Given recent events surrounding PokerStars, we thought it'd be a good time to have a look at the competition. In particular, we wanted to analyze effective rake (i.e. factoring in rakeback and reward programs).

In this first part of the series we're going to investigate several popular Sit & Go formats:

• Headsup Turbos
• Headsup Hypers
• 6-max Hypers
• Spin & Gos

at Stars (PokerStars/FullTilt), WPN (America's Cardroom, BlackChip Poker, etc.), iPoker, and the MPN (e.g. Betsson).
In future parts of the series, we'll look at other sites, formats, and cash games.


TL;DR

• Overall Stars is still best if you have at least 30% rakeback
• WPN (with a 27% rakeback deal) is close to Stars overall and considerably better than Stars at low-mid stakes if you don't have a high VIP status on Stars
• WPN is best for 6-max hypers up to the 100s
• iPoker is best for low stakes spins (≤ 10s) and nosebleed headsup hypers (above 1000s)
• MPN is not competitive. Generally, you'd need at least a 0.5% better win rate to get an ROI similar to the other sites

Methodology
We gathered the buyin data directly from the game clients in may 2016. For rakeback, we used the following numbers:

• Stars: 30% (new Supernova)
• WPN: 27%
• iPoker: 40%
• MPN: 30%

These rakeback deals are straightforward to get (i.e. no need to message anyone on 2p2, just sign up via any of several easy to find rakeback affiliates).

iPoker skins generally advertise 60%+ rakeback deals, but they use a "source-based" system which we don't model in detail. As an approximation, we use 40% rakeback which is close to the worst-case of 0.5 * 60% and seems to be what people get in practice. Any feedback on this point is appreciated!

All of the R code used to generate the tables displayed in this article is available on GitHub. We tried to make it as readable as possible, please have a look.


Headsup Turbos
We'll first explain the data, then present the data, and finally draw some conclusions.

In the tables below, pp is the prizepool contribution and rake is the pre-rakeback rake (as shown in most clients, buyin = pp + rake). effective rake is the post-rakeback rake, rake % is the pre-rakeback rake expressed as a percentage of the buyin, and effective % is the effective rake expressed as a percentage of the buyin.
breakeven is the winrate needed to break even; the ROI N% columns indicate the ROI achieved with the given winrate. These values assume that you're only paying the effective rake (i.e. you're Supernova on Stars or you have a rakeback deal on the other sites).









Notes:

• MPN isn't worth playing unless your winrate is at least 0.5% higher there than on other sites
• Both iPoker and WPN are better than Stars if you don't have a high VIP status
• iPoker has the lowest nosebleeds rake (1000s +)
• WPN has lower low and midstakes rake than iPoker (up to 300s)
• If you have at least 30% rakeback on Stars rake is still lowest there at all stakes below (1000s) and close to iPoker at 1000s+
• If you have a high VIP level on Stars, play turbos there, otherwise play on WPN
• If you're a losing or breakeven player, iPoker might be better than WPN (since you'll be able to realize more of your 60% rakeback deal)

Headsup Hypers
Data is the same as for turbos.









Notes:

• Even with no rakeback on Stars, low-stakes rake there is better than or close to both iPoker and MPN with a rakeback deal
• In particular, 7s are best on Stars
• MPN again isn't competitive
• iPoker again has the lowest nosebleeds rake and the highest stakes available
• Play WPN unless you have 30% rakeback on Stars
• If you do have 30% rakeback, Stars is best on all stakes it offers

6-max Hypers
pp, rake, effective rake, rake % and effective % are as for HUSNGs. There are three breakeven columns, indicating the ITM frequency needed to break even given that you headsup winrate is as indicated in the column header (e.g. the breakeven 50% column shows the ITM frequency needed to break even given that you finish first half of the time once you get headsup).

We assume the standard 65% / 35% payout structure.

The ROI columns show expected ROI for the given ITM frequency assuming a 50% headsup win frequency.









Notes:

• There's no point in playing on iPoker or MPN since even with a rakeback deal, rake is higher than Stars with no rakeback whatsoever (unless winrate improvements can make up for the difference)
• Rake on WPN is better than or equal to Stars at all stakes WPN offers (up to 100s).

Spin & Gos
The colums are as for the previous formats, execept that we add a pessimistic variant for most values. This gives the numbers assuming that multipliers with probability of < 1/10,000 aren't hit. Note that this is a fairly crude approximation (which penalizes some payout structures more than others) and subject to future improvements.

The ROI N% colums indicate expected ROI given an ITM frequency.







Notes:

• iPoker is best for lowstakes spins. Also the antes mean that the games will play less deep headsup, which is easier to learn
• Once you have rakeback, Stars is best at all stakes above 7s
• WPN is very close to Stars and offers the possibility to make deals in the highest multipliers
• It's probably optimal to start out on iPoker, then move to WPN and finally to Stars once you're rolled for 60s

Conclusion
So that's it for now. Stars still has the lowest effective rake overall (if you have a high VIP level) and is fairly competitive even without any rakeback in most cases.
For casual players, WPN is probably the best option (very close to Stars but no volume requirement for rakeback).
For low stakes spins, give iPoker a try.
MPN isn't interesting unless your winrate there is substantially better than on the other sites.

cheers!

