open challenge

[47]

i challenge anyone to a 21 game series.

if it finishes 11-10, the conclusion is that the two players are pretty evenly matched.

if it finishes 13-8 or worse, then we know who the better player is.

if it's 12-9, we argue about whether that result is significant


first taker gets the bid

[mattmonyo]

i challenge anyone to a 21 game series.

if it finishes 11-10, the conclusion is that the two players are pretty evenly matched.

Why did you choose an odd number if you're trying to find out if you and your opponent are even? The only reason sets are (almost) always comprised of an odd number of games is because there is always a clear winner. I think this would work better if you picked an even number. If you tie, you tie and you're even. If you don't tie, refer back to your list of what is what.

[47]

i challenge anyone to a 21 game series.

if it finishes 11-10, the conclusion is that the two players are pretty evenly matched.

Why did you choose an odd number if you're trying to find out if you and your opponent are even? The only reason sets are (almost) always comprised of an odd number of games is because there is always a clear winner. I think this would work better if you picked an even number. If you tie, you tie and you're even. If you don't tie, refer back to your list of what is what.

see, that's not a good question. as you recognize, the series is meant to to provide insight into who is the better of the two players.

suppose you walk into a room full of coins, some of which are fair, some of which are not. you pick one up and want to know whether it is fair or weighted to one side.

you can flip it twice, but if it comes up heads once and tails once, it doesn't tell you much. similarly, if it comes up heads twice, that doesn't tell you much either.

to gain any meaningful information about the fairness of the coin, you will have to flip it many times. the more times you flip it, the more you learn about its fairness.

the series i have proposed is analogous to this situaiton.

the coin's turning out to be weighted (one side lands face-up more frequently) is analogous to the result that one of us is better than the other.

essentially, you are saying... 'to try to figure out if the coin is fair, you should flip the coin an even number of times, like 100. that way, if it comes up 50 and 50, you will know the coin is even... and anything else, you will know which way it is weighted. but we now see that argument to be absurd.

it doesn't matter if my series has an even number or an odd number of games... what matters, is that there is enough to give us some meaningful information.

[Rubik87]

I accept the challange.

What about playing 20 games instead, and switching "who goes first" every time, like in a chess match. Counting draws as draws, instead of replaying them (yes, a tie is possible in quad).

[Pimpwin]

Id be down

[47]

I accept the challange.

What about playing 20 games instead, and switching "who goes first" every time, like in a chess match. Counting draws as draws, instead of replaying them (yes, a tie is possible in quad).

i accept your amendments. but how will we ensure that the order alternates?

[mattmonyo]

see, that's not a good question. as you recognize, the series is meant to to provide insight into who is the better of the two players.

suppose you walk into a room full of coins, some of which are fair, some of which are not. you pick one up and want to know whether it is fair or weighted to one side.

you can flip it twice, but if it comes up heads once and tails once, it doesn't tell you much. similarly, if it comes up heads twice, that doesn't tell you much either.

to gain any meaningful information about the fairness of the coin, you will have to flip it many times. the more times you flip it, the more you learn about its fairness.

the series i have proposed is analogous to this situaiton.

the coin's turning out to be weighted (one side lands face-up more frequently) is analogous to the result that one of us is better than the other.

essentially, you are saying... 'to try to figure out if the coin is fair, you should flip the coin an even number of times, like 100. that way, if it comes up 50 and 50, you will know the coin is even... and anything else, you will know which way it is weighted. but we now see that argument to be absurd.

it doesn't matter if my series has an even number or an odd number of games... what matters, is that there is enough to give us some meaningful information.

Everything you've said is sound, but I still don't see the sense in making it odd. By making it odd, you remove the possibility of the games to arrive at a dead even final. By making it even, you allow this conclusion to occur, whilst also allowing a lopsided one. A tie result isn't bad. It's just as a valid solution as one player being better. The series being even just allows for an extra outcome.

[Rubik87]

I accept the challange.

What about playing 20 games instead, and switching "who goes first" every time, like in a chess match. Counting draws as draws, instead of replaying them (yes, a tie is possible in quad).

i accept your amendments. but how will we ensure that the order alternates?

We re-load the page and re-challange each other if it is wrong. It may be a bit annoying, so if you don't want that it's ok, we can just make it random everygame. Starting first is a small advantage, but over 20 games if the same player starts first many times more than the other all those small advantages will stack to make a pretty significant one.

Standard settings, with 2 minutes timer?

[47]

sounds good. i don't mind the reload.

[Mammalman]

i'll butt in with a suggestion: just keep track of who goes first for the first 7-8 games (or basically until someone goes first 5 total times) and only THEN bother restarting games. no need to alternate as long as the totals end up even.

[47]

i'll butt in with a suggestion: just keep track of who goes first for the first 7-8 games (or basically until someone goes first 5 total times) and only THEN bother restarting games. no need to alternate as long as the totals end up even.

i did think of that....

there is something to be said for the psychological advantage or disadvantage that can come with getting to go first several times in a row...

but i am willing to do it this way if sanzo is

[driven2sin]

so it should work both ways.. one guy goes first 10 times in a row and then you set it up so the other goes first the next 10

I'll give you the first TEN moves and still beat you