Join In On The Action "Register Here" To View The Forums

Already a Member Login Here

Board index Forum Index
User avatar
Dignitary
 
Posts: 3486
Joined: 02 Oct 2000, 9:01 am

Post 05 Oct 2015, 2:32 pm

Doctor Fate wrote:I'm not sure that anyone in the finals really "punts" a category, but okay.


There is always a team or two that does not pursue saves or SB.
User avatar
Ambassador
 
Posts: 21062
Joined: 15 Jun 2002, 6:53 am

Post 05 Oct 2015, 2:44 pm

geojanes wrote:
Doctor Fate wrote:I'm not sure that anyone in the finals really "punts" a category, but okay.


There is always a team or two that does not pursue saves or SB.

Right, but do they make the finals?
User avatar
Ambassador
 
Posts: 21062
Joined: 15 Jun 2002, 6:53 am

Post 05 Oct 2015, 2:46 pm

geojanes wrote:
Doctor Fate wrote:I'm not sure that anyone in the finals really "punts" a category, but okay.


There is always a team or two that does not pursue saves or SB.


Furthermore, if they have punted one category, shouldn't they be stronger in other categories?

Oh, and it seems to me that it is nearly impossible to completely punt S/H. If you have a reliever, you're going to stumble into stats.
User avatar
NASCAR Driver (Pro V)
 
Posts: 7810
Joined: 08 Apr 2002, 9:45 am

Post 05 Oct 2015, 3:48 pm

Doctor Fate wrote:Oh, and it seems to me that it is nearly impossible to completely punt S/H. If you have a reliever, you're going to stumble into stats.

See: Turtle Soup. I'm not saying it was a successful strategy but if luck had turned my way this year, I wouldn't be entirely out of the playoff conversation.

I don't really buy the fairness argument (Cubs v. Pirates is much more egregious than Nick winning a tie-breaker) but having said that, if there's a groundswell to change the tie-breaker, I won't stop it.

So, what is transparent, simple, dependent upon current performance, but does not advantage any particular category?

My aforementioned 'pick a team'. That fits but may be too... random?
Pick your two best players (one offense, one pitching) and go mano a mano? Stars sometimes sit the final week.

I'm stumped...
User avatar
Dignitary
 
Posts: 967
Joined: 30 Aug 2003, 5:17 am

Post 05 Oct 2015, 3:59 pm

Doctor Fate wrote:Again, how about a situation where the two teams are separated by half a game?


How about a situation where the #6 & 7 seeds are separated by half a game, and the latter misses the playoffs entirely? Or the #2 and #3 seeds are separated by half a game, and the #3 misses out on a bye? The tiebreaker is small potatoes compared to that. But it's hardly the sort of thing that seems like it needs fixing.

Also, I completely agree with Mike's critique of the notion of looking at category differences. It's logistically infeasible and privileges some strategies over others. No thanks.

The only thing suggested so far that makes sense to me is going from seeding-based tiebreakers to pure record. Even there I'm not convinced it's a change we need to make, but I'm open to it.
User avatar
Dignitary
 
Posts: 967
Joined: 30 Aug 2003, 5:17 am

Post 05 Oct 2015, 4:02 pm

Actually, here's a thought. Why don't we award the tiebreaker to whichever team has the most players on the winning team in the All-Star game?

:razz:
User avatar
Adjutant
 
Posts: 3646
Joined: 17 May 2013, 3:32 pm

Post 05 Oct 2015, 4:10 pm

Look at how each owner did in each category in the regular season. Calculate the mean and standard deviation for each category (e.g add each owner's run totals and get the mean and standard deviation). For the quantitative categories divide 21 to get the weekly totals and calculate the mean and standard deviation from that. if a match-up is tied you look at the number of standard deviations from the mean an owner gets in each category (e.g. If an owner got 50 runs and the average was 40 with a standard deviation of 10 then the owner would get credit for 1 standard deviation). Then just add the total number of standard deviations for the ten categories and the player with the highest total wins.

Of course this would be slightly cumbersome...
User avatar
NASCAR Driver (Pro V)
 
Posts: 7810
Joined: 08 Apr 2002, 9:45 am

Post 06 Oct 2015, 7:59 am

Largest donation to the League Commissioner's charity (read beer) fund gets the tie-breaker.

