Tanking in Hockey (a proposal from the Sloan Sports Analytics Conference)

[The Legend]

I'm not sure if this research has been posted here before, but I thought I'd share it given that I feel its a worthwhile concept for preventing teams from tanking.

I'm sure some (if not many) of you are familiar with the MIT Sloan Sports Analytics Conference(SSAC), which occurs yearly and attracts the best and brightest minds in sports. Many conference attendees will present and discuss their statistical research showing different approaches to being successful. Hockey has been behind the other sports (especially baseball) in adopting some of these methods, but that's not really the discussion I want to get into.

We all know tanking is a problem. The researcher quoted in this article, Adam Gold, says historically teams winning percentage will decrease by 16% once they are eliminated from the playoffs. Thus, he argues for a new method of determining the draft order. Once teams are eliminated from the playoffs, their counter would start. The draft order would be determined based on points each team gets from its respective date of elimination until the end of the season.

Read the article here:

http://blogs.edmontonjournal.com/20...nt-fall-for-hall-and-fail-for-nail-campaigns/

Click to read more...

 

[Jumptheshark]

there are flaws in the theory--some teams are just that bad and no matter what they do they won win.

I have been in favour for years in redoing the draft order, I like the NBA where all non playoff team have a shot at first overall and would like the "ENTIRE" draft order done via the lotto and so a team could go from draft say first to 14th

 

[SidGenoMario]

Once you know you can't make the playoffs, you have no incentive to try hard.

 

[Freudian]

The main flaw with this theory is that teams get eliminated so late in the season, there is only ten or less games left. And anyone knows that in a ten game stretch anything can happen, no matter how hard you try.

I disagree with the notion that tanking is a problem. Teams don't tank. Some teams are just bad/unlucky/injury stricken/poorly constructed. The last place team in the league spent more money than they ever did. They added two big names in Carter/Wizniewski and still are last in the league.

There are a couple of cases in the last 30 years where teams might have been tanking (Penguins in the 80s and Ottawa in the 90s). But it doesn't really happen anymore. It's not a problem that needs solving. No matter what you do, you'll have a last place team every year.

 

[Lonewolfe2015]

There's way too many factors which go into determining why a team's winning percentage might decrease post playoff elimination.

Injuries, young callups, new coaches/systems, simply being terrible... etc.

It makes no sense to penalize a bad team further and force them to not get a top 5 pick.

What I might be in favor of is a tiered lottery system.

1-5 go into a lottery, order is shuffled based on a weighted drawing where you continue to draw until each team is picked. So right now, say the BJs get picked first, then the next 6 balls are all Columbus but the 7th is Carolina, they go 2nd overall. Until the top 5 are set.

Next is a weighted drawing, but lower percentage of weight, for the 6-10 tier. 11-15, 16-20, 21-25, 26-30. Etc.

That way you get blocks of up to 5 picks each team can move up in the draft, the weak can still get stronger and the stronger cannot become the strongest by shear luck.

 

[Brian28]

there are flaws in the theory--some teams are just that bad and no matter what they do they won win.

I have been in favour for years in redoing the draft order, I like the NBA where all non playoff team have a shot at first overall and would like the "ENTIRE" draft order done via the lotto and so a team could go from draft say first to 14th

The biggest problem with changing the draft system is accounting for legitimately bad teams. Who really believes Columbus would win enough games after elimination to retain the 1stoff overall they desperately need?

Tanking would still occur anyways.
If I know my team is a bubble playoff team but injuries have us on the verge of mathematical elimination maybe I lose 2 games then start trying when I'm working for the 1st overall. Any changes to a system designed to promote parity will have significant repercussions. Teams in less desirable locations could be put in as permanent bottom 10 teams only because playoff bubble teams push them from the picks they need to improveget and they are also unable to attract high end free agents.

 

[barneyg]

The main flaw with this theory is that teams get eliminated so late in the season, there is only ten or less games left. And anyone knows that in a ten game stretch anything can happen, no matter how hard you try.

This is a flaw but I don't think it's the biggest one. Under his proposal, the best way to get to #1 overall is to be eliminated as early as possible. That way, you can play more games that count toward the reverse standings. So instead of tanking the last quarter of the season, some teams would tank for the whole year.

 

[The Legend]

This is a flaw but I don't think it's the biggest one. Under his proposal, the best way to get to #1 overall is to be eliminated as early as possible. That way, you can play more games that count toward the reverse standings. So instead of tanking the last quarter of the season, some teams would tank for the whole year.

I don't think teams would necessarily tank the whole year. It would be incredibly stupid and difficult to start tanking from day 1 of a season.

Once you know you can't make the playoffs, you have no incentive to try hard.

Under the proposed system, you would have every incentive to try hard once you are eliminated from the playoffs

I disagree with the notion that tanking is a problem. Teams don't tank.

While I agree that it is difficult to accomplish, I think we are all naive if we think it doesn't occur in some fashion. Look at the statistics....teams are 16% worse once they are eliminated from the playoffs. That's pretty significant empirical evidence I'd say....at the very least it can't be ignored.

 

[FlyingPantherExtreme]

Oh no!!! Once eliminated losses go up by 16% thats like 1 in 6 games? Who cares. Guy with to much time on his hands trying to fix problems that don't exist. BIG NEWS: Someone in the NHL HAS to finish last every year. Someone HAS to have a bad record. That doesn't mean their tanking.

 

[Mayor Bee]

I touched on this briefly in my magic number thread, but the proposal leaves a couple of questions unanswered.

1) When is a team's elimination date?
2) Does the date matter more than the time?

For the first point, I'll use Columbus as the example. By losing on Sunday, Columbus was eliminated from the playoff hunt...but not really. It's possible for Columbus to finish with as many as 77 points, which kept them above 8th. However, making the playoffs was impossible because the two teams tied for 8th (76 points, Colorado and Calgary) would play each other before the season ends, so Columbus would automatically be eliminated at the moment that game is played. But that game is still several days off. Is Columbus' elimination date on March 11, or is it the day of that Colorado/Calgary game if Columbus somehow stays above water before then? That's a huge span of time, with multiple games in between.

For the second point, let's say that Team A is going to be eliminated if Team C beats Team D, regardless of whether Team A wins. The game between Team C and D begins at 7 PM, Team A and Team B at 7:30. The Team C/D game has an electrical failure that causes a mid-game delay of 45 minutes. Team A defeats Team B, and 5 minutes later, Team C defeats Team D (and eliminates Team A). Does Team A's win count in their post-elimination standings, even though their game started 30 minutes later and a strange situation caused the issue?

 

[Freudian]

While I agree that it is difficult to accomplish, I think we are all naive if we think it doesn't occur in some fashion. Look at the statistics....teams are 16% worse once they are eliminated from the playoffs. That's pretty significant empirical evidence I'd say....at the very least it can't be ignored.

Teams out of it might dump a UFA or two at deadline, which might make them slightly weaker. It's hardly tanking though, unless you dilute the term into meaninglessness.

The 16% number might look significant until you start to think about who they are playing down the stretch. They are playing against teams that have gone into playoff mode because they are trying to get into the playoffs/get a good seed. Points are harder to get for anyone in March-April and certainly will be so for the worst teams in the league. Add that the teams in the bottom five are among the most injury stricken (Oilers, Avs, Islanders have been recently) it is not surprising that the worst teams in the league struggle down the stretch. Interpreting it as them tanking is just silly.

Tanking is a non-problem.

 

[Kashie14]

Under the proposed system, you would have every incentive to try hard once you are eliminated from the playoffs

The fans do, the players not so much. They may not be around on the team next year/the first overall might knock them off the depth chart. This also hurts the trade deadline by forcing teams to retain UFAs and get nothing for them.

 

[Raccoon Jesus]

The biggest problem with changing the draft system is accounting for legitimately bad teams. Who really believes Columbus would win enough games after elimination to retain the 1stoff overall they desperately need?

Tanking would still occur anyways.
If I know my team is a bubble playoff team but injuries have us on the verge of mathematical elimination maybe I lose 2 games then start trying when I'm working for the 1st overall. Any changes to a system designed to promote parity will have significant repercussions. Teams in less desirable locations could be put in as permanent bottom 10 teams only because playoff bubble teams push them from the picks they need to improveget and they are also unable to attract high end free agents.