This post was provided by Omni Poker

Great info, thanks very much for posting this!

it's not the same format at Spin & Gos
iPoker games are quicker so the rakeback is indeed not better

Thanks! Do let us know if there are any other sites we should look into (888 and party already done and coming soon)!

Future work includes modeling VIP-style rakeback programs and how different structures affect ROI, any ideas on how to do this properly (I don't think one can capture these effects in a single number like breakeven winrate, but graphing return per time unit might be good enough) are welcome

cheers!

Thank you for doing this!

MPN has 33% larger starting stacks for turbo and reg speed with same blind structure (starts at 100bb), so the possible ROI should be higher.

Does anyone have any numbers on traffic for different sites and when it's most traffic for HUSNGs?

I am currently developing a site which will do exactly this, will update when it goes live.

You need to account for leaderboards (ex. WPN's SnC). Also, any chance at WPN numbers with 35% rb?

WPN 35% rakeback:

PokerScout for SNGs would be super useful!

We're working on an app which let's you browse the data a little more interactively and set your own rakeback percentages etc.

This is the second part in our series on poker industry rake in 2016. The first part covers PokerStars, WPN, iPoker, and MPN, as well as our methodology (all source code and data is available on GitHub).

We're again covering popular sit & go formats:

• Headsup Turbos
• Headsup Hypers
• 6-max Hypers
• Spin & Gos

this time at 888, PartyPoker, Winamax, and Intertops (Cake/Revolution/Horizon).
In future parts of the series, we'll look at cash games.


TL;DR

• Don't play Sit & Gos on PartyPoker
• Probably don't play on Winamax either (except maybe 3-max Hypers)
• 888 is pretty bad, but gets slightly better at higher stakes (still need substantially better winrates though)
• Try Intertops, especially the (non-lottery) 3-max Hypers have low rake

Rakeback
We assume the following Rakeback deals, which are only available through VIP programs on 888, Party, and Winamax. You can get a flat rakeback deal on Intertops via an affiliate.

• 888: 27%
• Party: 21.2%
• Winamax: 22%
• Intertops: 36%

Headsup Turbos
In the tables below, pp is the prizepool contribution and rake is the pre-rakeback rake (as shown in most clients, buyin = pp + rake). effective rake is the post-rakeback rake, rake % is the pre-rakeback rake expressed as a percentage of the buyin, and effective % is the effective rake expressed as a percentage of the buyin.
breakeven is the winrate needed to break even; the ROI N% columns indicate the ROI achieved with the given winrate. These values assume that you're only paying the effective rake (i.e. you're a VIP or you have a rakeback deal).









Notes:

• party poker has the highest rake of any of the sites surveyed, Winamax second, 888 third
• At least party and 888 go down at high buyins (500s +), Winamax seems pretty unbeatable
• Intertops not fantastic but pretty reasonable (profitable with 52% winrate)
• Generally, none of the sites are better than those from part 1

Headsup Hypers
Data is the same as for turbos.









Notes:

• Party again really insane rake (almost no difference between turbos and hypers!)
• Winamax almost as bad
• 888 gets tolerable at 200s (still worse than Stars)
• Intertops is pretty good, especially at low stakes (and has the lowest 200s rake overall, including Stars etc.)

6-max Hypers
pp, rake, effective rake, rake % and effective % are as for HUSNGs. There are three breakeven columns, indicating the ITM frequency needed to break even given that you headsup winrate is as indicated in the column header (e.g. the breakeven 50% column shows the ITM frequency needed to break even given that you finish first half of the time once you get headsup).

We assume the standard 65% / 35% payout structure (except on 888, which uses 70% / 30%).

The ROI columns show expected ROI for the given ITM frequency assuming a 50% headsup win frequency.







Notes:

• Party again unplaybable...
• 888 also not good
• Intertops just as good as WPN from part 1 and hence competitive overall

Spin & Go's
The colums are as for the previous formats, execept that we add a pessimistic variant for most values. This gives the numbers assuming that multipliers with probability of < 1/10,000 aren't hit. Note that this is a fairly crude approximation (which penalizes some payout structures more than others) and subject to future improvements.

The ROI N% colums indicate expected ROI given an ITM frequency.

3-max Hypers are just like Spin & Gos, but without the lottery aspect (i.e. similar structure and winner-takes-all).







Notes:

• Winamax worst of all sites (one full ROI percentage point worse for a given winrate)
• Winamax 3-max hypers better than their spinsngs
• Intertops 3-max hypers rake lower than any other site's lottery SNGs

Conclusion
There's a reason why Party/888/Winamax are only popular with MTT players: SNG rake is extremely high, to the point of being unbeatable.
Intertops is quite good overall, and is one of the few sites both supported by PokerTracker and open to US players. Their 3-max hypers have good rake (compared to the competition's lottery SNG rake, which is always high). Spin players should consider playing these if they can get action (since they already have the skill set for them).
Winamax 3-max hypers are better than their spin equivalent.

cheers!

This post was provided by Omni Poker