Congrats to Sam Ventura

[chrisp200]

"Hiring Sam Ventura as a consultant responsible for analytics only is the start, Penguins vice president of hockey operations Jason Karmanos insisted.

With hockey becoming more data-driven, building an analytics team is one of the Penguins' priorities, one over which Karmanos will have significant influence.

And one that started with the hiring of Ventura, a 27-year-old Swissvale native who, with Andrew Thomas, cofounded hockey analytics website war-on-ice.com."

http://triblive.com/sports/penguins/8786885-74/karmanos-ventura-penguins#axzz3gjWmIwZv

Click to read more...

 

[Cherpak]

Great. hockey is going "Geek" as well now.

 

[zero8771]

I appreciate that the org is at least doing different things

 

[Coastal Kev]

I would love for this guy to prove his value, let's see the data.

 

[Penguinator]

I appreciate that the org is at least doing different things

This analytical overdrive hiring could have been inspired by the Després for Lovejoy trade fiasco. Ventura knows his **** very well, his input will gradually become more valuable as this science evolves.

 

[WayneSid9987]

Good for the Pens to snatch Sam up before anyone else, seeing as he was right in thier backyard.

Hope they take Sam's recommendations for thier next hires for this department. He has alot of solid connections with very smart analytics people.

 

[Ugene Magic]

This analytical overdrive hiring could have been inspired by the Després for Lovejoy trade fiasco. Ventura knows his **** very well, his input will gradually become more valuable as this science evolves.

My thoughts are, it's fine to use those stats to check what you are seeing on video and through regular stats, but it just seems for the term being used, "geek statistic" they don't go far or deep enough to accurately without a doubt paint the whole picture. I would think if you went this far, go the extra mile and finish it.

That's using every combination of every situation in terms of hockey. They have to break it down in it's many forms, and it's possibilities they have with every single player to the fullest. That would take a lot of time, more than even the Analytic Scientists would probably want to go. That's quantifying each player by situation, line-mates individually and total team accounts don't even come into play at all skewing everything.

Like Sutter having a good year statically but playing with different line-mates of every variety.

I call it going in circles to come to the same conclusion.

Think of it like a rabbit. I can stand in the same spot, send off the dog/Analytic scientist to chase/flush out the rabbit/stats, and it will always circle back to its origin.

They basically brought in a dog to flush out the rabbit, but the rabbit was there already, shoot him then!

It's checking to check to make sure you check, but hey...lets check some more.

Analytics.

 

[Jacob]

If you need analytics to recognize that Despres for Lovejoy is a bad idea I don't know what to say.

 

[Penguinator]

If you need analytics to recognize that Despres for Lovejoy is a bad idea I don't know what to say.

That deal happened because of bad analytics being misused by org & ...experience!

 

[Jacob]

Did it? I thought it was Guerin's idea. I doubt he can even spell analytics.

 

[SEALBound]

"Mr Johnson, Mr Rutherford...please listen to me...every single stat known in the hockey world is screaming at you that Rob Scuderi is just not a good defenseman. Not only is he worse than 29 teams #8 dman, but he's pulling down Pouliot worse than the Titanic. He doesn't even pass the most BASIC eye test!"

"Ventura my boy...I think we'll go with our gut on this one. And our gut is telling us that Rob Scuderi is a winner. Thank you though for your concern."

 

[SEALBound]

Did it? I thought it was Guerin's idea. I doubt he can even spell analytics.

I think Guerin was one of the ones behind the Lovejoy trade...I highly doubt he's for analytics. Wide spread use of analytics would have killed his career 2 years early.

 

[Penguinator]

JR said his analytical guy(s?) were high on bringing Lovejoy back but Guerin might have pushed for the buddy thingy as well, thus saying in words what the other(s?) said in numbers...

Read: Lovejoy makes others better.

 

[Coastal Kev]

JR said his analytical guy(s?) were high on bringing Lovejoy back but Guerin might have pushed for the buddy thingy as well, thus saying in words what the other(s?) said in numbers...

Read: Lovejoy makes others better.

EXACTLY... Which is why I will never buy into analytics involvement in hockey. The day of trade deadline when someone posted the trade before it was official, I truly thought it was a failed attempt at humor. Any analytics that showed the trade was a good deal should be cast out immediately never to be heard from ever again.

 

[djt153]

EXACTLY... Which is why I will never buy into analytics involvement in hockey. The day of trade deadline when someone posted the trade before it was official, I truly thought it was a failed attempt at humor. Any analytics that showed the trade was a good deal should be cast out immediately never to be heard from ever again.

i think most people would recognize the limitations present when GMJR speaks on topics like that and discount his comments sharply as a result rather than citing them as immutable proof

 

[Penguinator]

EXACTLY... Which is why I will never buy into analytics involvement in hockey. The day of trade deadline when someone posted the trade before it was official, I truly thought it was a failed attempt at humor. Any analytics that showed the trade was a good deal should be cast out immediately never to be heard from ever again.