The LA Kings won themselves out of Steven Stamkos. True story. ****ing Brian Willsie.

But I generally agree.

 

[SmellOfVictory]

While I agree that it is difficult to accomplish, I think we are all naive if we think it doesn't occur in some fashion. Look at the statistics....teams are 16% worse once they are eliminated from the playoffs. That's pretty significant empirical evidence I'd say....at the very least it can't be ignored.

Put yourself in a player's shoes: you've spent over five months playing an average of around 3 games per week, constantly traveling and practicing in an attempt to make the playoffs/go on a cup run. You're probably tired, likely injured, and you get mathematically eliminated. Do you honestly think you're going to have the motivation to bust your ass for those final few games of the season? That's not intentional tanking, that's just relaxing your standards because the end result will be the same whether you win or lose your games.

Then these teams that are mathematically eliminated are going up against some teams that are in the middle of a hard competition for one of the final playoff spots, and you think they're going to win those games?

 

[eklunds source]

Once you know you can't make the playoffs, you have no incentive to try hard.

Who doesn't have incentive? The players?

"oh, I hope we finish last so we can get an elite talent who will fight me for my job..."

The coach?

"Last place sure isn't embarrassing, I'm confident I can finish last and not get fired.." (well, maybe Renney).

This isn't NHL12. Players hate losing - if you're not a competitive type of person, you don't make the NHL unless you have insane skills a la Daigle.

Players are working - its their JOB. And despite guaranteed contracts, not many want to uproot their family and move across the country because they were traded for mailing it in.. You think hockey players don't have pride? You think they HOPE their contract is bought out?

Put some freaking thought into it..

 

[CantTouchThis]

Who doesn't have incentive? The players?

"oh, I hope we finish last so we can get an elite talent who will fight me for my job..."

The coach?

"Last place sure isn't embarrassing, I'm confident I can finish last and not get fired.." (well, maybe Renney).

This isn't NHL12. Players hate losing - if you're not a competitive type of person, you don't make the NHL unless you have insane skills a la Daigle.

Players are working - its their JOB. And despite guaranteed contracts, not many want to uproot their family and move across the country because they were traded for mailing it in.. You think hockey players don't have pride? You think they HOPE their contract is bought out?

Put some freaking thought into it..

''Players want to win'' You think players don't realize when their team is simply not good enough? You think they want to go another season losing because they were heroes in the last 15 games of the year and screwed their chances of getting a lottery pick? ''players want to win''. The only ones that will be trying hard are the few fighting for a spot on the team next year. Other than that, they'll do what necessary to win at some point. No one wants to be on the oilers squad without hall, RNH, eberle... imagine if they made late heroics how crappy they'd be. Tanking is not as far fetch as you make it sound.

 

[Mayor Bee]

Teams out of it might dump a UFA or two at deadline, which might make them slightly weaker. It's hardly tanking though, unless you dilute the term into meaninglessness.

The 16% number might look significant until you start to think about who they are playing down the stretch. They are playing against teams that have gone into playoff mode because they are trying to get into the playoffs/get a good seed. Points are harder to get for anyone in March-April and certainly will be so for the worst teams in the league. Add that the teams in the bottom five are among the most injury stricken (Oilers, Avs, Islanders have been recently) it is not surprising that the worst teams in the league struggle down the stretch. Interpreting it as them tanking is just silly.

Tanking is a non-problem.

Not only that, but teams tend to be eliminated very late in the season. Even if everything goes against them, it could be more than halfway through March before the second team to be eliminated is actually eliminated. For the most part, we're comparing a 70-75 game sample size against a 7-12 game size and calling it statistically significant. I don't think that's accurate in the least.

Even Columbus, which has lagged well behind the league all season, wasn't eliminated until Game #69. And there's some debate as to whether or not they've truly been eliminated, as I outlined above.

 

[x Tame Impala]

I don't like that STL would have gotten the 1st overall pick (Nugent Hopkins) because they went 3-0 the last 3 games of the year.

It completely goes against team rebuilds and makes it impossible for Columbus and Edmonton to get out of their perpetual cellar.

 

[Pilky01]

I'm not sure if this research has been posted here before, but I thought I'd share it given that I feel its a worthwhile concept for preventing teams from tanking.

