So I was bored and made a quick spreadsheet on excel to see how badly regs hurt your profit, I wanted a quantitative answer after that thread we had about game selecting at the $16s.

You can get it here : http://www.megaupload.com/fr/?d=M0MU3RE7
If there is a better host than that, let me know (I'm an upload noob).
Also don't read the first part if you don't care about the math.


How the Calculation Works


We first calculate, for you and the other regs, what I called your "skill edge" in a game filled with randoms. That is, the ratio between your average prize won and their average prize won (disregarding any rake considerations, the prize is what you win before rake is taken into account). For example, 10% ROI leads to a skill egde of 1.20, meaning you win 1.2 times more than your fellow donk.

Then we consider a game where there are a given number of regs, and use these skill edge numbers in the following formula, that just states the game is zero sum :

Sum over all players [skill_edge(player i) * profit_of_a_random] = Prize Pool

An example with you, one regular, and 7 randoms in a 9-seater $16 turbo on stars, would yield :

(your_skill_edge + other_reg_skill_edge + 7) * profit_of_a_random = 135

From there we calculate the average prize of a random, then deduce yours, multiplicating by your skill edge. And then we get your ROI. Easy game.


Notes about using the spreadsheet


You can toy around as much as you want. Just be careful that you have to input your ROI assuming there are only randoms in the game. So you should be somewhat generous wth yourself (and for the other regs as well), because if you play SNGs with an average of 2 regs at your table with a 7% ROI, well, your ROI in an ideal game is significantly higher. Up to you to work out the guesstimate.


The interesting (and scary) results


Here's a quick summary if you want to just keep a couple numbers in mind. These numbers are pretty insensitive to your ROI, what really matters is the number of players in the game. See how a player just above the average donk can already hurt. Oh and it's pretty linear, the cost of two regs is about exactly two times the cost of a single one as long as the table isn't full of them.


Summary, for 9-seater SNGs
Cost of a good regular (12% ROI) ________ -2,80%
Cost of an average regular (6% ROI) ________ -1,80%
Cost of a barely winning player (2% ROI)_____ -1,20%

Summary, for 6-seater SNGs
Cost of a good regular (12% ROI) ________ -4,50%
Cost of an average regular (6% ROI) _________-2,80%
Cost of a barely winning player (2% ROI) _____ -2,00%


Cliff notes


Regs are super costly to your profit. Game select more, and stay out of my games. I'll do the same with yours. kthx.

Congrats. And nice post I hope it helps and guys will try to atleast make some effort to table select

nice and congratz, but the i doubt the problem is within 2p2 members so reachin out to those who dont read the forum will be a bit hard. But if they did know about it, that would only mean the read the forum/are gettin better/games get harder/we dont want that...
just accomodate to a changing enviroment, and move on, and get better.

good post, ty!

I have to think more about it, because it sound too scary to me!

I've been thinking about game selection in lower volume situations. (Such as regular speed comps on FullTilt)

You've got a model whereby there's 15 regs with 10% roi wanting to 4-table.
There is the traffic for 30 comps to be starting per hour, and the reg lasts an average time of 1 hour in each one.

If you are the 16th reg, would you want them all to be practising perfect game selection, and have every comp start with exactly two of the 10% players.
Or would you want a situation whereby none of them practised game selection, and they all joined the first available comp when they were ready.

In the latter example, some games will have less than two regs (some with none), and some will have more than two regs (some with seven).

Assuming you are happy enough to wait for the good comps with no or only one other reg, and avoid the nit-fests, it perhaps wouldn't be in my best interests for the other regs to all start practising rigourous game selection?

An over simplified model I know, but it got me thinking whether or not other people practising game selection is always a good thing for you?

ty 4 the post. what do you guys recommend for game selection? I have PT3, do any of you guys use the game selector and if so how do you like it?

real men don't game select.

Game selection is the difference between a winning player and the winningest player imo.

If I already see 4 regs ready to multitable I'm sometimes waiting out 10 mins to take the next train, but if I'm first to the tables I just take it as it comes. Then again, with my current roi in $27s people are probably cueing up to play me... :-D

Thing is I can either be fancy and want to have a 10% ROI over 2k hands as I'm waiting to game select, or I can have a 6% ROI over 4k hands as I'm not sat around waiting to game select. I'm not fussed about how impressive my ROI is it's my hourly that's most important.

