Know your Cat Rarity (PIM/SOG/HIT/BLK Multi-Cat Advice)

[Pengwin7]

PIM, SOG, HIT, BLK.

Two exercises to see how well you understand their value in multi-cat leagues.

Exercise #1, Order these Stats from: Least Plentiful to Most Plentiful (as they occur for the entire NHL in a year... or "per game"... or generically, in most fantasy leagues).
1 (LEAST):
2 (2nd LEAST):
3 (2nd MOST):
4 (MOST):

Exercise #2, You have the choice between 4 defensemen in your ROTO league counting PIM/SOG/HIT/BLK.
Without getting into where your team is in the standings of each cateogry, generally speaking, WHO is the best addition?

1. Goon McGoonly: 20 PIM, 5 SOG, 5 HIT, 5 BLK.
2. Shooter Shotz: 5 PIM, 20 SOG, 5 HIT, 5 BLK.
3. Hit Hitterson: 5 PIM, 5 SOG, 20 HIT, 5 BLK.
4. Bob Blockchuk: 5 PIM, 5 SOG, 5 HIT, 20 BLK.


Answers tomorrow.

[gainerama]

1. SOG
2. PIM
3. BLK
4. HIT

I'll take McGoonly or Blockchuk, probably more Blockchuk personally.

[Canucks_fan18]

1 (LEAST):BLK
2 (2nd LEAST):PIM
3 (2nd MOST):HITS
4 (MOST):SOG

1. Goon McGoonly: 20 PIM, 5 SOG, 5 HIT, 5 BLK.
2. Shooter Shotz: 5 PIM, 20 SOG, 5 HIT, 5 BLK.
3. Hit Hitterson: 5 PIM, 5 SOG, 20 HIT, 5 BLK.
4. Bob Blockchuk: 5 PIM, 5 SOG, 5 HIT, 20 BLK.

I am taking the shooter although shots are plentiful there typically not from a D and when a D shoots a ton it usually means he's getting points/PP time.

[skiteam14]

1 (LEAST):BLK
2 (2nd LEAST):PIM
3 (2nd MOST):HITS
4 (MOST):SOG

1. Goon McGoonly: 20 PIM, 5 SOG, 5 HIT, 5 BLK.
2. Shooter Shotz: 5 PIM, 20 SOG, 5 HIT, 5 BLK.
3. Hit Hitterson: 5 PIM, 5 SOG, 20 HIT, 5 BLK.
4. Bob Blockchuk: 5 PIM, 5 SOG, 5 HIT, 20 BLK.

I am taking the shooter although shots are plentiful there typically not from a D and when a D shoots a ton it usually means he's getting points/PP time.

+1 for all this. Fantasy hockey school is in session!

[tomastop]

1 (LEAST): PIM
2 (2nd LEAST): BLK
3 (2nd MOST): HIT
4 (MOST): SOG

1. Goon McGoonly: 20 PIM, 5 SOG, 5 HIT, 5 BLK.
2. Shooter Shotz: 5 PIM, 20 SOG, 5 HIT, 5 BLK.
3. Hit Hitterson: 5 PIM, 5 SOG, 20 HIT, 5 BLK.
4. Bob Blockchuk: 5 PIM, 5 SOG, 5 HIT, 20 BLK.

(I'd probably still be inclined to add Shooter though, for the reasons mentioned above... bigger chance for more points)

[Shoeless]

1 (LEAST): PIM
2 (2nd LEAST): Blocks

3 (2nd MOST): HITS




4 (MOST): SOG

1. Goon McGoonly: 20 PIM, 5 SOG, 5 HIT, 5 BLK.
2. Shooter Shotz: 5 PIM, 20 SOG, 5 HIT, 5 BLK.
3. Hit Hitterson: 5 PIM, 5 SOG, 20 HIT, 5 BLK.
4. Bob Blockchuk: 5 PIM, 5 SOG, 5 HIT, 20 BLK. I can find shots, pims and hits at forward position as well, not so for blocks.

[RizzeeDizzee]

PIM and Hits can be found from a wide variety of defensemen, and from your forwards as well. The delta in defenseman shots can be wide, but the drop off from a top guy to the middle of the pack guys is precipitous, plus it's a category that will be carried by your forwards. That leaves blocked shots, where you have to keep in mind that the top forward last season had a mere 89, which is a total that was bested by 117 different defensemen (plus even more d-men if you go by a per game basis).

