Mathematical Elimination Draft Orders

[Winning Unlimited]

[youtube][/youtube]

I presented at the MIT Sloan Sports Analytics Conference last month.

http://www.sloansportsconference.com/?p=5496

Abstract
Encouraged by fans’ desire to cheer passionately for their favorite teams every game of the season, I respectfully propose that draft order should be determined by performance after mathematical elimination. Analysis of historical data suggests optimistic statistical methods based on functions of winning can end tanking in sports. I hypothesize that when the incentive for teams to intentionally handicap themselves is removed, the expected elimination draft order is highly correlated with the expected reverse standings. The risk of allowing a poorly performing team to lose the top draft pick is outweighed by the benefit of eliciting intense competition, inspiring fans’ interest, and preserving sports’ integrity. If a championship can be awarded based on the result of a subset of games that have injuries, variability, and inequities, then draft order can be determined with comparable dynamic mechanics. This method of holding professional franchises accountable justifiably quantifies a minimum level of success that teams must possess. The number of games awarded to earn the first pick is dependent on parity and competitive balance. With warm personal regards and best wishes for a successful conference, I am proud to submit enriching research that has the potential to initiate an era where teams make the pursuit of winning unlimited.

Slides
http://www.sloansportsconference.com/wp ... nking1.pdf

If you would like to read more, here are a few articles written on this topic in the past month, starting with the beat writer for the Edmonton Oilers, David Staples.

http://blogs.edmontonjournal.com/2012/0 ... campaigns/
http://www.grantland.com/story/_/id/775 ... icture-nhl
http://www2.macleans.ca/2012/03/04/the- ... hemselves/
http://blogs.thescore.com/mlb/2012/03/0 ... nk-nation/
http://thewhitemamba.wordpress.com/2012 ... r-tanking/
http://www.montrealgazette.com/sports/T ... story.html
http://sports.yahoo.com/nba/teams/cha/blogs
http://www.pressherald.com/sports/MIT-h ... obile&id=2
http://nba-point-forward.si.com/2012/03 ... n-tanking/
http://hfboards.hockeysfuture.com/showt ... p=45512135
http://hfboards.hockeysfuture.com/showt ... ?t=1126719
http://oilfans.com/forum/index.php?t=ne ... msg_613005
http://espn.go.com/blog/truehoop
http://bbs.clutchfans.net/showthread.php?p=6670568
http://baseballmusings.com/?p=79569
http://hfboards.hockeysfuture.com/showt ... ?t=1126719

[Winning Unlimited]

I only bring this up now, because there is an interesting dynamic to tonight's games.

With every team having 1 game left to play, I've attached the current proposed draft order.

We need Edmonton to beat Vancouver for a chance at West #1.
Under the current system, if Edmonton wins, they can fall from pick 2 to pick 3.
Under my system, if Edmonton wins, they can catch Columbus and get the first overall pick.

[cardsfan04]

Did you post this last year? It sounds really familiar.

It's intriguing, but I'm not sold on this. I know it makes sense to us fans for a team to tank games to get a higher draft pick. But, I don't think it happens that much. IMO, the purpose of reverse record draft order is to promote parity. Taking the top pick away from teams because they can't win games after being eliminated seems to work against that. I get that it's designed to still allow the worst teams the biggest chance at top picks. But, if they're bad enough to be the worst in the NHL, they might lose those extra games by sucking, not by tanking.

[Winning Unlimited]

Did you post this last year? It sounds really familiar.

It's intriguing, but I'm not sold on this. I know it makes sense to us fans for a team to tank games to get a higher draft pick. But, I don't think it happens that much. IMO, the purpose of reverse record draft order is to promote parity. Taking the top pick away from teams because they can't win games after being eliminated seems to work against that. I get that it's designed to still allow the worst teams the biggest chance at top picks. But, if they're bad enough to be the worst in the NHL, they might lose those extra games by sucking, not by tanking.

The minimum cap in the NHL is $48.3 million. Fans are asked to pay that, plus the cost to open the doors. So, $70M+ per team per year. Every team should have a reasonable capacity to win any game. There are ways to keep teams from slipping too far, such as only awarding the first 3 picks from mathematical elimination. But in the NHL, there is very good parity, and the full model should be used.

[cardsfan04]

I guess my biggest issue is that I don't see teams tanking. The best way to look at that is PPG before elimination vs after elimination.

Here are the teams (listed in projected NHL draft order). PPG before elimination/PPG after elimination

Columbus: 0.75/0.92
Edmonton: 0.89/1.11
Montreal: 0.93/1.00
NY Islanders: 0.99/0.67
Toronto: 0.97/1.25
Anaheim: 0.99/1.00
Minnesota: 0.96/1.50
Carolina: 1.00/1.33
Tampa Bay: 1.01/1.00
Winnipeg: 1.03/1.00
Buffalo: 1.09/NA
Colorado: 1.09/NA
Calgary: 1.08/1.50
Dallas: 1.10/NA

Of those 14 teams, 3 of them have had no games since being eliminated yet have their draft position affected. Of the 11 teams that have played since being eliminated, 8 of them have earned a higher PPG than prior to elimination which suggests the opposite of tanking. Of the 3 that have lower PPG 2 of them went from 1.03 and 1.02 PPG to 1.00 PPG which essentially the same thing. The Islanders are the only team that has done significantly worse since elimination and they have a whopping 3 games. If they win their last game of the season, they will move into the group of 8 that has improved their performance after elimination.

This is only one season and no team has a great sample size. But, collectively, they have 55 games played since elimination. 55 games isn't a ton either, but you generally know how good your team is by game 55 in a season, so it does have some significance. I'm not going to take the time to see what their collective PPG was before elimination and their collective PPG is after elimination, but I am certain that they have performed better as a group. So, basically, I just see this as an intriguing idea (it really is interesting) to fix a problem that I don't see evidence of existing.

[Winning Unlimited]

Last year, performance after elimination dropped 35%. (Pre-elim 45% to post-elim 27%)

Don't overlook fan perception, either. Draft order methods should be structured to give teams that need help an advantage, while still keeping fans interested in their teams. The worst teams in the league would sell a lot more tickets with my method.

[cardsfan04]

That's a fair point. I'm not totally against that method. I'm just not completely convinced of the necessity to change either.

[Winning Unlimited]

edited the first post by adding youtube discussion.