Hey guys, this is a question for advanced SnG-pros 
I have just done a little push/fold training with the ICM trainer when I stumbled upon a hand that I found quite surprising. Unfortunately I didn't find a way to upload a screenshot of the spot so I will have to describe it here in detail:
Blind Level: 100/200/25
Stack distribution:
CO (Hero): 1190
BU: 4080
SB: 4040
BB: 4190
Hand dealt to Hero (in the CO):
K♣:T♦:
I opted for a push with KTo for 6 BB from the CO but was awarded with the infamous red cross. According to ICM trainer I should push 24.6% here: 22+, A2s+, A8+, K7s+, KJo+, Q8s+, QTo+, J8s+, JTo, T8s+, 98s, 87s
So KTo is right outside the Nash-pushing range according to ICM trainer in this situation. Ingame however, I am absolutely convinced that one should almost always push KTo in this spot because of the fact that Hero is shortstack and ICM does not take into account that Hero will hit the blinds in the next hand...
However, this is not the issue I would like to discuss (I am sure that most of you guys would agree that KTo must be a push here ingame due to the stack dynamic...)
What surprised me was the fact that according to ICM trainer we should push QTo and JTo but not KTo... I was really wondering how this can be explained so I pulled up Equilab and calculated a little bit:
According to ICM trainer our opponents' calling ranges against a push from us should be as follows:
BU: 7.8%
SB: 11.5%
BB: 28.2%
I compared the equity of KTo to the equity of QTo against each of these ranges:
Equity agains BU's 7.8%:
KTo: 32.95%
QTo: 31.06%
Equity against SB's 11.5%:
KTo: 34.01%
QTo: 33.09%
Equity against BB's 28.2%:
KTo: 46.68%
QTo: 42.24%
As we can see, if our shove gets called by any of our three opponents, our equity with KTo is better than our equity with QTo...
Furthermore, all three calling ranges contain more kings than queens, so due to card removal effects, our fold equity with KTo should be higher than with QTo which should be yet another argument to prefer a shove with KTo to a shove with QTo in this spot...
It is also due to card removal effects that in case if we fold KTo, our opponents will collide less often than if we fold QTo since their pushing- and calling-ranges against each other should also contain more kings than queens (does ICM take such things into account at all?)...
I thought about this a little bit and came to two different hypotheses how this mistery could perhaps be explained:
1. QTo performs slightly better against overcalling-ranges in this spot than KTo (ICM trainer does not show overcalling-ranges) so in the rare case that we get called by two (or even three) opponents instead of just one, we have a better equity with QTo than with KTo.
2. The calling ranges ICM trainer bases its calculations on are composed slightly differently than in Equilab (same percentage, different combination of cards (a few more aces here, a few more suited connectors there).
Do you think one of these explanations is adequate or do you have another idea why we should prefer QTo to KTo in this spot?
One can of course argue that this whole discussion is superflous nitpicking and knowing about such subtleties will earn us at most 10 cents over 10k games but I find it fascinating nonetheless - rather because of my general passion for the game than because of anything else...
I am looking forward to the ideas of you other poker-nerds on this spot... :f_drink::f_drink:
