Question to advanced stats proponents

[Bear of Bad News]

Ah, I see - you're saying that you see more hockey lines at -300 or more than in baseball.

This has not been my experience, although I have not studied it to any extent.

Rest assured that I need no education on the subject of how sports betting works; I'm actually writing this from the sports book at Mandalay Bay.

 

[Mathletic]

Ah, I see - you're saying that you see more hockey lines at -300 or more than in baseball.

This has not been my experience, although I have not studied it to any extent.

Rest assured that I need no education on the subject of how sports betting works; I'm actually writing this from the sports book at Mandalay Bay.

Just to give you an idea, in the last full NHL regular season (2011-2012) there were 22 games rated at -300 or lower (lower being -350 for example). While in the last 4 MLB seasons (2010-2013), there has been a total of 12. This pattern can be generalized to odds in general. For example, the average favorite in hockey is rated at -150 while in baseball it's -110. Random figures here.

 

[Bear of Bad News]

That does stand to reason - even a typical "top" baseball team (100 wins) only wins 61% of their games.

 

[EVBetting Site]

By the same vein, a top hockey team (100 points) only wins around ~47-50 games (~57-61%) as well.

 

[The Mighty Duck Man]

Blame Moneyball

People were so focused on the numbers they forgot that baseball numbers make sense because it's pretty much an individual sport whereas hockey depends on so many variables that the math is impossible.

It's actually pretty crazy how few people realize how little "Moneyball" actually had to do with advanced statistics itself. I used to cringe every time someone like Mirtle would use the term "moneypuck", just brutal.

 

[here come the]

I think stats like CORSI and FENWICK can be helpful when considered with some context, but aren't they really team-generated numbers?

How comfortable are we using possession stats to judge individuals when they can be largely dependant on who you play with or the situations you're put in?

Can't the same be said of every hockey statistic? It really is the ultimate team sport (in NA at least). PPP are highly dependent on the position coaches put certain players and where the system is designed to shoot the puck from, IMO points are also a statistic that without context can be very misleading as well. I love and am intrigued by the movement to understand hockey better with use of numbers, but part of me hopes we never really find any of the detailed, precise measurements that they have in baseball because I enjoy the intangible factors of the game.

 

[Fish on The Sand]

By the same vein, a top hockey team (100 points) only wins around ~47-50 games (~57-61%) as well.

100 points in hockey is much easier to achieve than 100 wins in baseball.

 

[hatterson]

By the same vein, a top hockey team (100 points) only wins around ~47-50 games (~57-61%) as well.

In terms of rarity a 100 win MLB team is about on par with a 120 point team in the NHL

In the 7 full seasons since the 04-05 lockout there have been 2 120+ point teams, 6 115+ point teams, 18 110+ point teams and 62 100+ point teams.

In the last 7 MLB seasons there have been 3 100+ win teams, 25 95 win teams, and 57 90 win teams.

 

[EVBetting Site]

In terms of rarity a 100 win MLB team is about on par with a 120 point team in the NHL

In the 7 full seasons since the 04-05 lockout there have been 2 120+ point teams, 6 115+ point teams, 18 110+ point teams and 62 100+ point teams.

In the last 7 MLB seasons there have been 3 100+ win teams, 25 95 win teams, and 57 90 win teams.

Because of the intro of the loser point. An OT/SO loss is still a loss from a gambling perspective. You have to eliminate OT/SO loss and count it as zero for this exercise.

100 point team in terms of win% is dependent solely on wins/total games.

So a 50 win team is only slightly more rare than a 100 win team in MLB.

 

[hatterson]

Because of the intro of the loser point. An OT/SO loss is still a loss from a gambling perspective. You have to eliminate OT/SO loss and count it as zero for this exercise.

100 point team in terms of win% is dependent solely on wins/total games.

So a 50 win team is only slightly more rare than a 100 win team in MLB.

In those same seasons there have been 24 50 win teams. 100 wins would likely be equivalent to 54 or 55 wins.

100 wins = .6173 win percentage
54 wins = .6585 win percentage
55 wins = .6707 win percentage

Highest number of wins since the lockout, 58 by Detroit in 05-06 = .7073
Highest number of wins in recent MLB seasons, 105 by St. Louis in 2004 = .6481

That .6481 would be 53 wins in the NHLs 82 game season and every year since the lockout, except 11-12, a team has won 53 or more games. 2x53, 3x54 and 1x58.

The top teams in the NHL win a larger percentage of their games than the top teams in the MLB. In fact, the best team in the decade in the MLB is roughly equal, in terms of winning percentage, to the normal best team in a given NHL season.

Edit: Similarly, the worst teams in the NHL are worse than the worst in the MLB when graded by win percentage, although not by as much.

 

[Ho Borvat]

I have literally yet to see an advanced stats proponent that used the stats simply as a guide. It seems like you guys seem to use it as black and white and seem to think that common stats and legitimate watching of the game is not important.

I've said it before and I will say it again, bad shot differential is no different than bad goaltending.

You can still win games with sub.900 goaltending... but your fighting an uphill battle and you would be way better off with .920+ goaltending.

If your constantly getting outshot, you can still win games... but its definitely an uphill battle and you would be way better off to outshoot the other team.

 

[Master_Of_Districts]

I've said it before and I will say it again, bad shot differential is no different than bad goaltending.

You can still win games with sub.900 goaltending... but your fighting an uphill battle and you would be way better off with .920+ goaltending.

If your constantly getting outshot, you can still win games... but its definitely an uphill battle and you would be way better off to outshoot the other team.

That would depend on whether you're referring to a team's results or its underlying true talent.

If you're referring to underlying true talent, then, yes, the effect of poor goaltending is analogous to the effect of having a poor shot percentage/ratio - both handicap a team but can be overcome in other areas of the game.

If you're referring to results, then the two are not comparable in the same way.

A team that posts, say, a 0.890 save percentage at the halfway point in the season is not likely to have a true talent save percentage of 0.890. Their true talent save percentage is likely to be much closer to the league average.

The same is not true of shot percentage/ratio. A team that posts a shot ratio of 0.450 halfway through the year is likely to have a true talent shot ratio near that figure. The effect of random variation on shot ratio/percentage is minimal over ~1500-2000 shots, so you won't expect to observe much regression toward the league average.