THE BOOK--Playing The Percentages In Baseball

One quick and easy fix for the NHL “incentive to tie” problem would be to remove the bonus point given for taking the game to overtime. Currently, the winning team gets two points and the losing team gets one point. If this were changed so that the winning team (in OT) got one point and the losing team nothing, there would be no reason to want overtime to occur. Critics of this could argue that this would increase emphasis on the shootout, which is not real hockey, but I would argue that it would cause the opposite. With any luck this would decrease the number of overtime games and therefore reduce the number of shootouts. No team would want to win in the shootout, because it would be one less point than winning outright is regulation. This idea is proven out in soccer, where a win is worth three points, a loss zero, and a tie one (not that they ever score in soccer, but that is a different discussion).

I’d say the changes in rules post-lockout, combined with the more stringent application of said rules, helped nudge things the right way. That is how we saw enforcers go the way of the dodo and how classical tough opposition / low scoring checking lines are also a disappearing breed.

So I guess some adjustments in the calling of interference and reining in the comeback of clutch and grab would help nudge things a little more.

Then, there is the fact that teams make do with the personnel they have on hands. Which is why Olympians can go further in that way: more stringent application of the rules, larger pool of players to choose from, total absence of monetary considerations. That’s a lot they have going for them.

How about splitting the NHL into two leagues, one with 12 teams, and the other with 24?

The 12 teams would mimic the talent level you see in the Olympics.

Basically, if you had a 12-team league, would a team REALLY want that checking guy or goon guy?

Minor penalties are 3 minutes (and a full 3 minutes...none of this Edmonton Oiler influenced penalty ends after a score business.)

I think that would greatly increase the value of skill players.

(Full) three minute minors.  Fascinating thought.

Is there going to be too much of a shift toward referees, knowing they have so much control over a game now?

Basically, referees would prefer to have enough chippy play so they can call penalties on both sides.  I am sure there is an evening-up.

Gabe, that something you can test?  Can you give us the frequency of single-team-game minor penalties?  For example,
1% 0 minors
5% 1 minor
10% 2 minors
etc

Given those frequencies, we can figure out the expected differentials in minors for each game, so say:
10% 3+ minors difference
20% 2 minors difference
40% 1 minor differece
30% equal minors

(All numbers for illustration.)

Then, we compare to the empirical.  I presume we’re going to see a huge difference.

Now, there could be a legitimate reason: chip play begets chip play.

You could try to include score of game and period.

And is it different between regular season and playoffs?

I did consider the possibility that if penalties are 3 full minutes, referees would simply call fewer penalties. But in theory, the league can control the behavior of referees if the league is so inclined.

My other thought, if you are one of the folks who wants to get rid of fighting, is that whenever a major penalty for fighting is called on your team, the other team gets to select a non-goaltender (ANY non-goaltender) to serve a game misconduct. So, when your goon gets into a fight, you lose your superstar in the process. You’d somehow have to mitigate a team trying to goad a player into a fight in order to eliminate an Ovechkin for example. But it would pretty much eliminate fighting.

Brian: oh, I like that.  You can make it so that it’s no so harsh by letting the opponent choose anyone that is currently on the ice (other than goalie).  Sure, it’ll be some 4th line guy, but it might be a top defenseman.

Tango,

0 mins difference = 15.6%
2 mins difference = 27.2%
4 mins difference = 20.5%
6 mins difference = 12.1%
8 mins difference = 6.9%

All others = 17.8%

This includes majors and misconducts.

Hawerchuk, can you also show just the frequency of minors (per team-game)?

Wow, Tango.  You just nailed the reason I don’t really watch hockey anymore.  It wasn’t the stoppage.

I only really enjoy playoff hockey.  Regular season games are often sluggish and, well, a bit boring at times.

I like the proposed changes to penalties, and I never liked the OT bonus point.  3-1-0.

Rob, could you clarify on the 3-1-0 part of your post?

3 - win
0 - loss
1 - is this tie at the end of regulation, OT win, or OT loss?

IIRC, they added the bonus point for the tie in an effort to reduce the amount of ties.  The idea was, if a team was guaranteed the point anyway, they’d “go for it” more in the OT since there was nothing to lose.  Of course, now that we have the shoot-out, there aren’t any more ties in hockey anyway.  And it is questionable if the “guaranteed point” has had the inverse effect, that there are now more regulation ties since teams play for that.

The variation of Rob’s suggestion I’ve seen is 3pts regulation win, 2pts overtime win, 1pt for regulation tie. 

Also, isn’t one of the biggest difference in Olympic Hockey the fact they use a bigger rink?

In the playoffs, a win is worth 1 (OT or no) and a loss is worth 0 (OT or no).  If we want playoff-style hockey, wouldn’t a good place to start be 1 point for a win and 0 points for a loss, whether in OT or not?

Olympic hockey uses NHL size arenas, at least in 2002 and 2010.  I don’t remember what they did in 2006, but I think it was NHL size as well, wasn’t it?

