Hello all,

I'm a little bit confused as to how the EV line and SD+NSD lines interact. Looking at my PT4 graph and running it through filters to isolate 3 handed and HU all in situations I get some worrisome results, but ones that also seem unrealistic. Below are pics which will illustrate my point. These are taken from a recent ~2k sample, which I know is statistically insignificant but it is what it is.



Here is a HU graph and the filters which I activated are (Raised All in Pre) and (Players Dealt Into Hand Between 2 and 2). So if my understanding is correct, the non-SD winnings being so high under these circumstances makes sense since the majority of shoves go uncalled and we just pick up the blinds, or opponent will min-raise and I jam, he folds, again picking up the pot without SD. The SD winnings being so low feels strange? But I suppose can be explained by negative variance in which case okay whatever.


Now here's the same graph but with the addition of (Hands Went to Showdown) filter




So as you can see the nose dive of the EV line is quite worrisome but more importantly very puzzling. Again, if I understand correctly, the EV line shows our adjusted chip expectation taking All-In equity into account. So apparently in situations where get cards get turned up (i.e. there's a showdown) I'm getting it in on average super horribly.


My question is: am I missing or misunderstanding something? I can well imagine the green line (actual chips won) being low, that can just be explained by negative variance. But the EV line being so low seems completely unrealistic; I have access and routinely study/go over GTO as well as readless preflop charts which contain variations for 3 handed spins play as well as HU. I'm sure that I've jammed some hands that I technically wasn't supposed to or called when I wasn't supposed to, but it's not like I'm open shoving 43s at 15bb on average and then wondering why I'm seeing such results. To have such low EV expectation just feels weird.


Any help would be greatly appreciated.

Hi, it seems like you are overprotecting your good and medium value hands, which results in villains folding their weaker hands, allowing you to increase your red line. However, when facing stronger ranges that do not fold, your showdown blue line decreases. The same is true when running too many bluffs in bad spots; while it increases your red line, your showdown line goes down when called.

Hey, thanks for the response. I think I understand what you're trying to get at but it sounds as though you are referring to postflop play, but keep in mind this is filtered for situations in which we are all in preflop.

I'm just concerned that the EV line is on such a downward trajectory and it just doesn't seem appropriate. Yes, once we get short stacked- especially below 10bb -a lot of our strategy revolves around Push/Fold (of course there's still limping and playing post some, but generally we will be all-in preflop a good amount of the time) so we can't expect to have a huge equity advantage on average as both we and opponents will go with many speculative hands.

But if we take the EV line at face value it can mean a few possible things:

1. I am getting crushed by card distribution. Which would be fine.
2. I am shoving/calling too many hands that should be folds. Highly unlikely, since as I've mentioned I put a concentrated effort into studying and implementing preflop ranges, and also attempt to adjust those to the various opponents I face.
3. The EV graph actually should be this way and I'm just not understanding something

For more reference here's a graph with the same parameters but filtered for 3 handed instead of HU:





And here's the same graph but filtered for showdown:





I also added the average all-in equity stat to the info box and you can see it's ~49%. My initial post with the HU graphs lacks the stat but looking at it now its ~47%.

Is the roughly 2 percent difference enough to explain the disparity between the graphs? In the 3 handed category we at least have a positive expectation albeit not a great one, but nonetheless. Whereas looking at the heads up graph you'd think we were open shoving hands like 72o only as standard, which obviously I know that I am not doing haha.

You jam hands that should be limp/calls denying your opponents chances to push. When they call they have stronger range because of this. Used to struggle with this a lot when playing too many tables. You really shouldn't be playing mainly push fold around 10bbs. Threshold for that is 5bbs. Like I mentioned you should be doing a LOT of limp/calling instead of shoving around 8-15bbs.

Don't know what stakes you are playing and oc this depends also a lot about the pool you play in but in general.

E try filtering out pre flop allin true that should take the iso allins also. Imo raised pre allin is just you pushing.

Lot of BS in this thread lol.

This is a lot simpler then you're making it out to be. Think of it from another angle: You go all-in, and villain is deciding whether to call or fold. Assuming they are reasonably sensible, they will be trying to call with all hands that have >X% equity, where X depends on stacks and is <50% but usually >40%. So your opponent only calls hands with close to 50% OR BETTER versus your open shove range. So yeah, makes sense that your open-shove range, when called, will face only hands with at least X% equity and therefore only have so much equity itself.

Another important thing to think about is that you're not open-shoving AA,KK, etc

All-in all this is basically just a complete waste of your time and there are wayyyy more interesting and useful things to study/consider/research whatever.

Coffeeyay!!!!


So glad you responded to this thread, I was actually thinking of messaging you directly if I didn't get a reasonable response from someone on here. I realize that this is not something which merits a big time/effort investment. I suppose I was tinkering around with this since it feels as though I've been running somewhat badly especially over the second half of this sample. I'm sure much of that is mental bias, but wanted to look at the data and see what I can find.

Incidentally, the preflop charts which I referred to are the ones that come with your Spin n Go pack. Ever since I started seriously using the pack (and studying the charts) I feel that my game improved noticeably, so thank you first of all.

I'll continue doing what I'm doing, seeing as based on your comment I can imply that it's unlikely I'm leaking a lot in these situations. And yes you're correct about the top of our range being absent since we are predominantly trapping with it.

I guess I was just surprised to see the EV line on such a trajectory when, given everything you wrote and what I already suspected, I would have assumed it should should hover around breakeven.

You mentioning that this is mostly a waste of time already answers the question somewhat, but I can't resist asking: is the graph being this way ultimately not something to worry about?

The EV of open shoving comes from a combination of chips won from folds and equity when called. If you only look at half of the equation, it's necessarily going to look smaller than the whole.

My point was to consider what you are looking for and whether you have realistic expectations. Why should, or would, your chipEV line when the all-in is called look break-even or winning? Why is that relevant? Like, if your opponent is calling only TT-AA vs your all-in then open-shoving 54s is incredibly good 12bb deep--but obviously the small % of the time you're called it's by TT-AA and your chipEV will look bad

I'd recommend focusing in on individual parts of your range and seeing how they perform rather than getting lost in trying to look at performance of whole ranges. A good way to use your database is to look for instance at the bottom of your VPIP range and make sure it's > folding. That can give added confidence and understanding of what's going on. Another example is filtering for your 3b nai bluffing range and seeing if it's doing better than folding. A bad example is filtering for your total 3b range and getting confused because it's a combination of premium and bluffs and the total filter doesn't make that distinction and will obscure what's going on.

I feel a little silly for making the wrong assumption initially, but I think I understand it now. My naive thought process at first was like "well we're getting it in generally with a wide range and our opponents will be calling with a relatively wide range also so shouldn't our expectation be better than what the graph shows?" Which is why I assumed it should be breakeven-ish.

But now I realize I evaluated the relative ranges completely wrong. Our open-shove range is going to first of all be wider than opponents calling range and also it will be capped (i.e. missing the top of the range) whereas our opponents calling range is completely uncapped, AND the weakest portion of their calling range will be on average stronger than the bottom portion of our open-shove range.

That is all to say that, when called, we can expect to face a stronger range than what our open-shove range is - therefore our EV should actually look the way that it does. But as you mentioned, and as the comprehensive graph I have up shows, we more than make up for it given our fold equity pre and the amount of chips we pick up routinely when our opponents fold.

Thanks for clearing that up for me!

You got it! yay!