Reno 911 - day 1

[magratheazaphod]

I came third! I played well except for phonying a lot and making bad challenges. also, I had mucho fun gambling/pokering. unfortunately I left my shoes in nevada, but besides that and my cold the trip went well. yay escapism!

I'm just going to throw up interesting positions that come up as they come along. day 1 had a lot of fun moments.

in game 1 I

Comments

[wantonhalo]

Maybe RARED instead of AE against Lester? No VIEWY or plays from DE after TARED. The only think that would concern me after that is all the vowels in the pool.

JIG instead of JAG against Dominick still wards against Q trouble but keeps your A for rows 1 and 4. You probably don't use it in most endgames, but I bet it wins you a couple more.

Why DIVAN against Jon? I don't really see a great reason to ramp up the entropy. It seems like you have better tiles and better skills than your opponent, so VANS would be better, scoring well and going for a lead on a somewhat closed board. Also, what about KELTER in the endgame? I don't really think your reasoning for LET and LECH is good enough to justify it as the best play. On the other hand, I don't really know if KELTER is right either, just that it seems like a good option. If you look at some typical draws after either play, they both do pretty well. One advantage LET has is one-in-the-bag, which would probably perplex your opponent more than a straight endgame.

[magratheazaphod]

I'll do these plays backwards

[wantonhalo]

I'll do the plays in an order that matches neither mine nor yours.

If you're going to do AE you may as well do RARED. They score the same, and ADERT isn't going to be 15 points better on this kind of board. There's not nearly as much "firepower" left after RARED.

I don't think the DIVAN scoring spot is particularly scary. My reason for VANS is more to start closing down the board now because you have the advantage with the blank in hand, and it'd be nice to get a closed board.

The sim is not going to tell you much about the pre-endgame here. In fact, the sim says KELTER wins 70% of the time while LET wins 45%, which seems like too large a gap. If you look at some typical racks for Jon, you'd be surprised how often he doesn't have a winning endgame after KELTER, and some of the winning endgames are tough. I really think your heuristic arguments for LET are too weak to properly justify it. We need some more heavy-duty reasoning, but I don't know what.

Nothing to say about Dominick. Well, except that I have nothing to say about

[magratheazaphod]

yeah, rared probably looks better, wanted to be more flexible but blocking viewy is a big priority.

I'll definitely still take divan.

I'll have to take more of a look at kelter later. what kind of reasoning method do you have in mind? even kelter requires 152 different endgames to be analyzed according to my back of the envelope calculation, let alone if we don't empty the bag...I'm a pretty huge fan of using heuristics to do pre-endgames. I know kenji isn't.