http://j-archive.com/showgame.php?game_id=6843
Alex hit the Daily Double in "The Cubists" on the last question of the round on #26. He knew at question #23 (Medical History for $1200) there was a minute left, so this was going to be the last question.
The score before the DD wager was Kevin: $14,400, Alex: $5900, Natt: -$400.
So Natt was out of it. Alex had below $7200, so, he was out of it too if he didn't get the question right, because Kevin was going to have a "lock game" (could wager $0 in FJ and not get caught up to) with the scores they way they were.
So, Alex. had to wager
at least $1300 on The Cubists.
At that point, what's the difference in saving anything? Why not make it a true Daily Double?
If he wagers $1300 to everything, and gets the question
right, he ends up with 7200-11800, Kevin has 14400, and, Alex needs to get FJ right, and Kevin needs to miss it, for Alex to have any shot.
But if he gets the question
wrong, no matter what he wagers, it's game over anyway. If he wagers $0 to say $3000 on the DD and gets it wrong... he goes into FJ down 14400 to 5900-to-2900. He has to stand there through FJ looking like an idiot in a game that won't matter. Did it really make a difference if he's actually there or not? He's locked into second place.
Basically... a wrong answer on the DD leads to the same result no matter what the wager is. So he had to get the question right. Knowing that... might as well get as much cash as you can, because you're banking on putting pressure on Kevin to get his FJ wrong and then you win more money at the end of the day.
Knowing that he had the question all to himself, and, that he knew what the category even was... not knowing what subject might come up in FJ... he has to bank on getting it right and in that case, might as well bet it all. No point in saving anything in reserve "in case you're wrong" because it's of no use in FJ.
The error Alex did make was in picking the question to begin with.. having gotten question #23 right, then missing on #24 for $1600 and #25 for $2000, if he had done the math quickly enough, he would have realized he needed a question of high value to get back to the $7200 halfway point. He should have picked Cubists for $1600 or $2000. He was lucky it was a DD--had it been just a regular $1200 clue, it would have been awful: he could have
answered correctly, the buzzer would have sounded, and he would have ended the round $100 short of forcing FJ to matter.