Recounting the Tie

Reprinted from the March 96 issue of the JG Newsletter.

An anonymous reader from Texas queries: "When is it correct to ask for a recount on a tie?" The general answer is when you have a good spread. Since when you have a tie, you are most likely the only player in your score group (except for mega tournaments like the national championships), accepting the tie makes spread a nonfactor for the rest of the tournament. If you have a crummy spread, you'd like to hide it away; but with a nice spread, you want the whole world to see it. This also can be measured by what you're risking vs what you're hoping to gain on a recount. The following partial standings demonstrate what our heroine Plain Jane is risking vs trying to achieve in a typical scenario:

CASE I
PLAIN JANE HAS A GOOD SPREAD

	SHE TAKES THE TIE			SHE RECOUNTS & WINS			SHE RECOUNTS & LOSES

1)	Johnny Mack	5-4	+400		Johnny Mack	5-4	+400		Johnny Mack	5-4	+400
2)	Cason Pointe	5-4	+250		PLAIN JANE	5-4	+300		Cason Pointe	5-4	+250
3)	John Nunn	5-4	+80		Cason Pointe	5-4	+250		John Nunn	5-4	+80
4)	Dave Skalansky	5-4	-100		John Nunn	5-4	+80		Dave Skalansky	5-4	-100
5)	PLAIN JANE	4.5-4.5	+300		Dave Skalansky	5-4	-100		PLAIN JANE	4-5	+300
6)	Nick Vanos	4-5	+150		Nick Vanos	4-5	+150		Nick Vanos	4-5	+150
7)	Travis Bickle	4-5	+20		Travis Bickle	4-5	+20		Travis Bickle	4-5	+20
8)	Topper Headon	4-5	-100		Topper Headon	4-5	-100		Topper Headon	4-5	-100
9)	Mo Delhome	4-5	-200		Mo Delhome	4-5	-200		Mo Delhome	4-5	-200

In the above example, Plain Jane risks going down no places while freerolls at going up three spots.

CASE II
PLAIN JANE HAS A BAD SPREAD

	SHE TAKES THE TIE			SHE RECOUNTS & WINS			SHE RECOUNTS & LOSES
	
1)	Johnny Mack	5-4	+400		Johnny Mack	5-4	+400		Johnny Mack	5-4	+400
2)	Cason Pointe	5-4	+250		Cason Pointe	5-4	+250		Cason Pointe	5-4	+250
3)	John Nunn	5-4	+80		John Nunn	5-4	+80		John Nunn	5-4	+80
4)	Dave Skalansky	5-4	-100		Dave Skalansky	5-4	-100		Dave Skalansky	5-4	-100
5)      PLAIN JANE	4.5-4.5	-150		PLAIN JANE	5-4	-150		Nick Vanos	4-5	+150
6)	Nick Vanos	4-5	+150		Nick Vanos	4-5	+150		Travis Bickle	4-5	+20
7)	Travis Bickle	4-5	+20		Travis Bickle	4-5	+20		Topper Headon	4-5	-100
8)	Topper Headon	4-5	-100		Topper Headon	4-5	-100		PLAIN JANE	4-5	-150
9)	Mo Delhome	4-5	-200		Mo Delhome	4-5	-200		Mo Delhome	4-5	-200

In the second example, Plain Jane risks going down three spots with no upside potential.

There are cases where considerations other than spread factor in. At the Superstars tournament, I tied my ninth round game against Ed Halper after going 7-1 in the first eight games. In this situation, instead of risking spots based on spread, the risk is $ based on final standings. Since being in the top three at the end of the tournament was not an impossible expectation at this point, it is correct to recount the tie. Let's say hypothetically that a tie would give me second place at the end of the tournament, while a win would give me first, and a loss, third. Given a payout of $50k, $20k, $10k for the top three spots, it's conceivable that recounting a tie in that situation would be like betting $10k to win $30k. In tournaments, it's generally correct to take risks that will push you to one end of the standings or other. The reward for winning the tournament one week and taking last the next is much greater than finishing even two weeks in a row. World-class tournament poker players know this. They are often knocked out early in a tournament due to their aggressive style but certainly end up in the winner's circle more often than the solid player. I'm certainly not advocating you play insanely every time you sit down at the board during a tournament, but from a game-theory standpoint, fortune favors the brave.

Last Modified 2/7/00


Back to Scrabble Correspondence Page

Back to Scrabble Main Page

Back to Jim's Main Page