The 3-2-1-0 scoring system is used in this year’s Olympics.  I don’t remember what they used in the past.

The commentators said yesterday this was the first Olympics played on an NHL size rink.

I’ll get back to you on the penalties.  I need to scrape that data.

Aloysius, there is a lot of beauty to that approach. Imagine, hockey could have simple W-L records like baseball. That would be pretty.

In defense of 2-1-0, I was thinking two things.
1) There was a lot of controversy over having a shootout in the NHL because it was argued that a shootout isn’t hockey. Having a SO victory only worth one point should reduce the amount of unhappiness over a SO loss.
2) In terms of incentives, one point for an overtime win is only half as good as a regulation win, so both teams should want to avoid OT as much as possible and play aggressively near the end of regulation.

Brian K posted: “Minor penalties are 3 minutes (and a full 3 minutes...none of this Edmonton Oiler influenced penalty ends after a score business.)”

In fact, the rule to cancel remaining time on a minor after a PP goal was not implemented because of the Oilers. It was implemented before the 1956-57 season, before the Oilers were even a team!

Before 56-57, a team could score as many goals a as possible on a minor PP. The great Montreal Canadiens of the time were doing just that, pouring goals in with a magnificent power play that included defenseman Doug Harvey (along with Orr, arguably the greatest ever); forwards LW Bert Olmstead; center Jean Beliveau; and the incomparable Maurice “Rocket” Richard; and pointman forward Bernie “Boom Boom” Geoffrion.

In the playoffs, a win is worth 1 (OT or no) and a loss is worth 0 (OT or no).  If we want playoff-style hockey, wouldn’t a good place to start be 1 point for a win and 0 points for a loss, whether in OT or not?

Sure, but the wrinkle is what do do when it’s tied.  In the playoffs, the game goes on until a team wins, but this is a feature that most people like to keep exclusive to the playoffs.  Nobody is interested in a triple-overtime regular-season game.  So having a binary system just isn’t feasible since there will be ties (or SO w/l)

"The commentators said yesterday this was the first Olympics played on an NHL size rink.”

Really?  Salt Lake had Olympic sized arena?  I’m going to have to check that.  I don’t remember, but it’s been 8 years.

I very much remember Salt Lake being International sized.

Nice hockey game tonight (USA vs Canada).  Look forward to another possible rematch in the medal round.
vr, Xei

@ PaperBoyz: Thanks for that. I have no idea why I was under the impression that the rule was changed in the 80’s.

As for last night’s game...when can we start to acknowledge that it is not GIVEN that Canada has the best team in international competition? I don’t mean to say that the USA is better simply because they won last night. But it has been apparent to me for some time that there is much less distance between Canada and the other nations than has been assumed. Canada has the best talent suited to the NHL game. When you are playing the international game (subtle but important difference) the gap narrows and perhaps even closes completely with a couple of countries.

Am I wrong?

I presume the Vegas odds at the start of the tournament was something like this:
18% Canada
18% Russia
18% Sweden

12% Czech
12% USA
12% Finland

9% Slovakia

1% the rest

Something like that.

Basically, Canada has as much chance of winning the gold as the Yankees have of winning the World Series.

Given the current draw (Sweden/Russia/Canada on one side of the draw, USA getting Switzerland/Belarus in the quarters) should the US team be the gold medal favorites? I’m curious how the draw affects the odds.

The draws, presuming no upsets in the elimination round, all games on the same day, Wednesday (a GREAT day):

USA-Switzerland
Finland-Czech

Sweden-Slovakia
Russia-Canada

Let’s gives USA an 80% chance against Switzerland, then 50% against Finland and 50% against whoever comes out of the other group.  That sets the odds at 20%.

Let’s give Canada (and Russia) 50/50/50, for 12.5%.

Sweden gets say 60/50/50, for 15%.

Because Canada has to qualify againt Germany, you multiply their 12.5% by .95.  And, maybe because they are playing back-to-back, it should be: 95/45/50/50, for 10.7%.

Something along those lines…

Where those the Vegas odds? If so, I am stunned. Almost all informed mainstream commentary has made Canada out to be a prohibitive favorite.

I don’t mean to imply that the “informed mainstream” is to be taken serious...I’m simply using it as a proxy for the general consensus. Based on what I have heard or read, I would have assigned odds something like this:

Canada: 50%
Russia: 20%
Sweden: 10%
US: 5%
Field: 15%

My numbers are all for illustration.

In no way is Canada v Field 50/50 possible though.  If the tournament was ONLY Canada/Sweden/Russia, I don’t think that any of them would be at greater than 40%.  As you add more teams, that 40% has to drop.  Adding Finland means the best of the 4 is at say low 30%.  Adding USA means the best of the 5 is at high 20%.  Adding Czech and Slovakia, and now you are close to 20% as your best odds.

Absolutely agree that Canada @ 50/50 is ridiculous. That’s the point I was trying to make in my original post. The media carries on as though Canada not winning the gold medal would be a surprise.

