Advanced Stats

[stan the caddy]

I think PFF is one of the biggest shams out there. A bunch of guys who admit to not having scouting experience watch highlights and assign an arbitrary grade to indidual plays. Then people on the internet act like you're an idiot if you don't don't subscribe to it.

 

[Ail]

I don't think that most fans think the stats are useless. People just say that they're not the be all end all. Yet the advanced stat guys tend to SAY the same thing, yet every argument they have is "CORSI brah".

Yeah and then the eyeball guys come in and every argument they have is, "OMG HENRIK LUNDQVIST SOFTIE, BRAN BOYLE SUK! GIRARDI SO DUM!"

Very convincing.

 

[Ail]

Yeah, its a good thing those advanced stats came along, because nobody knew what they were watching for 80 years.

People knew they were watching the sun rise and set everyday for millions of years. Now we know the exact time it will every dusk and dawn, how far it is from Earth, how hot it burns, what it is made of, the kind of light it puts out, how fast that light travels, so on and so forth.

Sorry that your eyeballs might not be the most reliable source of unbiased information anymore.

 

[pld459666]

People knew they were watching the sun rise and set everyday for millions of years. Now we know the exact time it will every dusk and dawn, how far it is from Earth, how hot it burns, what it is made of, the kind of light it puts out, how fast that light travels, so on and so forth.

Sorry that your eyeballs might not be the most reliable source of unbiased information anymore.

and I tell THOSE people to ****, put the lap top down and enjoy the ****ing sunrise.

 

[Ail]

and I tell THOSE people to ****, put the lap top down and enjoy the ****ing sunrise.

Okay that's fine, but don't try and argue with them about how the sun operates when you're only looking at it with your eyes.

 

[Tawnos]

Shots come in all shapes and sizes, agreed.

Shots while McDonagh is on the ice come in all shapes and sizes. Shots while John Moore is on the ice come in all shapes and sizes. Given a sample large enough, both collections of shots will be pretty similar.

Don't believe me? Here's a list of the Rangers' defenseman from 2012-13 listed in order of on-ice save percentage:

Does that look more like a list of the Rangers’ best defensive defensemen listed from top to bottom, or like a list of Rangers’ defensemen put in random order

Show me a shot quality heat map for each defenseman when they're on the ice and another for when they're off and we will talk. Someone is going to quantify that eventually, and work on it has already begun, but it's not there yet.

The problem I have with that stuff is that it's all negative quantifiers and doesn't actually account for the positive things a player does. Yes, being great positionally and making a lot of good defensive plays will be reflected on the other side of the ledger, but it's just a reflection and, like all reflections, is imperfect.

Advanced stats are great and are important, but a solid eye remains more important.

 

[Hire Sather]

Okay that's fine, but don't try and argue with them about how the sun operates when you're only looking at it with your eyes.

Using advanced stats doesn't make you more qualified to talk hockey. Think what you want, though. I guess its an easy way to try and feel like you have a leg up.

 

[Ail]

Using advanced stats doesn't make you more qualified to talk hockey. Think what you want, though. I guess its an easy way to try and feel like you have a leg up.

Doesn't make you more qualified but it certainly gives you a perspective the people looking at it with just omg eyeballs don't have.

It's not about being superior, it's about using all the tools available to you to make an informed decision or opinion on something.

 

[aufheben]

Using advanced stats doesn't make you more qualified to talk hockey. Think what you want, though. I guess its an easy way to try and feel like you have a leg up.

Additional information doesn't make you more knowledgeable on a subject? I really don't care, just playing Doval's advocate.

 

[mike14]

Doesn't make you more qualified but it certainly gives you a perspective the people looking at it with just omg eyeballs don't have.

It's not about being superior, it's about using all the tools available to you to make an informed decision or opinion on something.

Completely agree, stats are just one of the many things that should be taken into consideration when evaluating hockey (or anything else).

On the bolded, I'd say that on this site (in general not just NYR) a lot of the time it does seem to be about being superior. A lot of attitude seems to be "I understand advanced stats better than you so I'm better and my opinion is more valid". Not everyone is like that, but there are quite a few.

Personally, I don't use advanced stats much. I enjoy watching the games, talking about it afterwards, but really don't feel that that type of analysis will make me enjoy the game any more than I already do, but, as I stated at the start, stats are always a useful tool if used correctly (and not just on their own)

 

[BlueshirtBlitz]

and I tell THOSE people to ****, put the lap top down and enjoy the ****ing sunrise.

That's ridiculous.

Just because (to take the sun metaphor further) I know when the sun is gonna rise, doesn't mean I can't watch the sunrise.

In fact, it probably means i'll be better at watching the sunrise since i'll know exactly when to look for it.

The "LOL VIDEONERDS DON'T WATCH THE GAEMS" argument is the worst of all.