I'm sure some (if not many) of you are familiar with the MIT Sloan Sports Analytics Conference(SSAC), which occurs yearly and attracts the best and brightest minds in sports. Many conference attendees will present and discuss their statistical research showing different approaches to being successful. Hockey has been behind the other sports (especially baseball) in adopting some of these methods, but that's not really the discussion I want to get into.

We all know tanking is a problem. The researcher quoted in this article, Adam Gold, says historically teams winning percentage will decrease by 16% once they are eliminated from the playoffs. Thus, he argues for a new method of determining the draft order. Once teams are eliminated from the playoffs, their counter would start. The draft order would be determined based on points each team gets from its respective date of elimination until the end of the season.

Read the article here:

http://blogs.edmontonjournal.com/20...nt-fall-for-hall-and-fail-for-nail-campaigns/

I am in favour of any proposal that punishes losing as opposed to rewarding it.

The way North American entry drafts work is an affront to the ideals of competition and sportsmanship.

 

[sharks9]

A big reason for the drop in winning percentage late in the season is the fact that teams that know they won't make the playoffs often trade away some of their good players at the deadline for younger players and draft picks. This makes them lose more games so it seems like they're losing on purpose.

 

[engy]

I frankly doubt players care enough about tanking so that another player next year gets drafted and takes their spot. They probably aren't as motivated and rightly so among other reasons.

Fans and GMS care but when you are an UFA, why would you want to play bad?

 

[The Legend]

The 16% number might look significant until you start to think about who they are playing down the stretch. They are playing against teams that have gone into playoff mode because they are trying to get into the playoffs/get a good seed.

I agree that teams that are in playoff mode might try harder than those who are mathematically eliminated, but why are we accepting this as a given? If you give the bottom teams something to play for then the games will become more competitive.

Add that the teams in the bottom five are among the most injury stricken (Oilers, Avs, Islanders have been recently) it is not surprising that the worst teams in the league struggle down the stretch. Interpreting it as them tanking is just silly.

Tanking is a non-problem.

I'll ignore the fact that the Avs aren't in the bottom five right now....

Are you saying that teams in the bottom five are among the most injury stricken this year? Obviously this year's data would not have gone into the research....

If you are saying that teams in the bottom five are historically more injury stricken (and thus ended up in the bottom five to begin with) this would have obviously affected their complete season's results. It would not necessarily cause a drop in performance in the last dozen or so games.

 

[Mayor Bee]

If you are saying that teams in the bottom five are historically more injury stricken (and thus ended up in the bottom five to begin with) this would have obviously affected their complete season's results. It would not necessarily cause a drop in performance in the last dozen or so games.

A player who's fighting an injury late in a non-playoff season is more likely to be shut down for the rest of the year than if the team is still in the hunt. One of the side effects of implementing this idea is that injured players would be forced back into the lineup, all to possibly get a better draft pick.

I don't think that's right at all. A team is being asked to choose between the health and future health of a current player versus the ability to pick a certain prospect.

 

[Finlandia WOAT]

''Players want to win'' You think players don't realize when their team is simply not good enough? You think they want to go another season losing because they were heroes in the last 15 games of the year and screwed their chances of getting a lottery pick? ''players want to win''. The only ones that will be trying hard are the few fighting for a spot on the team next year. Other than that, they'll do what necessary to win at some point. No one wants to be on the oilers squad without hall, RNH, eberle... imagine if they made late heroics how crappy they'd be. Tanking is not as far fetch as you make it sound.

Players that do their job badly don't last long in the NHL.

You want to take advantage of every opportunity to play well so that when your 2 way deal runs out and the Oilers/Blue Jackets/'Canes are saying "Sorry, too expensive", someone out there will take a flier on you based on what they saw (case in point- Tim Brent and the Hurricanes).

This is a League of players who are 3rd/4th line tweeners, who may or may not have a job next year.

Why the hell would they want to risk their job security so that the team will be better off in a couple years?

 

[Ashasx]

My god. That proposed draft order system is fantastic.

The crappy teams still have the highest chance of picking #1 because they're eliminated sooner, but at least there will be much more reason to watch hockey after playoffs are out of the picture.