As you know, there is a big difference between being the favorite within a population vs being the favorite against the field. Everyone acts as if Canada is the favorite against the field.

I googled the odds...one popular online betting site had the following odds listed before the tournament started:

Canada: Even money
Russia: 2 to 1
Sweden: 6 to 1
US: 6 to 1
Finland: 15 to 1
Czechs: 16 to 1

I agree with BrianK that the world in general gives the Canadian team far too much credit going into international competitions like this. Naturally the Canadian squad will be among the contenders, but the gap is much narrower than the general public (and the pinheads at NBC and other MSM) seem to think.

One factor not mentioned much in this thread is the sprint-vs.-marathon nature of the tournament. Like the NHL playoff tourney, a hot goalie can take a team a long way, which seems like a great leveler when you’re trying to set probabilities. Over 80 games, it would all even out, but over the six or seven it takes to win this thing, especially in lose-and-you’re-out mode, it might not. In that respect, it reminds me of the World Baseball Classic or whatever it’s called, where the US gets beat by a great pitching performance from Dice-K in a one-game situation.

Canada pretty soundly outplayed USA Sunday night (2:1 shot advantage, and the whole third period played in the USA end), but Ryan Miller made saves and Brodeur didn’t. In hockey, that’s the difference. Canada similarly outplayed Switzerland, but scraped out with a shootout win because their goalie Jonas Hiller also outplayed Brodeur.

Brian/29.

I added “30:1” for Rest (which includes Slovaks).  The total odds comes in at 127%.  So, that’s alot of juice.  If I multiply all the odds by 1.4 (Canada would be 1.4-1, Russia 2.8-1, etc), the total odds comes in at 100%.  Here’s what that looks like:
42% Canada
26% Russia
11% Sweden
11% US
5% Finland
4% Czechs
2% Rest

100%

That sure doesn’t look right.

Actually, Slovaks were 30:1 on the site...I just didn’t list them. The rest were all 1000:1 or more I think. Good guess.

Even removing the ‘juice’, as you did, has Canada at 42%. Wow. Lot’s of foolish money going there.

There’s no way you could justify placing money on Canada at those odds, even if you knew that Canada conclusively had the best ‘team’, because the outcome of any one game is so dependent on a single individual (goalie).

In order to be 42% favored, you have to have these odds:
bye: 100%
quarter final: 90%
semi final: 69%
final: 69%

That doesn’t make sense.

Russia makes more sense:
bye: 95%
quarter final: 74%
semi final: 60%
final: 55%

The top odds should have been close to 25%.  Very surprised…

Kind of off-topic here.

but here’s why olympic hockey was so great and the NHL is not so great..

US vs Canada in 2 games........11 total minor penalties called.

Flyers first 2 games after Olympic break........29 total minor penalties called.

It’s actually so very easy to fix: turn the league into 12 teams.  What you will have left is the high quality players. 

And you can also adopt the fighting rules at the same time.

And turn the league into a 48-game regular season.

30-second commercial break.

Basically, we all loved Olympic hockey, so all you have to do is match it.

You then make a Second-Division league of 20 more teams.  And a Third-Division of 30 teams, etc.  And you have relegation/promotion.  Up to two teams for First/Second, and up to four teams for Second/Third.

Love the idea Tango.Even a 20 team league would be better. Shorten the schedule. Adopt the relegation system like soccer.’nough sed.

How are you going to get 10,000 people in (insert non-Canadian city here) to show up for a Second-Division game? Assuming you can’t, then what we really are talking about is simple contraction, which is fine by me.

Another option, which is something I’ve talked about for years but have never found anyone else who liked the idea, is a team-sharing arrangement. Let’s pare the NHL back to 20 teams. The following cities maintain permanent franchises:

Montreal
Toronto
New York X 2
Philadelphia
Pittsburgh
Chicago
Detroit
Los Angeles
Vancouver
Boston
Colorado
Washington
Ottawa
Minnesota

That’s 15 teams. The remaining 5 teams rotate between cities. For example, you combine the Anaheim and San Jose franchises. In even years they play in San Jose (with a handful of games in Anaheim to maintain interest) and in odd years they play in Anaheim (w/handful of games in San Jose). Some possible combinations for the remaining teams:

Anaheim/San Jose
Calgary/Edmonton
St.Louis/Dallas
Colorado/You get the picture

I think the relative scarcity of games in the non-traditional cities would keep attendance high enough to support this. Perhaps you could include some kind of incentive that the number of games you get the following year is based on the attendance figures in the current season.

It’s probably too radical of a change to work. But since we are throwing suggestions up against the wall here…

I wanted the team-sharing idea for the Canada Expos.  One month in each town (say, Montreal, Ottawa, Hamilton, Calgary, Edmonton, Vancouver).  Schedule it so that they are out west when they play the NL West and home games in the east when they are on their NL East tour.

Playoffs will always be an issue.  You can award it to whichever city generates the most ticket revenue.