below are some results (bold # = $ ev in a 200nl game) from simulations i ran on the donk/3-betting line:
bb donks,
|
| btn raises {>=tp, fd, oc+gs, Var2% others} |
| bb 3bets all-in w/ {>=tp, fd, oesd, pr+gs, Var3% others} |
| btn calls w/ {>=tp, fd, oesd, pr+gs} |
|
| Var2 % |
|
|
|
| Q94r |
| 10 | 20 | 30 | 40 |
| Var3 % | 30 | 2.1 | 1.61 | 1.13 | 0.65 |
| 40 | 3.88 | 3.32 | 2.77 | 2.21 |
| 50 | 5.58 | 4.91 | 4.23 | 3.55 |
| Th98h |
| 10 | 20 | 30 | 40 |
| Var3 % | 30 | 9.35 | 9.72 | 10.09 | 10.46 |
| 40 | 9.86 | 10.39 | 10.93 | 11.47 |
| 50 | 11.06 | 11.7 | 12.34 | 12.98 |
|
| Var2 % |
|
|
|
| 5h43h |
| 10 | 20 | 30 | 40 |
| Var3 % | 30 | 10.47 | 10.63 | 10.8 | 10.97 |
| 40 | 9.06 | 9.24 | 9.42 | 9.6 |
| 50 | 9.7 | 9.91 | 10.11 | 10.32 |
|
| Var2 % |
|
|
|
| KJ6r |
| 10 | 20 | 30 | 40 |
| Var3 % | 30 | 0.96 | 0.32 | -0.31 | -0.94 |
| 40 | 2.57 | 1.78 | 0.99 | 0.2 |
| 50 | 3.75 | 2.86 | 1.96 | 1.06 |
as you can see, this line shouldn't lose money on most boards. that said, it's not necessarily the optimal way to play the hand.
anyone want to try picking a flop or two and looking at alternate lines?
Is there any way that by looking at these numbers that we can figure out why there is an inverse relationship between 5h43h and Th98h with the 3 bet bluff %? These hands don't seem dissimilar enough to me to have an opposite response to bluff %. That's assuming that I'm reading these correctly. Any insight?
ReplyDeletebtw, I just had my kid, so I've been worthless the last couple of weeks. Catching up on the postings now. Great stuff. Since I'm never going to be able to leave my house again I figure I might as well re-focus my energy on some poker. Maybe I'll start contributing something in the next few weeks, or at least try to feed some info off of your stuff.
yeh, i'm not sure. i think it probably has to do w/ the btn calling range, since on the 5-high board there are (probably) more combos of >=tp, but fewer ways to make combo draws w/ the top X% of hands.
ReplyDeletei think this format of trying to present results for group interpretation might not really work. if you had the .stx file and could tinker w/ the results and see which exact hands are folding/calling etc. it would be much better.
i'll look for you online...