a whole bubble situation, is more important than how you play a _specific_ bubble hand. This is a bit strange, but it's true, although the _way_ is built out of many specific hands. In other words: _strategy_ is a key word in SNGs, as opposed to ring game. And you ROI is the best indicator for how good (or how close to "optimal") you play an SNG, over-all, or how good is your strategy.

Surely, in many cases you are willing to give up some of your ROI, for the ability to play more games simultaneously, or for a chance to play shorter games. And therefore, ROI is not the only thing that matters for $/H, obviously. But even if you are willing to give up some ROI for a better win rate, it is important for you to know how much ROI you're giving up, and if there's a way to keep your ROI still higher.

In any case, if we discuss SNG play, with the idea of "optimal play" in mind, i.e, gatting as much out of every game we play (And it doesn't matter if you are multitabling or not), ROI is definitely the best way to approach it.

[cagedman]

Fair enough.

Looks like it boils down to the dichotomy of interest in performance vs. interest in earnings.

I think the post I read right before making this one caused me to assume ROI was used as a way to discuss earn rate rather than a way to discuss relative performance.

I agree it's a sensible metric to measure relative performance; I still don't like the term though.. I still think 'return on investment' is still misleading in the sense that it implicitly refers to a line of thinking that doesn't truly apply here... (Outside of this forum, ROI is usually discussed in making allocation decisions when capital is the 'scarcity' [in a relative sense])

That's just more of a nitpick at this point though... I would call it PBR (Profit to Buyin Ratio) :-). [That and pabst blue ribbon is my beer of choice]