Where Do Stats Matter Most? Goalies?

[Goose]

After reading about twenty "omg, do you only look at the stats you idiot?" posts today, it got me thinking about when stats ARE the least, and conversely the most, relevant, and in which sports?

I have to think that, being a game of percentages, goaltending statistics have to be some of the more powerful statistics out there. Obviously you can't measure everything about a goalie's performance statistically, but it's awfully hard to make the argument that a sub .890 Save % and plus 3.5 GAA goalie is good in the same way that you can argue that a d-man with 100 PIM and 2 points in 80 games is good.

Anytime you repeat an action over and over and over stats become more powerful (shooting percentage in basketball being much more meaningful of a stat than it is in hockey for example), but some sports seem more inherently suited to statistical quantification than others.

Baseball has sabermetrics because so much of the game happens one on one and so much of it is quantifiable, whereas in hockey, being a team sport, and taking place five on five, much less of it is.

Which stats in hockey are the most representative, and how does statistic usage in hockey compare to other sports?

Click to read more...

 

[KidCanada*]

I actually find GAA useless. Generally better goalies will have a lower GAA, but if you coupe GAA with save %, GAA becomes completely meaningless.

I find save % to be the most useful stat. It's not foolproof, but it's the closest thing to it.

Goals are obviously very useful. Even if they're garbage goals, it still shows that youc can get into the dirty areas and get it done.

Faceoff percentage is a good one. It isn't foolproof because of match-up problems, but it's still very indicative.

As for overrated stats, look no further than wins for a goaltender. I don't really think I need to explain this one. It should be obvious.

Assists are a bit overrated. It's not a true representation of passing ability, especially when you throw secondary assists into it, even though you generally will see what many consider to be the best passers up at the top.

+/- is very overrated. It's a team stat, and doesn't consider match-up issues (ie. Pronger and Chara always going up against other teams top lines).

All that being said I think you can look at all stats within a certain context and make most of them somewhat useful (ie. analyze a tabbed puck-moving defensman's turnovers within the context of time on ice or something)

 

[The Fuhr*]

SV% is what I like

 

[joe89]

Goals, hits, blocked shots, faceoffs, takeaways and giveaways are rock solid stats in hockey imo. Assists are a bit behind but also informative.

+/- is only good if you compare the player with the team.

sv% is the best goalie stat, but requires a bunch of other factors to be taken into consideration. GAA & Wins for a goaltender are interesting but not very telling. Shutouts are flashy but don't mean much if the goaltender is not consistent and lets in 5 goals the night after. Goaltending stats are very subjective otherwise it would be very easy to crown a #1 goaltender year in and year out.

 

[KidCanada*]

After reading about twenty "omg, do you only look at the stats you idiot?" posts today, it got me thinking about when stats ARE the least, and conversely the most, relevant, and in which sports?

I have to think that, being a game of percentages, goaltending statistics have to be some of the more powerful statistics out there. Obviously you can't measure everything about a goalie's performance statistically, but it's awfully hard to make the argument that a sub .890 Save % and plus 3.5 GAA goalie is good in the same way that you can argue that a d-man with 100 PIM and 2 points in 80 games is good.

Anytime you repeat an action over and over and over stats become more powerful (shooting percentage in basketball being much more meaningful of a stat than it is in hockey for example), but some sports seem more inherently suited to statistical quantification than others.

Baseball has Sabremetrics because so much of the game happens one on one and so much of it is quantifiable, whereas in hockey, being a team sport, and taking place five on five, much less of it is.

Which stats in hockey are the most representative, and how does statistic usage in hockey compare to other sports?

I find the NBA has the most telling stats. FG % is a big one, and almost any stat combined with time played is telling.

The NFL not so much. So much depends on teams gameplans on how much weight to put into a certain stat. Even something like yards per carry for a RB isn't that useful as certain teams or backs are used in short-yardage situations as opposed to other teams that like to run sweeps which generally will pick up more yards.

MLB also not so much. Pitching statistics have many flaws in them. Batting average depends on if you're just a guy who wants to get on base, or if you're a power hitter. Same with HR's.

I think hockey might place second out of the big 4.

