Deuce-to-Seven starting hands--I think Daniel was wrong here

[weirdboy 0]

A while back, I was doing a bit of reading on available material for triple-draw lowball, and came across the following thread in RGP:linkIn this thread, Daniel and GCF trade barbs back and forth and at one point in the discussion of starting hands, Daniel says the following:

Of course, a good simulation would show you that these three cardholdings are not as strong as many he left out, namely (234-235-245). Theimportance you give to holding a 7 is clearly overrated. Sure drawing toa 7 is always nice but you are going to miss more often than not.Realistically, you should be more than happy to end with a good 8, ratherthan a rougher 8-7. If you ran your 237 against my 234 heads up a milliontimes... I'd bust you.

Daniel's conviction in this statement interested me, so I did what he said; I ran this as a simulation on twodimes.net's hand analyzer. Here are the results:

pokenum  -l27 2c 3s 4d  - 2d 3c 7h 5-card Draw 2-7 Lowball: 3916440 enumerated outcomescards          win           %win    lose           %lose     tie      %tie     EV3s 2c  4d  1925332  49.16  1987004  50.73  4104  0.10  0.4923c 2d  7h  1987004  50.73  1925332  49.16  4104  0.10  0.508

As you can see, 237 in fact has a slight edge over 234. It's not much, but if you were to play as he suggests in his post heads up a million times, he'd go bust.As PL deuce-to-seven is one of my favorite games, I thought the thread was interesting in terms of strategy. And, of course I'd also love to hear any comments Daniel has on the subject a year and a half later. I've also heard that he wrote the lowball section of the new Super System, so I'm looking forward to that as well.