 

[SnowblindNYR]

Yeah and then the eyeball guys come in and every argument they have is, "OMG HENRIK LUNDQVIST SOFTIE, BRAN BOYLE SUK! GIRARDI SO DUM!"

Very convincing.

So if you think that's ridiculous why don't the stat people think it's ridiculous to rely solely on stats?

BTW, I heard that if the Rangers don't score first they are incapable of winning.

 

[sousuffer]

Additional information doesn't make you more knowledgeable on a subject? I really don't care, just playing Doval's advocate.

Not if that information is too difficult for some people to understand...some people just don't want to put the effort in (or are incapable of it) to at least attempt to understand what they are criticizing - it's easier to just close your eyes and say that advanced stats suck because I can see everything with my eyes without having to spend time learning something new.

I personally enjoy both sides, but the dinosaur mentality of "ain't no doggone kid from Yale gonna tell me that a bunt ain't a good play" that was so prevalent in Moneyball and still comes out constantly just annoys the hell out of me. Of course, many of the stat guys at time come across as holier than thou (which annoys me as well), but the OP didn't, he started a rational discussion, and a dinosaur came out of the woodwork to denounce the OP because he/she "doesn't need advanced stats".

 

[BlueshirtBlitz]

I can understand advanced stats with baseball because it's just so much easier to separate the individual from the team. ABs are individual accomplishments. Having said that, stats like WAR are incredibly unreliable and always being changed. I can get on board with wOBA but it's basically just looking at players singles, doubles, triples, hrs, and walks, and putting it all together in a neat little package. It's not very "advanced".

Sports like football and hockey are just so much harder to quantify.

That's not how WAR works.

WAR is a measure of how good a player HAS been, not how good they will be. You can't look at someone's WAR and say they'll have a good year next year, but you sure as hell can look at someone's WAR and say how good they were that season.

It's why Jimmy Rollins winning the MVP when he did was such a sham.

Like I said, I really wish there was an equivalent of WAR for hockey, because WAR is a wonderful little stat.

 

[Ail]

So if you think that's ridiculous why don't the stat people think it's ridiculous to rely solely on stats?

BTW, I heard that if the Rangers don't score first they are incapable of winning.

They do think that is ridiculous. I have not seen one person say "advanced stats are the best and only way to properly evaluate players." If someone did say that and I missed it, they are wrong.

-31- said re: Klein 3 years of data is better than watching him play 3 games, and in that regard, yeah it's true. Which would you trust more, watching 3 games, or 3 years of information about games someone has played? Which would be more reliable in predicting their future play?

Edit: Also, as far as the sarcasm stat goes, no stat is infallible, they are a guide. There are anomalies in all data. They will never be 100% accurate, that doesn't make them wrong or irrelevant.

 

[iamitter]

Show me a shot quality heat map for each defenseman when they're on the ice and another for when they're off and we will talk. Someone is going to quantify that eventually, and work on it has already begun, but it's not there yet.

The problem I have with that stuff is that it's all negative quantifiers and doesn't actually account for the positive things a player does. Yes, being great positionally and making a lot of good defensive plays will be reflected on the other side of the ledger, but it's just a reflection and, like all reflections, is imperfect.

Advanced stats are great and are important, but a solid eye remains more important.

That's not really that hard. You can actually already do that now fairly easily. And you can bet your ass several teams have done it already.


Every game in the NHL gets a log. For example: http://live.nhl.com/GameData/20132014/2013020001/PlayByPlay.json

You can reverse engineer the data relatively simply (no, I'm not going to post my code - it's a good exercise for anybody actually interested in this). Each shot comes with an x and y coordinate. The problem is that the NHL tracker data is often off by a lot because the people who track shots often get them very wrong.

I'm planning on publishing a paper for my senior capstone project on this very topic next year. My adviser has been incredibly helpful in getting in touch with the right people - you'd be surprised how far this field has gone. Addition/Subtraction is how it started, but it goes way further than that. Example: http://scholar.harvard.edu/files/pettigrew/files/pettigrew_nhl_conf_gap.pdf
The above guy uses Monte Carlo methods of simulation.

This is never going to be an exact science. Statistics depends on large sample sizes and 82 games is not nearly enough to get a decent statistical set - let alone 7. It's usually "good enough" to work with though (this is why classes in college curve even if there's <50 people in the class). Just like the eye test, you can never fully predict whats going to happen in a small sample size. Nash may be a good player, but he may go 10 games without a goal. He may go 10 games with 15 goals.

There's always going to be naysayers who say, oh we had 40 shots but only 2 goals, clearly shot percentages don't generally hover around x%. By the same token, you can say the eye test says Nash is a terrible floater who turns over the puck and never scores goals (because that's how one game looked).

Hockey statistics are still at a very young stage. They won't tell you THAT much yet. But they do tell you a decent amount. They tell you who has the puck at most times. That's valuable. No coach would want to not have the puck as opposed to have it. With time, however, they'll develop.

