I can really only do wild guesses without seeing the hand history. Maybe your HH language isn't English, maybe the game type isn't supported, you might even be trying to import tournament summary files instead of hand histories.

Please email an example file and we can clear this up in a minute.

Yes, thats the problem then, because i'm loading the tournament summary files.

But how/where do i find the right .txt to load the tournament like in your video?
Been searching allot, but can't find it.

Using HM3.

HM3 is probably moving the imported HH files to some archive location, I think that can be configured within HM3 but don't know the exact steps, sorry.

when setting up the bet sizes could you add an option "x times previous raise" so we can just type 3x/2.5x for example - this would make the setup process of trees much easier

also, great to see you have added the "prune action" option, could you please also add an "add action" - again this would make setting up trees much much easier

thanks!

one more small thing: just above the range grid where it shows actions and frequencies right now it shows the raises in chips while everything else is in bbs, so can we please have the option to show them in bbs instead? would also be nice to say after the call like calls x bb (y%) instead of just saying calls (y%)

imo in the betting setup the "add all-in" box should be optional for all streets, and not just all or nothing, since i think it would often only be neccesary for the river

The betting configuration UI is still to be rewritten within the next 1-2 months, we'll make sure there is an easy way to setup raises like that in the new version. In the meantime you could try to use something like "0.75pot", that's already supported.

We'll allow more options to edit the game tree after the initial creation, adding actions is a bit complicated right now though. We'd either need to query the user for the line configuration of the added subtree or keep the initial configuration options around, which is currently not the case. But that's also planned once the sizing UI rewrite is complete.

We'll change all action descriptions to big blind amounts to keep things consistent, and the call amount also makes sense I guess.

The all-in sizing: Not sure if we'd want three separate settings for this, adding all-ins for earlier streets is not a big deal in terms of tree size anyway. Maybe replace the current setting by an "Add all-in if the current stack-to-pot ratio is below x" option, do you think that would be sufficient?

Is there any way to change de colors on the range? The green to raise and the blue to call is driving me crazy....

Haha, the color scheme is just temporary. We'll have the option to customize the colors in one of the next 1-2 updates.

Thankssss and congratz for the beta version

Hello there!

Can you please explain / give an example of what is Table Equity and how it is calculated?

Thanks a lot!

Table equity is simply the sum of prize equity for all players on your table. For single table calculations that's equal to the sum of all remaining prizes (regular + bounties).

For MTT calculations it's more complicated, but with x tables remaining the table equity will be around 1/x'th of the total remaining prizes. Exact amount depends on the chip/bounty distribution though.

In any case, you can also display the $EV amounts in the tooltips of the hand grid / outline table.

Does all calcs methods uses card removal effects or only Monte Carlo?

The regular math mode only considers card removal from active ranges. Card bunching, ie removal from folded ranges, is only considered in Monte Carlo mode.

Do i need to set all players here to Monte Carlo to consider card bunching?



ie 8 players when using a 8handed table

No, bunching is always considered when using Monte Carlo mode, that's not related to this setting.

This max player setting controls how many players can actively enter the pot, once that limit is reached everyone else is forced to fold.

In most cases a player limit of 3-5 will be perfectly fine, only go higher if you think 6+-way pots actually occur with any significant frequency.

Even when i run "update selected tree" does it keeps bunching considered? Or do i need to run "update entire tree" to get full bunching considered?

still wondering why discord is not being used..... feels like this forum is like going back to windows 98

Yes, bunching is still considered when using subtree sampling.

Discord is still planned. We've put this off until later this summer though, prime vacation season didn't seem like the right time to launch this.

Can you get an idea of ​​why, running simulations that just changed this player limit (3 to 8), keeping all other settings the same and running 1000m samples (with set reset and default decay), generate significant differences?

Here are screenshots of Hand Settings and calculated strategies running "update entire tree", with reset checked, 1000m samples and decay "1E-6".

On the left, max active players=3 and on the right max active players=8.









I don't know if the number of samples can cause a difference like that, or I've done something wrong.

You can expect a direct effect of that setting, especially on the early position opening ranges. When using max=3, a simple raise/call/call is enough to reach the player limit and close the Preflop action. After that second call there is no option for a squeeze play, the blinds will be forced to fold.

You need to make sure that players aren't easily locked out of the pot like that. I think using a limit of 4 or 5 is more reasonable, especially when allowing two flat calls or limps.

Another aspect to keep in mind is that the max players=8 setting will result in a much larger game tree. The smaller max=3 tree will converge more quickly, you'll need to run additional sampling for the max=8 tree to reach a similar level of convergence. (Check the tree size during the setup, the # of samples you run should grow roughly proportional to the tree size.)

So if we are interested in say a BTN/SB/BB spot, leaving the max players = 3 makes sense. However, if we're interested in an EP strategy it makes sense to increase the max players limit to 4 or 5?

That's reasonable, although max 3 would still slightly affect the BU/SB/BB subgame due to card bunching. But even with max 3 the early opening ranges should be close enough for bunching purposes.

hi plexiq

i've been running a sim to see how our strategy reacts vs 3bets COvsBTN. When applying the RFI range for CO, i have not included comboes like like 67o, 85s, 63s etc. And as you see some comboes appear in the late gametree without me adding it to the RFI of CO.

So our counter strategy contains comboes that was not even included in the RFI range of CO. How is that possible? I've run the sim with 1000m samples.

Here is a quick pic:

https://ibb.co/gSxb3fr

The combos you mention are not in the node, they would be in bold / bordered otherwise. If you select the range display mode (second from left), they should be excluded there.

A non-folding strategy is still defined if these hands were to get here, that's what you see in the strategy display mode. ie IF the hands get there, this is how they should be played.

So, the solver will still find the strategy for hands, even if they don't reach the current node with the selected strategy. That's perfectly normal.