Preface
I had a very solid 8-6 finish in division 1 in Season 13, but I wasn’t the most prepared for this season after being busy.
Game 1
In this position, down 28 points, I was holding a solid rack but had no playable bingoes.
There are no standout options here. In this position, I had to make a play I hated but didn’t see anything better than. I played 12L (JA)R for 10 points, setting up my S on the triple lane. In a vacuum, this play seems pretty alright. There are only 5 tiles unseen that hook onto JAR (the blank, 2 S’s, the P for JARP#, and 1 L). If he doesn’t have any of these tiles, then he’s going to be in a major bind and I’m super likely to hit a very high scoring bingo on the triple. The major problem with this is how many tiles he’s played. He’s played 23 tiles so far in the game without playing any S’s. Also, he’s made multiple short, cagier plays in a row, suggesting strong tiles. An S is definitely in his range at this point. Even if he didn’t have the S, he could easily have drawn one of other tiles that hook onto JAR, only making my deficit even worse. Only inferring from his most recent play of MNA#, Macondo suggests that all of the tiles that hook JAR are more likely than random to be on his rack.
Using these inferences and performing a 5-ply sim, exchanging down to ERS? or multiple other bingo prone combinations solidly outperform playing (JA)R.
In the pre-endgame, I’m down 36 points in the following position:
The unseen tile pool is AAAEGINRTTY and there are only 4 tiles in the bag. I like having the Z to score with, but there are no strong scoring spots for it in this position. Also, the V is extremely clunky here and can’t be dumped for a solid score. The first play I considered was to burn the blank and empty the bag to play C9 EVZOnE for 39 points. This takes a 3 point lead, but given that I’ll have 5 tiles on my endgame rack, I’m not very likely to go out, and andy is pretty likely to have a out-in-two sequence that beats me.
The next play I looked at was 9B ZO(N)E 14. This play sets up the case blank (with no O’s unseen) to make o(ZONE) hooking plays on the triple. It also sets up my D on the other side with ZONED that may also yield some bingoes. The problem, I felt with this play was that it was easy to block if andy had the R to hook ZONER, which would eliminate both my threats at the same time. The problem with this line of the thinking is that I’m already in some trouble in this position. He’s signaling a decently strong rack with his last play of just 2 tiles that score 1 point each. I need to assume that he doesn’t have the R if I want to win. If I can draw the R myself or leave it as the last tile in the bag, then I’m threatening multiple big plays that will be very tough for him to block.
One other Macondo suggested option is the simple M1 ZO# 16. This option is mostly all-in on bingoing. There are not a lot of ultra conclusive blocks of the lanes for bingoes starting with vowels next to TOUN#. With a lot of vowels unseen, my chance of drawing a bingo is actually pretty good, 70.4% according to Macondo (though in practice likely a bit lower because of cagier plays by andy).
In the actual game, I played D2 (R)OE for 20 points. My goal with this play was to try to draw any of the A, I, or Y to play AVIZED#/AVYZED# for upwards of 40 points. The problem with this is that I’m still down 16 points plus whatever andy scores on his next turn. Even if he empties the bag, I’m just not winning that often. Macondo rates this as only 25.7% to win, down 34.5% from its favorite plays of ZO(N)E and ZO# mentioned above.
In the actual game, andy bingoed with E7 GRANITE 74 and though I got to play C9 AVYZED# for 50, andy played out the next turn to win by 52.
Game 2
With 6 in the bag, I’m up 44 points against jellomochas (Alec Sjoholm).
Unfortunately, DISSERS* is not a word, so I have a tougher decision to make. The unseen pool is AABDHIIILLOR?, so not too scary, especially with how the lane hooking s(PIT) with a blank S scores very little. In this position, I really need to be playing a lot of tiles. If I leave just 1 tile in the bag, if he bingoes immediately, he’ll draw 1 tile only and I’ll be down only around 5-10 points and will get the first shot to hit any open multiplier squares he opens. This also means that his outplays won’t score almost any points, and I’m very likely to win easily.
To accomplish this, I should be playing something like 13D SIDES 14. This play blocks rare bingoes hooking PIGS with a blank S that score more and opens absolutely nothing. I can also accomplish something similar with J1 SI(L)DS 24 at the top of the board. This leaves 2 tiles in the bag, so less ideal, but also scores an extra 10 points.
In the actual game I made a huge mistake. I ended up settling on the odd looking 13I S(U)S for 12 points. My thinking was that by leaving 4 tiles in the bag, jellomochas was less likely to go out if he bingoed immediately. I was also thinking that my leave of DEIRS was pretty strong and with multiple common draws like AI, I was threatening a bingo in 2 spots (in this case DAIRIES with SPIT or AIRSIDE with PIGS) that would guarantee me the win. The problem is that 4 tiles in the bag just isn’t really what I want here. 4 tiles means he’s still likely to draw an outplay that I will struggle to block while scoring and also means this play is vulnerable to him fishing if he doesn’t already have a bingo, for example, something like 3E D(AL) sets up the O(DAL)# hook in an annoying way. The plays Macondo suggested of the aforementioned SIDES or SI(L)DS win about 99.3% of the time while S(U)S may only win 88.8% of the time. I can’t be affording to make such big errors in a single move like this.
After I played S(U)S, jellomochas played C4 RIBALDs 67, and I had no winning pathway even with a perfect endgame sequence blocking his outplay of OHIA, and I lost by 4 points. If I had played either of the other plays, I win easily after RIBALDs.
Game 3
Against merlion I’m in this position down 77 points, running out of time with 8 in the bag.
To start, I completely whiffed on the bingo of N3 COSHERS for 80 points. It’s not Macondo’s favorite play but it’s certainly better than the play I actually made. The actual move I played was M1 HORS(Y) 31 holding CES. My idea was to make a second dangerous lane perhaps with some -CHES triple triple possibilities while also preserving the lane hooking YAYS. As with my game vs jellomochas, one of the problems is timing. Leaving 4 in the bag, he can play 3 tiles and score while closing at least one of the lanes, and if he closes the upper lane, my bingoes hooking YAYS will score very little. Much better is to play N2 CHOSE 34. This makes a much stronger opening with the lane next to the E. A big part of this play is also that it leaves 3 in the bag. Merlion cannot hit the triple here without emptying the bag. If he opts to make a low scoring 1 or 2 tile play next to the E, I may still be able to hit a bingo through the U at the base of the board or try to fish for a bingo in 2 spots by opening a second lane at the top of the board. Macondo suggests that not making this adjustment basically throws away almost half of a game.
My Favorite Play of the Season In this game, though I’m already lost as I’m down 206 points and MGI is holding only AA, I made my most stylish find of the season.
In this position, with no outplays available, I made the very hidden find of K3 V(I)TE# for 36, 25 points better than my next best option.
Summary
From this season it’s clear that beyond word knowledge, the area I need to improve the most on is making pre-endgame decisions. I’m planning to try to do more practice over the next month if I have time.
Stats
Lost Endgame Points: 14 (12/14 optimal endgames)
Missed Bingo Turns: 8
Suboptimal Bingo Turns: 7