So basically if you live in Arizona you have a better chance of winning Powerball than any of your teams winning a draft lottery.— Luke Lapinski (@LukeLapinski) May 17, 2017
Isn’t there a better way for leagues to handle draft order?
By now many Coyotes fans are familiar with an idea that Captain Shane Doan has popularized to change the draft lottery. It’s been discussed in articles, on the radio and on podcasts. Many fans may not be aware that he was not the first to suggest it.
In the wider hockey analytics community, the draft order system the captain has endorsed is known as the ‘Gold Plan’ or ‘Gold Drafting’. It was presented at the 2012 Sloan Analytics conference by Adam Gold as a method to eliminate tanking. Basically, the method goes like this: once a team is mathematically eliminated from the playoffs they would begin accumulating points for each win. At the end of the season, the team with the highest point total would be awarded the first overall pick and so on down the list of point accumulators. It would serve the dual purpose of removing the incentives for tanking and keeping the season competitive for eliminated teams.
There are many proponents of the system besides Doan that feel very strongly that Gold Drafting should be implemented in the NHL. One such person is Micah Blake McCurdy, the brain behind HockeyViz.com and who frequently posts charts of various statistics on his site and twitter. He also tracks what the draft order would be if Gold Drafting were implemented, with this year’s shown below.
With Gold Drafting:— Micah Blake McCurdy (@IneffectiveMath) April 10, 2017
1st overall pick: WPG, 12 Gold points
2nd: ARI, 9
3rd: COL, 7
4th: DAL, 6
5th: L.A, 5 (in 4 gms)
6th: DET, 5 (in 6 gms)
There are some common criticisms of the plan that are sometimes raised. Doesn’t it affect league parity? Doesn’t it hurt weak teams that can’t win games? Won’t teams just start tanking sooner? Luckily, McCurdy has also taken on the task of answering these questions.
In 2016 he participated in Rochester Institute of Technology’s Hockey Analytics Conference and presented his work attempting to answer these questions.
McCurdy ran several simulations to see how the Gold Draft system compares with the current lottery-based system, both with and without tanking. His probability model found that Gold Drafting would affect league parity only slightly more than the lottery system without tanking and slightly less than the lottery system with tanking. The differences are small enough to be considered negligible at best.
Next, he ran a simulation looking at how Gold Drafting would affect weaker teams. It showed that these teams would take approximately 5 to 6 years of receiving high draft picks before they would get better. This falls in line with the real world evidence of teams in the current system taking approximately that long to get better under the current lottery system.
As for the idea that teams would tank sooner, well that’s a bit more difficult than one might think. It’s really hard to know when a team will be mathematically eliminated until the games are played because their position not only depends on their own wins and losses but also the wins and losses of other teams in their conference. So, as an extreme example, assume that a team can’t make it into the playoffs with less than 87 points. (That’s the lowest number of points that a team has made the playoffs with since the wildcard format was started in the 2013-14 season.)
If a team was deliberately tanking and losing all of their games in regulation, they wouldn’t be eliminated until after their 39th game, almost half the season. Under Gold Drafting at that point, in order to get the first overall pick, the team would have to then turn around and start winning games, flipping that switch would not be easy. This is not even considering that losing 39 games in a row would be grotesquely obvious and likely the league would take steps.
So, how much sooner could a team plan to lose in order to better their chances at a Gold Draft?
Yeah, McCurdy has a graph for that too, but it’s a bit confusing. The most important takeaway from this chart is that he found that the difference between teams that are deliberately tanking (the dotted line) and the same teams not tanking (the solid line). At best is a gain of one or two extra games. That isn’t even worth the time it takes to calculate how soon your team can be mathematically eliminated.
There is one area where the current draft lottery system beats Gold Drafting hands down and without question: The drama! In the Gold system, every team would have a fairly good idea where they would be selecting players in the last few days of the season. No more chances for teams to jump up in the system. No more televised lottery night with GMs waiting to around to find out if they won or lost. No more top draft prospects looking disappointed when a team they didn’t expect gets their pick. And perhaps worst of all no more draft lottery memes! Poor Taylor Hall would have to update his hockey resume again.
Officially adding "NHL lottery ball specialist" to my hockey resume.— Taylor Hall (@hallsy09) April 30, 2017
Since we know the effects of parity, weak teams and tanking are no worse than the current system. It might be worth giving all that up just to have an 82 game season where all 82 games mattered for every team and every fan base.