How to Structure the Ballot?
We decided this already:
Voters must vote for their top 10 candidates in order. We aren’t sure how to award the points.
In the extended text, I’ll post the last two posts from our earlier discussion.
I kind of like DanG’s idea and JimD points out that we aren’t splitting atoms. I guess we have to make a choice here, so let’s start moving in that direction . . .
Posted 2:34 p.m., January 16, 2002 - DanG
I thought I’d weigh in on the topic of MVP-style voting points.
I’ve always had a small problem with the 10-9-8-7-6-5-4-3-2-1 and similar formats used, because the point totals achieved by players relate poorly to the players’ actual values. The #1 man is not usually twice the value of the #6 man; the #9 man is not twice the value of #10; The #3 man is not four times the value of the #9 man, etc.
The hill is too steep. I’m thinking more along the lines of a distribution starting like this: 5-5-4-4-3-3-2-2-1-1. Along with this, 5 bonus points are to be distributed, no more than 2 to any one player, for a total of 35 points on each ballot.
The top-loaded extreme looks like this: 7-7-5-4-3-3-2-2-1-1.
The bottom-loaded extreme is: 5-5-4-4-3-3-3-3-3-2.
The tilt-the-ballot-to-one-man ballot is: 7-5-4-4-3-3-3-2-2-2. This one especially is reminiscent of a typical TPR leaders list.
The tilt-the-ballot-to-two-men ballot is: 7-7-4-4-3-3-2-2-2-1.
I think this gradual point distribution enables voters to better reflect the relative values of the candidates they’re voting for. There just is not that big a difference between the candidates in most years.
It also lessens the impact of a top vote. It’s much harder to skew the results to My Favorite candidate. A high point total depends more on the consistency of a player’s high ranking on the ballots cast. You would avoid an outcome such as the 1979 NL MVP vote, where the madcap Stargell-for-#1 voters pushed him into a tie with the more-consistently supported Keith Hernandez.
Posted 6:11 p.m., January 16, 2002 - jimd
The “these are my 5 guys ballot”
6-5-5-5-5-3-2-2-1-1 instead of 12-12-12-12-11-6-4-3-2-1
DanG: actually, the impact of a top-vote is the same in a 7 out of 35 system as in a 15 out of 75. You’re getting 20% of the points.
Scaling them to a common denominator (525 points):
(75) 15-15-14-8-7-6-4-3-2-1 becomes 105-105-98-56-49-42-28-21-14—7
(35) -7—7—5-4-3-3-2-2-1-1 becomes 105-105-75-60-45-45-30-30-15-15
(75) 10-9-9-8-8-7-7-6-6-5 becomes 70-63-63-56-56-49-49-42-42-35
(35) -5-5-4-4-3-3-3-3-3-2 becomes 75-75-60-60-45-45-45-45-45-30
I can get the same effect by requiring 10th place to get 2 points on the 75 point scale. I don’t think it makes a big difference.
Posted: April 03, 2002 at 03:49 PM | 9 comment(s)
Login to Bookmark