Hall of Fame Committee Change

In addition to a veritable smorgasbord of outstanding, colorful, tragic and brilliant candidates the Magic Hall of Fame offers this year there is a revamped voting process. Instead of having 2 separate committees (the Selection Committee and Players Committee) whose votes were valued differently (67% and 33% of the total vote, respectively) as has been done in the past there is a single committee and everyone’s votes are counted equally. Any time there’s a change made by the WOTC brass, it’s worth examining the change’s merits.

I’ve already looked at how the 2 committee system behaved and for our purposes I’ll be using a lot of the same data for today’s discussion.

In addition to revamping the process, the actual members have been revamped. The former PC members got reduced with the change from a 100 lifetime Pro Point threshold to 150 Pro Points. Similarly, the SC members were reviewed and presumably reduced (to my knowledge there isn’t a public listing of members, but I’ve seen folks mentioning that they no longer have votes so it seems to be a safe assumption Pat Cox pointed me here. Thanks!). I don’t know if the ratios of PC vs SC are still the same as they have been, but I can at least look at what they were. Below are the number of votes from each committee from year to year (the actual committees were likely larger, with people neglecting to submit ballots).

Year	SC	PC	SC %
2006	81	75	51.9%
2007	92	82	52.9%
2008	100	71	58.5%
2009	113	81	58.2%
2010	129	102	55.8%
2011	127	111	53.4%
2012	140	107	56.7%
2013	144	122	54.1%

As you can see, the actual mix varied from year to year but was always within the 50-60% range for the SC. They way it worked was that for each player they would have a percentage of votes from SC members and a percentage of votes from PC members. Those 2 percentages would be weighted 2:1[i] to form a final percentage. So the actual mix above wasn’t terribly important. Having more members in a committee didn’t result in more or less influence on the output. If anything, having fewer people in the PC made some sense, as it meant each of those individual votes were more important to the 33% of the vote that committee could influence than if there were more voters.

Now, however, the mix does become relevant. At least to the extent that people believe one group or the other serves the game better. I’m not really of that mind, I think both sides have done a good job protecting the game’s legacy. It is worth noting that based on these numbers, former SC members now have a lesser impact on the voting process than they did in previous years, though it would still be larger than the PC.

That entire paragraph may all be for naught. I don’t know the number of members of each group, nor do I know the new criteria for non-player members so the ratio could be entirely different this year.

Those in the selection committee this year via being a player with 150+ Pro Points or via HOF credentials represent 42% of potential voters, which is largely in line with how things have been in the past.

After looking at the actual composition of the voting body I looked at voting results. I calculated how each player from previous years’ votes would have fared using the new non-weighted method. All previous percentages and number of votes are public in the archives of the Wizards site, so I just math’d[ii] my way to the number of voters for each committee and then re-math’d the new percentages. The good news is that there really wasn’t a huge difference between the weighted and non-weighted results, which is perhaps why it was decided to eschew a complex system for a simpler one. There’s a famous sentiment most commonly attributed to Einstein that says to make things as simple as possible, but not simpler. That seems to be the axiom here.

While there isn’t a huge difference, there is a difference. The average delta in percentage was .03% – basically nothing – with the higher percentage under the old methodology. The median was .04% – still nothing – but in favor of the new methodology. The extremes were Neil Reeves in2006 would have gotten 12.2% of the vote instead of the 8.8% he did get. The extreme in the other direction was Rob Dougherty in the same year. He would have gone from 38.2% to 32.7%. That year, there was no 40% threshold for election; the top 5 vote getters were elected regardless of total percentage. This result would not have changed the top 5.

The only person who was elected under the old rules but would not be under the new ones is Bram Snepvaners. He got 49.6% of the SC vote and 20.6% of the PC vote which resulted in 40.03% of the total vote. Under this year’s rules it would have resulted in 36.8% of the vote. The following year, only 3 people were elected in (Shuhei Nakamura, Anton Jonsson, Steve OMS) and all were relatively comfortable under the new rules (Steve was lowest at 47%). So it’s conceivable Bram would have been elected in 2011. Both Patrick Chapin and William Jensen were at 36.1% of the vote under the new rules that year and both ended up eventually being elected in. I’m not sure anything would have changed there but it’s possible a domino effect would have happened where Bram got momentum in 2011 which took votes from Steve OMS who then got elected in 2012 and took votes from other people and so on and so on.

I also looked to see if there was anyone who was not elected under the old rules who would have been under the new rules. There was nobody who would have been pushed up to the 40% threshold, but I did find something fun to look at. Take a look:

Year	Rank	Gamer	SC	PC	SC %	PC %	Tot %	2014 rules %	2014 Rank
2007	5	Randy Buehler	39	18	42.39%	21.95%	35.58%	32.76%	6
2007	6	Ben Rubin	34	26	36.96%	31.71%	35.21%	34.48%	5

 

So if we’re going to use some unholy combination of 2014 and 2007 rules where there’s only 1 committee but we’re somehow still electing the top 5 vote getters, Ben Rubin would have been elected in 2007 instead of 2008. Randy would have had to wait.

All told I think the change makes a lot of sense. I think people can grok it a lot better than the previous process while, at least historically, there isn’t a big change in the result. I’d like to see who is on the committee every ballot posted publicly, but that’s just me being greedy.

 

[i] Actually it was 67% to 33%. Famously, the 2010 ballot was initially announced with 39.94% of the total vote for Bram Snepvangers because it was calculated at a 2:1 ratio. But the rules were posted as 67:33 and when calculated that way it comes out to 40.03%. Bram was elected using the latter calculation since those were the published rules.

[ii] Unrelated to this entire practice, I just have to mention that recently at work I was able to answer a question with, ‘because Math’ and it felt great.