Simple. Elegant. Corrupt. What's not to like?
User avatar
Ambassador
 
Posts: 21062
Joined: 15 Jun 2002, 6:53 am

Post 06 Oct 2015, 8:12 am

SLOTerp wrote:Largest donation to the League Commissioner's charity (read beer) fund gets the tie-breaker.

Simple. Elegant. Corrupt. What's not to like?


Bwhahaha! I'll never lose another tie-breaker!
User avatar
Administrator
 
Posts: 7374
Joined: 26 Jun 2000, 1:13 pm

Post 06 Oct 2015, 8:53 am

SLOTerp wrote:Largest donation to the League Commissioner's charity (read beer) fund gets the tie-breaker.

Simple. Elegant. Corrupt. What's not to like?


You mean all that money I spent doesn't help me if I don get to the final game? What kind of bribe is that!

So limiting...
Adjutant
 
Posts: 80
Joined: 22 Apr 2014, 5:27 pm

Post 06 Oct 2015, 11:21 am

Count me in the group that thinks the tiebreaker we use now is good. I would only add that it should be based on record rather than seeding. I would much rather have the tiebreaker come from my work over the course of the year than a fluky result one week in a certain category. That seems completely arbitrary. Right now we are rewarding the right thing: being consistently good. Want the tiebreaker? Build the best team, using any strategy you want. You'd rather have a team win because they happened to be two standard deviations over their mean in steals one week? I don't see it.
User avatar
Dignitary
 
Posts: 857
Joined: 13 Oct 2000, 9:42 am

Post 04 Jan 2016, 10:18 pm

I also hate the idea of having a designated tiebreaking category. We had ERA filling this function when our league started and that was arbitrary and dumb. I didn't mind losing to Nick on the tiebreaker in the finals. He won more games and earned the higher seed, and I knew going in that I needed 5-4-1 to win. It was a fantastic matchup that went down to the final weekend.

However, if there is general support for lessening the impact, then how about designating tiebreakers based on manager preferences? Here's the proposal:

Prior to each matchup there is a deadline for managers to submit tiebreaker preferences. Each manager will submit a list of the 10 categories in the order they would like tiebreakers applied. The tiebreakers are made public at the start of the matchup. In the event of a tie the winner of the designated category will win the matchup. If that category is tied, the tiebreaker moves to the next category in the priority list. The priority list is determined in alternating order of preference, with the highest ranked remaining category being used and the higher seeded player going first.

Example:

I play Nick in the finals. Prior to the start of the first game, we each send a list to the commish labeled "Super Secret Tiebreaker Prefs"

Nick's list
1. ERA
2. OBP
3. SVHD
4. R
5. K
6. WHIP
7-10. I don't care

Matt's list
1. SLG
2. OBP
3. K
4. RBI
5. ERA
6. W
7. SB
8. R
9. WHIP
10. SVHD

Now, nobody needs to actually do the following work unless the tiebreaker comes into play, and in 99% of the cases we wouldn't even need to go past the first step, but for completeness we can determine the whole list by these preferences. It's applied in alternating order starting with Nick, who had higher seed. If something has already been picked, take the next highest. Therefore the tiebreaker list comes out as:

1. ERA (chosen by Nick)
2. SLG (Matt)
3. OBP (Nick)
4. K (Matt, skipping OBP)
5. SVHD (Nick)
6. RBI (Matt)
7. R (Nick)
8. W (Matt, skipping ERA)
9. WHIP (Nick, skipping K)
10. SB (Matt)

As I said, I'm offering this full example for completeness. In most cases it won't need to go beyond the first tiebreaker, which is selected by the team with the higher seed. For even more completeness we can stipulate that if no preferences are submitted by one owner, then the other owner gets their full list used, and if no preferences are submitted at all then the higher seeded team automatically wins the tiebreaker.


Under this system, the team with the higher seed still gets a significant advantage by being able to select the first tiebreaker, presumably picking the category in which they have the biggest advantage over the other team. The lists are made public, so both teams also know exactly what the tiebreaking procedure is at the outset.

I'm not sure about the math, but spitballing some previously cited numbers, let's say there's a 25% chance of a tiebreaker coming into play between closely-matched teams. Let's also say that the higher seeded team is able to designate a category that they would win 75% of the time. In this case we'd see a home field advantage of roughly 56% (.75 * .25 + .375). If the categories were all closer than 75%, home field advantage would be closer to 50%.