Good post though

Do you guys just make notes on the players who you've seen play and look in the lobby to see who you're playing against?

I don't I just start up ft shortcuts and get it to continously reg me for 15 regardless of who's playing.
But that's part of the reason why I've only got 7% ROI.

I don't get why people are anti-game select. The one sng on stars where you can game select the best is the 16s. You wait, what, 6 minutes for other games to fill? Before I switched to 16s, my wait time at 20 and 30s was seriously at least 40 minutes for the set to be completely up and running. 16 load time is a blessing in disguise.

i try to find out the statistics of the grinders (regs who are the only ones, that registered on many SnGs ---> u know ... 1/9 (quasifiction) etc. ^^), cuz they are almost always good players, who have a decent roi.
then i make notes on them --> write their sharkscope statistic in the notebox.
And when they are registered, you could avoid registering yourself.
But i dont want to wait. I want to play, so i register too and avoid game-selection. I'm playing for fun, not for the money.

Btw. hi, Glitlr

FT shortcuts makes game selecting very easy Chillin'. Just use the sharklist feature. Of course you have to actually add names to it so it won't register you in those games with X number of sharks, etc. I currently have mine set to not register me if there are more than two people from my list, but the list probably isn't big enough yet to make much of a difference.

hooray for gameselection

Wanted to load a set yesterday and saw 4 winning players (+5-10% ROI) regged for a 22$ at FTP. It was only the four of them, so its not like they were after a bunch of fish or something. Longy was one of them and unregistered 15 seconds later - smart choice imo.
Dunno how it is on stars, but even the 11s on FTP are regfests like 60% of the time, eventhough some of the regs aren't DDBeasts, they still hurt your winrate.

It should be pretty interesting to see if it's really worth waiting some regs finish launching their set (bc the time you don't play you loose $)
Since I play continuousely I can't pay really attention to tableselecting when my 20 tables are launched but I always try not to begin my starting set with others 3-4 regs, even if most of 16's regs are not that good having 8 others nit at the table is never good (and theses days I met a ton of 9/6 type tables)
Anyway nice post (pretty scary numbers tho) and congratz on 500th

I think that pretty much sums it up.

Recently I started game selecting religiously in the $16s. My number of SNGs/hour has dropped from like 18 to 16 only (12 tabling continuously), and you wouldn't believe how softer these games are, and my ROI shows it even if I'm running hot as well. I think it outweighs the loss of game/hour by far imo.

Actually I'm not even sure it would be +EV, here's the reason : when regs stack up like 4 or 5 per SNG for a couple of minutes, well, they aren't in the next game anymore. Makes the game selection process easier, assuming you're not playing in sets obv.

Ha
Profit is for pussies obv.

yeah obv your ROI is gonna be higher but what about your $/hr ?
But well if you "only" drop from 2 tables/hr to get a slightly better ROI (how much did he increased ? 2-3%?) then for sure it should be more profitable I guess.

Lol cheers Jurass, I appear to be one of the few regs attempting to exercise table selection at the 22's.

I must admit to getting pretty frustrated with some of the regs who seem completely oblivious to this and are joining games which can barely be proftiable for anyone. Therefore forcing me to make a good decision for everyones benefit.

My rule is 3 winning regs and i unregister from the sng. Sometimes it takes longer to get the games moving but i am playing continuously so I can pick and choose.

Very interesting.
In higher stakes games a 6% ROI is really good. Does this mean just 3 other regs will make the game unprofitable even before rake?

I think part of the reason why 6% ROI is really good at high stakes is that those 3 players were already there anyway. Add another three regs for a total of 6 and yes, I can see those games being unprofitable.

Does that make any sense? I run bad at phrasing things.

Ah, true. The calc is before regs.

i am with you on gameselection - if i see 3+ names already sitting at 3 or more tables i just load behind - at the 16s games load so fast so you can afford to wait 5 mins imo - but if you load behind on the other hand some regs jdk about gameselection and sometimes you get 5+ of those instead of the 3 you would have had before - conclusion - i dont know :-)