 

[Hunter Gathers]

Once hockey advanced stats get close to baseball's, I'll begin to subscribe and take notice. I'm a huge fan of advanced stats in baseball. Not so much in hockey.

 

[Thirty One]

That's ridiculous.

Just because (to take the sun metaphor further) I know when the sun is gonna rise, doesn't mean I can't watch the sunrise.

In fact, it probably means i'll be better at watching the sunrise since i'll know exactly when to look for it.

The "LOL VIDEONERDS DON'T WATCH THE GAEMS" argument is the worst of all.

This.

Show me a shot quality heat map for each defenseman when they're on the ice and another for when they're off and we will talk. Someone is going to quantify that eventually, and work on it has already begun, but it's not there yet.

I'm not sure why I would need that to support my position that On Ice Save Percentage is random. But if you want that information, play around with this: http://somekindofninja.com/

 

[SnowblindNYR]

They do think that is ridiculous. I have not seen one person say "advanced stats are the best and only way to properly evaluate players." If someone did say that and I missed it, they are wrong.

-31- said re: Klein 3 years of data is better than watching him play 3 games, and in that regard, yeah it's true. Which would you trust more, watching 3 games, or 3 years of information about games someone has played? Which would be more reliable in predicting their future play?

Edit: Also, as far as the sarcasm stat goes, no stat is infallible, they are a guide. There are anomalies in all data. They will never be 100% accurate, that doesn't make them wrong or irrelevant.

Except, 2 problems.

1) Mint (I think) in the Klein thread had an aneurysm over Klein and every post of his was complaining about his stats. If he thinks that it's not the be all end all, why have a conniption over stats without watching him play.

2) I have yet to see a Predators fan say that he's not a good defenseman. These guys saw him play far more than 3 games. And usually fans of teams trading players talk **** about the player traded to justify the trade.

 

[stan the caddy]

That's not how WAR works.

WAR is a measure of how good a player HAS been, not how good they will be. You can't look at someone's WAR and say they'll have a good year next year, but you sure as hell can look at someone's WAR and say how good they were that season.

It's why Jimmy Rollins winning the MVP when he did was such a sham.

Like I said, I really wish there was an equivalent of WAR for hockey, because WAR is a wonderful little stat.

Did I say any of that?

WAR is unreliable due to a few reasons:

1. They're always tweaking the forumula. There's not even one concrete WAR. Different Sabres come up with different WARs.

2. It uses an unreliable fielding metric.

Sabres will tell you that their defensive metrics take three years to stabilize but it doesn't stop them from using it in a single season of WAR. They may have since changed their opinion on that, they do it a lot.

For example, Sabres concluded that Mark Teixeira was a bad fielder his first year for the Yankees. In reality, Teixeira played closer to second base which allowed Cano to also move over and Jeter benefitted greatly from it. Saying Mark Teixeira was a below average fielder was downright stupid. Ironically, the guy that Sabres love to hate (Jeter) finished the season with the best defensive grade of his career despite being older and slower than his younger years.

You can see how unreliable their defensive metrics are just by looking at the crazy fluctuations individual players have from year to year.

 

[Thirty One]

1) Mint (I think) in the Klein thread had an aneurysm over Klein and every post of his was complaining about his stats. If he thinks that it's not the be all end all, why have a conniption over stats without watching him play.

I think you're overstating his reaction a tiny bit.

2) I have yet to see a Predators fan say that he's not a good defenseman. These guys saw him play far more than 3 games. And usually fans of teams trading players talk **** about the player traded to justify the trade.

How many did you talk to? What was their scouting experience?

 

[SnowblindNYR]

I think you're overstating his reaction a tiny bit.


How many did you talk to? What was their scouting experience?

I don't know, is scouting experience the only way you can form a valid opinion of a player?

 

[Thirty One]

I don't know, is scouting experience the only way you can form a valid opinion of a player?

No.

Just illustrating how a poll of an indeterminate amount of people with an undisclosed aptitude may not be the best measure of a player.

 

[SnowblindNYR]

No.

Just illustrating how a poll of an indeterminate amount of people with an undisclosed aptitude may not be the best measure of a player.

I'm not saying it is. Neither is staring at a bunch of numbers (and I love numbers). Why not wait and see some games with our own eyes so that we can use those eyes in conjunction with his stats? No! That wouldn't be fun. How else would you rip Sather?

 

[Thirty One]

I'm not saying it is. Neither is staring at a bunch of numbers (and I love numbers). Why not wait and see some games with our own eyes so that we can use those eyes in conjunction with his stats? No! That wouldn't be fun. How else would you rip Sather?

OK, now you're just debating what should be discussed on this message board. And you seemed to conclude that neither the past nor the future should be discussed. How very mindful of you.