AA Draft Quarterfinal Round: #4 Saskatoon Sheiks vs #5 HC Krylya Sovetov

[seventieslord]

No, unfortunately the currently/mostly used formula is incredibly stupid, and only considers series winner.

How a series with 8 votes of 4-0 and two of 3-4 can be 4-1, yet one with 10 votes of 4-3 is a 4-0 has always been beyond me.

You are right that your far-fetched examples produce results of disproportionate fairness for the losing team.

Keep in mind that this workd quite well for the atd, particular in the opening rounds. If we took the "median" as I suggested, we would have pretty boring results, even some very lopsided series would evd in 6 or even 7. I think almost all would be one or the other.

When judging 8 or 16 series, it is easy to take the aggregate results, then avoid painting all the series with the same brush, by saying, these two were the most lopsided,I will call them sweeps, these two were the closes, I will call them 7 games, and this one was even closer so let's call it game 7 OT. (All subject to appropriateness of vote counts, of course)

Once we are in the MLD or even the late atd rounds when voting declines, though, these smaller vote samples make "sweeps" more common despite the greater likelihood of the series involving more evenly matched teams. That is when I say we should go by the median.

So, if we had 5 votes, for it to truly be called a sweep, three people would have had to have voted for a sweep. And if five all voted for a 7-game victory for the same team, that would be the result.

I can tell you for sure, mad arcand, that when I am in charge of any series voting I take seriously the responsibility of making sure the posted result fairly reflects the votes cast, either in the series on its own, or in relation to the other series, where applicable.

 

[BenchBrawl]

You are right that your far-fetched examples produce results of disproportionate fairness for the losing team.

Keep in mind that this workd quite well for the atd, particular in the opening rounds. If we took the "median" as I suggested, we would have pretty boring results, even some very lopsided series would evd in 6 or even 7. I think almost all would be one or the other.

When judging 8 or 16 series, it is easy to take the aggregate results, then avoid painting all the series with the same brush, by saying, these two were the most lopsided,I will call them sweeps, these two were the closes, I will call them 7 games, and this one was even closer so let's call it game 7 OT. (All subject to appropriateness of vote counts, of course)

Once we are in the MLD or even the late atd rounds when voting declines, though, these smaller vote samples make "sweeps" more common despite the greater likelihood of the series involving more evenly matched teams. That is when I say we should go by the median.

So, if we had 5 votes, for it to truly be called a sweep, three people would have had to have voted for a sweep. And if five all voted for a 7-game victory for the same team, that would be the result.

I can tell you for sure, mad arcand, that when I am in charge of any series voting I take seriously the responsibility of making sure the posted result fairly reflects the votes cast, either in the series on its own, or in relation to the other series, where applicable.

In this case I could suggest literally saying this team won for example 7-1 , 3-0 , 7-4 , just by counting how many peopel said team 1 won and how many peopel said team 2 won insteand of saying Team 1 won in 4 games.We could do this for the none-important rounds of the ATD and lower drafts , but make sure we do the real average when we're at 8 or 4 clubs left in the ATD and probably every draft's finals.

 

[MadArcand]

Man, you're taking this too personally.

Sure, with huge and varied samples, and with creative writing, it works quite decently.

But.

It's still a very, very flawed methodology, if one wants the result to actually reflect the votes. If everyone feels a series should be incredibly close, then it should be such, not a sweep.

I don't think 4-3s would be common or prevailing with median. 4-2s and 4-1s, with sweeps being maybe somewhat rarer (though OTOH now all it takes to prevent a sweep is a single contrarian vote, with median it wouldn't be enough).

 

[seventieslord]

I’m not taking anything personally, I’m just trying to be transparent for you, since you are questioning the process.

Let’s say that in theory in a round 1 or 2 ATD series, there were 18 votes for a series, and they all voted for team A to win in 7 games. You’re making the mistake of assuming I would call that a sweep, or that anyone else would. I would employ enough common sense to compare this result to the other results and judge it accordingly. If someone else won their series 15-4 but there were a bunch of 4 and 5 game victories for them, I would likely consider that series to not be as close. (of course, if someone really did win their series 18-0 there would never be 18 seven game votes, there would be plenty for five and even four)

Also, you say that the result should reflect the votes, if everyone thinks it should be close then it should be close. What about when someone wins 10-9 in votes, but most of them were for 4-5 games either way? No one thought the series should be Close, yet, what choice would we have there?

I’ve collected a lot of votes for these drafts and at the leafs chat site. When I say 6 and 7 game victories would become the norm I’m not just speculating. A lopsided victory for EagleBelfour against some noob would look like this: 5 6 7 6 7 6 6 6 5 6 6 7 5 6 5 6 6 7 5 7 6 6. The median vote would be for EB in 6 games, and this would be the MOST lopsided series of the round.

I wouldn’t be comfortable with just calling every series 6 or 7 games. There has to be some differentiation provided between actually close, and “closeâ€.

In other words, I can poke holes in the “median†method just as easily as you can poke holes in the “first past the post†method (or whatever we want to call it) that currently prevails in the first two rounds of the ATD. I think that each has their place at the right time, and I’ve explained why.

 

[TheDevilMadeMe]

No, unfortunately the currently/mostly used formula is incredibly stupid, and only considers series winner.

How a series with 8 votes of 4-0 and two of 3-4 can be 4-1, yet one with 10 votes of 4-3 is a 4-0 has always been beyond me.

Have you ever been in charge of tabulating the votes?

The honest reason we don't use your proposed method is because it adds significant time to the process of tabulating the votes