Open Side Menu Go to the Top
Register
Construction of Push/Call Ranges Construction of Push/Call Ranges

07-01-2019 , 01:56 PM
Hey there,

In my efforts to improve my game I spent some time on my pushing / calling ranges and, obviously, found various charts online, claiming to provide "optimal" ranges for whatever scenario.

In this time, I found that basically all ranges, regardless of the source, deviate from what PokerStove is recommending for a given percentage. The ranges from PokerStove regularly contain less PPs and less Ax, Kx hands but more broadway hands and high one-gappers etc.

Despite some strange ranges in general (am I really supposed to call with 44% of my range in the BB with a 6BB stack when UTG shoves 9-handed??) I am wondering:

-Why do these ranges differ? How are these "optimal" ranges constructed?
-Isn't Stove giving me the hands with the highest raw equity and therefore the correct ones?
-Are these (sometimes very marginal) deviations even relevant or am I overthinking it?

Thanks in advance
skeppi
Construction of Push/Call Ranges Quote
07-01-2019 , 08:42 PM
Just buy snapshove.
Construction of Push/Call Ranges Quote
07-01-2019 , 10:00 PM
Hm. All they did was packaging the ranges that are already out there for free in a small web interface I'd have to pay for. I can create something similar in my spare time without giving them my money.

Thanks for the link nevertheless. The Reshove tool might be worth a try.

But now I still don't know where these ranges come from...
Construction of Push/Call Ranges Quote
07-01-2019 , 10:49 PM
The ranges came from a computer. If we assume we're just playing cEv then we take our pot odds, their expected range (or rather what an app says they should shove) and what % of hands we need to call with to have proper equity to beat the pot odds.

So lets say a villain is shoving top 25% and we have 2-1 on our money we plug into pokerstove a range that gives us just above 35% or so. That's about it honestly. Then if we feel they're tighter we drop some of the bottom off the range etc.
Construction of Push/Call Ranges Quote
07-02-2019 , 07:13 AM
Yeah, I get that concept. My issue lies with how they are being constructed. My assumptions were:

1: Playability, equity realization etc. is not relevant so the ranges solely consist of the hands with the highest raw equity
2: PokerStove gives me exactly these hands for the percentage I put in it
3: Suggestions and charts that can be found online use these ranges

Instead the ranges differ:
1: Charts contain more PPs, Ax, Kx hands. Makes sense in terms of raw equity, but with what parameters does Stove work then?
2: The charts sometimes heavily favor suited connectors and one-gappers (even more than Stove)
3: Playability seems to be relevant? E.g.: SnapShove gives the following range for a 25BB shove UTG:

TT+, ATs+ A5s A4s A3s, AQo+, KJs+

Why the f are A5s to A3s in there? They have more playability but they lose to every higher A and this is a shoving chart...

I'm confused...
And I'm not intending to get better at the game just by accepting some random ranges I find online without understanding the logic behind them... :P
Construction of Push/Call Ranges Quote
07-02-2019 , 04:58 PM
They're in there to give you some balance for your jam range so it isn't just nutted and you block their ace hands like AK and aces etc
Construction of Push/Call Ranges Quote
07-08-2019 , 02:58 PM
Quote:
Originally Posted by skeppi
Yeah, I get that concept. My issue lies with how they are being constructed.
How would you construct a push / fold range for SB vs BB for a certain stack size? Let's assume you are far from the money and the BB always has you covered.

For an easy example, suppose you are in the SB and you know the BB will only call your pushes with AA or KK? What percentage of hands, with a stack of 10BB, can you profitably push with?
Construction of Push/Call Ranges Quote
07-09-2019 , 12:12 PM
Assuming 50/100 blinds with 12.5% antes and 9 players:

Hero range: 100%
Villain range: KK+ = 0.905%

Pot: 262,5

EV if he folds: 0.99095 * 262.5 = 260.12
EV if he calls: 0.00905 * [(1000 + 262.5) * 0,162 - 950 * 0.838] = -5.35

Total EV by shoving 100%: 254.77 = 2.55BB per shove so I shove any2 anytime

Last edited by skeppi; 07-09-2019 at 12:20 PM.
Construction of Push/Call Ranges Quote
07-09-2019 , 06:13 PM
Now start adding hands to BB's range until it is no longer +EV to push 100% of your hands at 10BBs, and you will start to get an idea of how these push charts are created.
Construction of Push/Call Ranges Quote
07-10-2019 , 06:15 AM
Thanks, but that wasn't my question. Please read my previous posts.

When I started spending some time with these charts, I assumed that they were, like ICM, based on some "unexploitable" calculation. But the more time I spent on them, the more I learned (and the more I got frustrated).

I can find charts everywhere. E.g. D+B, Snapshove, the new Harrington book I've just read etc. and they aren't congruent:

1) The percentages differ (often just slightly)
2) The construction of the percentages (!) differs in terms of what hands to push with an x% pushing range (see below)

So while charts can be found everywhere, their calculation basis can be found nowhere. And continuing my thoughts from my previous posts:

1) Shouldn't these charts include assumptions on the looseness of our opponents? If not, how are they calculated independently from the opponents? If yes, what are the assumptions and how do they end up with (mostly) similar percentages? Are there assumptions for a "standard" opponent?

2) How are the constructed??? See the image below and these are just the first two examples that I randomly picked. And please don't tell me that it doesn't make a difference whether I push A3s UTG with 25BB (which I would never do I think) or 99.

Comparison D+B Poker calling range with Stove:


Comparison SnapShove Pushing Range with Stove:


PS: I don't think Snapshove uses any kind of algorithm for its calculations or their algorithm is flawed as I found various smaller inconsistencies. They only rammed a whole lot of ranges into a database. Example: It gives a wider shoving range for a 25BB stack than for a 24BB stack but that's not the point.
Construction of Push/Call Ranges Quote
07-11-2019 , 03:31 PM
The difference in the charts may be due to a couple of things:

1. Sample size used in the calculations. Some calculations may use a larger sample size and be a little more precise.

2. Taking effective stack sizes and players remaining in the hand into consideration. This appears to the case in image #2. It looks like pokerstove is calculating their hand ranking vs a single opponent, where SnapShove is taking stack size and players remaining in the hand into consideration.
Construction of Push/Call Ranges Quote
07-11-2019 , 06:08 PM
Quote:
When I started spending some time with these charts, I assumed that they were, like ICM, based on some "unexploitable" calculation.
The mathematical term for this is called "equilibrium." It can be approximated.

Quote:
1) Shouldn't these charts include assumptions on the looseness of our opponents? If not, how are they calculated independently from the opponents? If yes, what are the assumptions and how do they end up with (mostly) similar percentages? Are there assumptions for a "standard" opponent?
The equilibrium solution does not make an assumption about opponents. If you play near equilibrium but your opponent plays too loose or too tight, you still make chips. If you play near equilibrium, it is harder for opponents to exploit you. This does not mean that there are not opponents (who do not play near equilibrium) that you can exploit by deviating from equilibrium play.
Construction of Push/Call Ranges Quote
07-11-2019 , 06:42 PM
Just get the ICMizer pro license and spend as much time as you can with their SNG Coach.
Construction of Push/Call Ranges Quote

      
m