Richard said
So you’re asking about the chip EV making a 15BB from the CO with 95s.

I mean it’s not good, why did you think this would be a good idea, what was your reasoning for shoving?  

Do you have a calculation that can back up your assertion that “it’s not good”?

Chances are it’s probably not +EV, but there are some circumstances in which it might be, and it benefits you to imagine what those might be.

HRC has a Nash range for vmacc in this spot at 30.2%, with calling ranges for the three villains at 12.7%, 47.5% and 19.3%. At Nash, 95s is -0.49bb, suggesting we’ll have to go some distance with our exploitative reads in order to make it +EV. Let’s take a look at how tight the ranges have to get in order for this to happen.

First, let’s take the range about which we have the greatest confidence, which is the tightest of the three – the BTN calling range. If we tighten BTN up to calling with 8.3% (66+, ATs+, AJo+, KQs), which isn’t an unreasonable range for many players in these events, then our shoving range expands to 32.1%. 95s is -0.44bb.

Next, let’s tighten up the BB’s range, since that’s the next-tightest. If we put the BB on 14.9% (44+, A2s+, A9o+, KJs+, KQo, QJs), then our shoving range expands to 34.8%, and 95s is now -0.37bb.

Last, we’ll do the SB – this is going to make a big difference, since the SB was supposed to be calling it off really wide with their short stack, and this is what prevented us from shoving wider. If we narrow them up to 23.7% (22+, A2s+, A8o+, all broadways, T9s-K9s, 98s), which isn’t unreasonable given how tight players tend to play when calling off a short stack, then we can now shove 56.3%, including 95s! 95s is now +0.11bb, a reasonable margin for 15bb of risk, since our edge with that stack is likely pretty thin.

If we also tighten up the overcalling ranges for the spots where it goes three ways all-in, we can shove even wider – 68.3%, with 95s doubling in profit up to +0.22bb.

I’m not advocating we shove 95s in this spot – I think we’d need a high level of confidence about these reads in order to justify it. But I am advocating that we shy away from arbitrarily stating that “it’s not good” to shove in a spot like this, if we’re not able to produce some numbers to support our conclusions. Clearly, it is possible to produce a set of plausible, logical circumstances in which shoving 95s would be profitable in this spot, and thus the more important question becomes, how closely does the spot we’re actually in resemble that situation?

Richard said

Do you have a calculation that can back up your assertion that “it’s not good”?

 

Yes, which was the next logical step in a discussion about shoving ranges. I could if required do manual calculations but I would start with the snapshove range from the CO for 15 giving us, 30.6%. I would then look up the software’s calling ranges from each position to see how accurate they looked and tighten them up slightly if they looked looser than what I’d expect villains to call.

AFter that I would be unsure of how to proceed, I’d probably ignore all cases with more than 1 caller, and do something like

– P(no callers) = P(buttonCall*) x P(SBCall*) x P(BBCall*)

– EV(No callers) = 1,5 BB+ antes

– EV(button call) = equity x potsize – (1-equity) x potsize

– EV(SB call)….

– Total EV = EV(No callers) * P(no callers) + EV(Button call) * P(button call) …

without going into the P(SB call | Button fold) since I’m not too confident going into that math right now, probably just slap a little bit of a lower EV on everything after getting my end result. how would you calculate this?

 

Anyway my default right now if to add 10% to my shoving range so if I’m shoving 30.6% here I’d shove closer to 34%, I havn’t done much math on it it’s just a rough estimate of how I’d adjust on the fly.

If I came across as hostile when saying it probably wasn’t good then that wasn’t my intention, I wanted to hear OPs thought proccess before continuing  

This is fairly sound math in some senses, but it all seems totally unnecessary when HRC can do the math in much more specific terms for you, as I’ve outlined above. There are a number of flaws with the method you’ve chosen, in comparison to using HRC or ICMIZER.

Ignoring the instances with more than one caller is problematic, since that will definitely happen a non-zero frequency and affect the EV. As you can see in the calculation I ran, changing the overcall frequencies made a 12% difference to the overall exploitative shoving range we can adopt.

SnapShove is also inaccurate with its shoving and calling ranges, to the tune of usually about 0.5-1% in every spot. The reason for this is that it works with the assumption that everyone has the same stack size, instead of recognizing that the calling range of one player in a certain spot will depend on the stack sizes behind them, and on whether they’re able to call the shove and then fold to future action, or they’re forced to reshove for a much larger stack size (e.g. if the BTN wants to call your shove here, he has to reshove for ~40bb effective vs the BB, since he can’t call your shove without pricing himself into also calling a BB reshove).

Adding 10% to your shoving range is a very arbitrary way to approach these spots. There are some spots where it will result in you shoving way too wide, and some where you’ll be way too tight. I would encourage you to step away from reliance on SnapShove (it’s not a bad baseline for beginners, but if you can do EV math the way you seem like you can, it’s holding you back) and get accustomed to honing your push-fold instincts via HRC or ICMIZER. HRC is a more in-depth tool if you’re seeking accuracy.