 

[Blades of Glory]

Every goaltending stat is flawed in a different way. Defensive play skews both GAA and save percentage, because a good defensive team will allow fewer high-quality chances and do a better job of clearing up rebounds, thus improving a goalie's save percentage, and in turn, influencing his goals-against. Shutouts are obviously tremendously influenced by defensive play, as well, and wins are a team stat.

If you want to know the leaders in save percentage in each post-lockout season, here is the rather under-whelming list. Cristobal Huet, Nicklas Backstrom, Dan Ellis, Tim Thomas, and Tuuka Rask. What is the common denominator between those goalies? Rask, Thomas, Backstrom, and Ellis play or did play in a decidedly defensive system. Cristobal Huet didn't even play 40 games during the year he led the league in save percentage, so we'll just leave him out of the discussion.

The truth is, stats are most unreliable when judging goaltenders, especially in recent years, considering the most "reliable" goaltending stat, save percentage, has been dominated by a plethora of not-so-elite goaltenders. The only way to judge a goaltender is to watch him play.

Giveaways are a completely useless statistic unless you apply the stat differently to every player, due to circumstances. Players, especially centers, that play a puck-possession style of game, are going to have a high giveaway total. That's why you see names like Crosby, Thornton, and Spezza dominate the giveaway list. Defensemen that are the primary puck-movers for their team are also going to be near the top of the giveaway list. That's why top defensemen like Andrei Markov and Scott Niedermayer have made the top 20 in giveaways quite often. It's when you see a stay-at-home defenseman like Hal Gill or Chris Phillips up there that you know something is wrong.

 

[KidCanada*]

Goals, hits, blocked shots, faceoffs, takeaways and giveaways are rock solid stats in hockey imo. Assists are a bit behind but also informative.

+/- is only good if you compare the player with the team.

sv% is the best goalie stat, but requires a bunch of other factors to be taken into consideration. GAA & Wins for a goaltender are interesting but not very telling.

For all of the things you listed above there's alot more you have to take into consideration. For blocked shots, do they get more PK time where you see a higher volume of shots? Are they used pimarily on hte PP where you would see next to none?

For giveaways, are they a player that is relied upon to moe the puck up ice, or is it a stay-at-home defenseman that will rarely be in a situation to turn it over relative to a puck mover? Of course pretty much everything has to be combined with time on ice too.

 

[Alexei Yashvalev]

Save Percentage gives a general idea, but has its problems since obviously a dump in and a breakaway count exactly the same. Some teams allow a lot of easy shots from the perimeter and others allow fewer but higher quality shots. It's not a perfect stat.

 

[KidCanada*]

Every goaltending stat is flawed in a different way. Defensive play skews both GAA and save percentage, because a good defensive team will allow fewer high-quality chances and do a better job of clearing up rebounds, thus improving a goalie's save percentage, and in turn, influencing his goals-against. Shutouts are obviously tremendously influenced by defensive play, as well, and wins are a team stat.

If you want to know the leaders in save percentage in each post-lockout season, here is the rather under-whelming list. Cristobal Huet, Nicklas Backstrom, Dan Ellis, Tim Thomas, and Tuuka Rask. What is the common denominator between those goalies? Rask, Thomas, Backstrom, and Ellis play or did play in a decidedly defensive system. Cristobal Huet didn't even play 40 games during the year he led the league in save percentage, so we'll just leave him out of the discussion.

The truth is, stats are most unreliable when judging goaltenders, especially in recent years, considering the most "reliable" goaltending stat, save percentage, has been dominated by a plethora of not-so-elite goaltenders. The only way to judge a goaltender is to watch him play.

What's your reason for that list being underwhelming though? You're disregarding save percentage in favour of something else, but what is it?

I agree that save % isn't foolproof though.

 

[The Waffler]

Shut-outs - That's how many games your goalie has won for you.

 

[Blades of Glory]

What's your reason for that list being underwhelming though? You're disregarding save percentage in favour of something else, but what is it?

I agree that save % isn't foolproof though.

It's an under-whelming list. Save percentage has been dominated by a rather pedestrian group of goaltenders. For that matter, so has goals-against average over the same time period, Kipprusoff excluded. That list is not exactly Brodeur, Luongo, Nabokov, Miller (though he is leading this year), and Kipprusoff. It goes to show you that even the most widely regarded goalie statistic does not really prove that much in terms of who the best goalie is.