Well, apparently one of Karmanos (the Pens one) guys was shipped to Europe afterwards. Now enters Lord Ventura:

....and that's why Ventura should be so useful, by presenting the data in a way that the org can make sense of it so their bias is filtered through analytical context.

All about interpretation of submitted data, Ventura should circumvent BS takes on it by context & superior analytics. He'll... educate the org so no more Després **** happens.

 

[drpepper]

My thoughts are, it's fine to use those stats to check what you are seeing on video and through regular stats, but it just seems for the term being used, "geek statistic" they don't go far or deep enough to accurately without a doubt paint the whole picture. I would think if you went this far, go the extra mile and finish it.

That's using every combination of every situation in terms of hockey. They have to break it down in it's many forms, and it's possibilities they have with every single player to the fullest. That would take a lot of time, more than even the Analytic Scientists would probably want to go. That's quantifying each player by situation, line-mates individually and total team accounts don't even come into play at all skewing everything.

Like Sutter having a good year statically but playing with different line-mates of every variety.

I call it going in circles to come to the same conclusion.

Think of it like a rabbit. I can stand in the same spot, send off the dog/Analytic scientist to chase/flush out the rabbit/stats, and it will always circle back to its origin.

They basically brought in a dog to flush out the rabbit, but the rabbit was there already, shoot him then!

It's checking to check to make sure you check, but hey...lets check some more.

Analytics.

Your rabbit analogy makes no sense.

Video highlights are essentially a handful of snap-shots that assistants and coaches attempt to compile into some meaningful narrative.

Traditional statistics are descriptive numbers that some person some time decided are important.

Analytics is a process that uses any type of data and enables the team to learn what is repeatable, predictive, descriptive, and whether the underlying systems and roster choices are working. Instead of a handful of snapshots, it's thousands. And instead of randomness or personal whim, usefulness is determined by numerical values.

The data used is limited only by current technology. There is a Canadian start-up that is using video processing to track player's touches on the puck and movement off-puck. There are two other corporations that work with special video equipment and sensors. There is data available from NHL tracking. There is data that could be available from players wearing sensors.

None of this data and the analytics to separate the important bits from the noise will fully describe and dissect a hockey game, but it doesn't have to. It will provide invaluable tools to scout other teams, isolate important aspects that need further dissection (more video), scout draft picks and potential trade or FA options, review internally prospects and players, evaluate player usage, systems, and coaching decisions, scrutinize in-person and video scouting, and eliminate personal biases.

Nothing else comes even close to the power of analytics.

 

[DoktorZaius]

EXACTLY... Which is why I will never buy into analytics involvement in hockey. The day of trade deadline when someone posted the trade before it was official, I truly thought it was a failed attempt at humor. Any analytics that showed the trade was a good deal should be cast out immediately never to be heard from ever again.

But that's terrible reasoning. The analytics don't show that Lovejoy is better than Despres. Quite the opposite. All that fiasco demonstrated is that our front office still has AT LEAST one more poison that needs excising.

 

[TNT87]

If you need analytics to recognize that Despres for Lovejoy is a bad idea I don't know what to say.

It was just anal.

 

[madinsomniac]

Adding analytics in the correct way is a very positive move way fro the stone ages hockey traditionally resides in. Advanced stats never tell the whole story, but they do provide clues to make life easier. Traditional scouting is great, but on its own there are too many games and too many players to really get a good feel for everything one needs to know, and far too much personal bias that can get interjected.

Using advanced stats can narrow focus onto the gaps and issues a team really has and let them target a smaller portion of players or strategies that may be the solution.

 

[HandshakeLine]

Awesome!!!1!! !12!1 !.123! I can't wait until we've reduced the whole game to spreadsheets and data mining! The future rules!















This posting paid for by your facebook likes. Please click "like" for more such quality posting!

 

[chrisp200]

How anyone could think this is a negative thing is beyond me.

 

[SomeDude]

While I'm sure they will get some fine tuning, analytics say that Tyler Kennedy and Dan Winnik are good players, so I have a hard time taking too much stock in them at this point.

 

[cassius]

I have a career in a numbers-focused discipline, and I gotta say.. I'm very skeptical about the whole hockey analytics movement. Seems like a lot of snake oil if you ask me.

I think the collection of the data is great, but trying to make inferences on the said data is where it gets sketchy.

 

[Asuna]

JR said his analytical guy(s?) were high on bringing Lovejoy back but Guerin might have pushed for the buddy thingy as well, thus saying in words what the other(s?) said in numbers...

Read: Lovejoy makes others better.

You keep saying that...do you have any proof of that? Cause I don't remember seeing JR say anything like that.

Also Karmanos wasn't shipped to Europe...Guerin was. So again...dunno why you keep saying Karmanos was.