Elise's ROAD C1 seems mysterious at first, but consider that it greatly reduces the possibility of a bingo by Quackle in the last few moves of the game. (Where would it likely fit?) At the same time, by making this move, Elise is creating another possible spot to play the Z, on square B6. At the time, Elise's other spots for the Z are A8 (the triple word square) or J6. The trouble with A8 is that Quackle is likely to play there, and if not, Elise could spend other tiles there and make almost as many points. The trouble with J6 is that it interferes with the hook at F6 and makes the endgame more awkward; also it is generally not good policy to rely on one space only for your 10-point tile in the very late game. Quackle's YIN M1 immediately afterwards is an attempt to open up a possible big play on the first row -- it's behind enough that it needs to try what it can to open scoring opportunities at the expense of immediate points. Elise recognizes this, and blocks with OH L1. A move like QAt B10 seems like a "computer move", a move that a human player would unlikely make in this situation. It scores 34 points -- but then, AY L14 scores 33 and doesn't use the blank. Alternately, if you want to use the blank you could play sHINY or wHINY at 14B for 42 points. But QAt B10 has its advantages; first, it keeps the triple word squares in that corner of the board out of play until the late game (and makes the TWS at A15 difficult even then), second, it plays the Q, for which there are not many places on the board, and this actually increases the probability of bingoes over the next couple of moves, despite using the blank. Elise gives QAt a win probability of 91.5% looking 7 plies ahead, versus 90.6% for sHINY and 90.4% for AY; the difference between the moves is not very large. Back

Move#	Player	Rack	Move	Sq	Leave	Score	Total	Notes
1	Elise pre-release	eioortx	OXO	8F	[eirt]	20	20	7-ply winprob simulation (1360), -8.43 / 54.8%
2	Quackle	adeippu	PUP	9F	[adei]	26	26
3	Elise pre-release	eeeirrt	REE	10D	[eirt]	12	32	7-ply winprob simulation (1840), -17.73 / 51.1%
4	Quackle	aadeiii	RADII	D10	[aei]	12	38
5	Elise pre-release	deeirtu	ERUDITE	10H		67	99	7-ply winprob simulation (480), +24.88 / 69.4%
6	Quackle	aaaeilr	AA	C12	[aeilr]	11	49
7	Elise pre-release	eegnrt?	mETERING	L5		68	167	7-ply winprob simulation (1360), +84.93 / 91.4%
8	Quackle	aabeilr	RADIABLE	K8		83	132
9	Elise pre-release	eijnqv?	JIVE	15H	[enq?]	42	209	7-ply winprob simulation (480), +53.61 / 85.7%
10	Quackle	imnttuv	VIM	K3	[nttu]	22	154
11	Elise pre-release	aehnqy?	QAt	B10	[ehny]	34	243	7-ply winprob simulation (2480), +67.88 / 91.5%
12	Quackle	amnttuu	UT	7E	[amntu]	8	162
13	Elise pre-release	eghnnty	THING	14B	[eny]	32	275	7-ply winprob simulation (1280), +81.65 / 95.4%
14	Quackle	abmnotu	UMBONATE	N3		64	226
15	Elise pre-release	aeeefny	FEY	O6	[aeen]	49	324	7-ply winprob simulation (1200), +64.39 / 95.0%
16	Quackle	ceforsw	COWS	G11	[efr]	23	249
17	Elise pre-release	aeennrz	NENE	O1	[arz]	22	346	7-ply winprob simulation (3440), +76.60 / 97.8%
18	Quackle	eflorsw	FLOWERS	D2		81	330
19	Elise pre-release	adgorsz	ROAD	C1	[gsz]	22	368	7-ply winprob simulation (1280), +34.31 / 92.1%
20	Quackle	iiiklny	YIN	M1	[iikl]	18	348
21	Elise pre-release	adghosz	OH	L1	[adgsz]	17	385	7-ply winprob simulation (80), +46.90 / 100.0%
22	Quackle	ciiklls	KISS	8A	[cill]	24	372
23	Elise pre-release	adegosz	ADZ	B4	[egos]	37	422
24	Quackle	cillt	CULT	J9	[il]	18	390
25	Elise pre-release	egos	SEGO	6F		16	438

Elise pre-release: 440, Quackle: 388