There are no stats that can do justice to a goalie. Watching a goalie play blows away every statistic out there. Only then can you take into account the actual play of the defense and the goaltender.

 

[joe89]

For all of the things you listed above there's alot more you have to take into consideration. For blocked shots, do they get more PK time where you see a higher volume of shots? Are they used pimarily on hte PP where you would see next to none?

For giveaways, are they a player that is relied upon to moe the puck up ice, or is it a stay-at-home defenseman that will rarely be in a situation to turn it over relative to a puck mover? Of course pretty much everything has to be combined with time on ice too.

Of course eveything has to be used in appropriate forum. Blocked shots are usually used to compare defensive defencemen who always play on the PK. Giveaways are usually for playmakers, alright this is maybe not "rock solid" but still gives an idea what players that are not coughing up the puck as often.

 

[983 others]

Cups.

Everything else is lame fantasy garbage.

 

[KidCanada*]

Of course eveything has to be used in appropriate forum. Blocked shots are usually used to compare defensive defencemen who always play on the PK. Giveaways are usually for playmakers, alright this is maybe not "rock solid" but still gives an idea what players that are not coughing up the puck as often.

Yeah I agree. I think you can make any stat somewhat useful if you use it in the right context. Nothing is ever foolproof, but it can be telling.

 

[JalapenoPeppa*]

It is only save percentage. GAA is a team statistic.

 

[Nine to Five]

Cups.

Everything else is lame fantasy garbage.

Chris Osgood is on HFboards?

 

[tarlinian]

For all of the things you listed above there's alot more you have to take into consideration. For blocked shots, do they get more PK time where you see a higher volume of shots? Are they used pimarily on hte PP where you would see next to none?

For giveaways, are they a player that is relied upon to moe the puck up ice, or is it a stay-at-home defenseman that will rarely be in a situation to turn it over relative to a puck mover? Of course pretty much everything has to be combined with time on ice too.

The problem with most of the stats you list (outside of goals and assists), are home scorekeeper bias. Generally, speaking to get a reasonable hit, takeaway, giveaway, etc. count you have to look at road numbers only which cuts your sample size in half.

 

[Lario Melieux*]

 

[mimic]

It's all relative and dependent.

Anyone can skew stats to their favor so stats can't be the end all be all of every argument.

The best players have stats and intangibles both.

 

[rumrokh]

Save Percentage gives a general idea, but has its problems since obviously a dump in and a breakaway count exactly the same. Some teams allow a lot of easy shots from the perimeter and others allow fewer but higher quality shots. It's not a perfect stat.

Agreed. You can get an idea, but every popular goalie stat is just as much a team stat.

Popular hockey stats are very raw and are so situational it's ridiculous. As such, they have to be combined with watching the games.

I know people want to be able to claim that so-and-so is the best, but people are also buttholes. Disregard that need for objective certainty when you only catch one or two games a week outside of your favorite team. It's okay! I swear! Almost everyone here is already a much more dedicated fan than most hockey fans. It's not a competition of how much hockey you watch unless you're making claims that require that knowledge. So...just stop it.

There are potentially great stats that are not kept or are not popular, such as possessions after rebounds. But even that would be a combination of goalie rebound control and team ability/positioning. The cause could differ a lot and you'd have to watch games and combine your observations with the stats. But people want a shortcut.

 

[mimic]

Cups.

Everything else is lame fantasy garbage.

So Kirk Maltby is one of the greatest players to ever play the game?

 

[umwoz]

Giveaways are a completely useless statistic unless you apply the stat differently to every player, due to circumstances. Players, especially centers, that play a puck-possession style of game, are going to have a high giveaway total. That's why you see names like Crosby, Thornton, and Spezza dominate the giveaway list. Defensemen that are the primary puck-movers for their team are also going to be near the top of the giveaway list. That's why top defensemen like Andrei Markov and Scott Niedermayer have made the top 20 in giveaways quite often. It's when you see a stay-at-home defenseman like Hal Gill or Chris Phillips up there that you know something is wrong.

Great post. Looking at the top 10 now, Crosby, Ovechkin, Thornton, Myers are all in there. Why? Puck movers.

 

[MatthewT]

Wins

 

[HockeyGuy1985*]

Save % and GAA.

 

[mimic]

Wins

Osgood is one of the top10 greatest goalies of all time in your opinion?