Of those four, give me Bob Blockchuk

I'll note that one factor which wasn't woven into this is the propensity of defensemen who block a lot of shots not to be among the better point producers, as only five rearguards among the top 30 in 2013-14 blocked shots tallied 40+ points. But that's a debate for another day.

[Pengwin7]

Answers from quiz.

1. In the NHL last year (2013-2014), skaters totaled this quantity in each statistic:
PIM: 26, 374 (least)
BLK: 34, 794 (next least)
HIT: 57, 927
SOG: 74, 423 (most)

So, shifting these to be based off of 1 PIM:
1.00 PIM
1.32 BLK (32% more BLK than PIM)
2.20 HIT (120% more HIT than PIM)
2.82 SOG (182% more SOG than PIM)


2. Which defenseman should you roster if your league counts all of these as single "categories"?
Generally speaking, 20 PIM is more rare than 20 BLK... which makes Goon McGoonly the most valuable addition.
If you ran his numbers in a software (like FHG), he would shake it to be the most valuable since he "stocks" that category (%-wise) more than the other options.


Pictured above: Goon McGoonly... and his cat

It is also debatable that Bob Blockchuk could be the most valuable.
Defensemen block (on average) about 3x more than forwards.
If you are not getting blocks from a slotted defenseman, you will have a very hard time getting them back from a forward (especially since most high-BLK forwards, not named Ryan Getzlaf, are low-offensive producers... like Boyd Gordon).
PIM can also be made up through rostering forwards... so depending on how well a GM can work "strategy", a good case can be made that Bob Blockchuk is more critical as a defenseman.


3. So... what's the point of this:
Well, a lot of times we go to the waiver wire and you'll see a 30 HIT / 10 BLK defenseman as well as a 10 HIT / 30 BLK defenseman. 40 vs. 40. It's important to always know who is more valuable.
If HIT+BLK (a commonly added thing, by some) are equal for two guys... the more valuable guy has higher BLK.
And... actually... a 10 HIT / 20 BLK defenseman probably holds more value than a 30 HIT / 10 BLK defenseman... because "HITS" are simply 2x more plentiful (naturally) than BLK.


Pictured above: Bob Blockchuk... and his cat

In most leagues (mine at least), defenseman on the wire have low offense. The PP D-men are gone.
So you are left looking at the high-quantity categories: PIM / SOG / HIT / BLK.
Here's what I do...
1. I total up SOG + HIT.
2. I total up PIM + BLK (and multiply it by 2)
The guy with the highest stats (per game basis... always "per game") is the best add.

My 2 cents.
Goon McGoonly & Bob Blockchuk.
These are your guys.
McGoonly if you don't have a forward roster spot to make up PIM.
Blockchuk if you can roster some PIM in a forward spot.

[Magicstew]

1 (LEAST):BLK
2 (2nd LEAST):PIM
3 (2nd MOST):HITS
4 (MOST):SOG

Agree with this order and I am going with Bob Blockchuk - the least amount category of BLK. forward will find shots. shot blockers will have chance at SHP because they will be out on penalty kill.
One good shotblocker can win that category in a week, so you can focus your Forwards on the scoring categories.

Good stuff as always Pengy - makes you think out of the box sometimes.

[mister_mcgoo]

The problem with this is that this mostly applies to old school roto leagues, leagues where you don't care where you get the numbers from and you just want to get the most of them because they're all equal in the end (in other-words it doesn't matter that goals are relatively rare compared to shots, winning the goals cat 350-349 is the same as winning the SOG cat 3500-3389, the rarity isn't factored into the overall scoring).

Most leagues are becoming more complex and realistic, not less so. All the leagues I play in weight the categories to get rid of this 1-1 bias, and that is becoming more and more the norm. You're also seeing increased positional weighting to bring balance to the format. For example: give 2-way forwards more value by awarding forwards more for BLKs than d-men...or make a offensive d-man carry more weight for scoring than a forward (i.e. forward pt =1, d-man pt = 1.25). IMO it's a much better way to do it because you're more closely mirroring the actual NHL where a winning team takes a mix of grit/toughness, finesses, scoring, defense and goaltending to win.

[Pengwin7]

Weighting is a neat concept.
But... "weighting" is usually created by a commissioner and most weighted leagues have different weightings and thus can not be discussed the same way.
In other words, this thread really only applies to non-weighted standard category roto (or H2H) pools.

There can be no generic thread that captures all differently weighted fantasy leagues.
The only solution to evaluating players in a weighted league is to go to FHG and input settings.

So... yes, for weighted leagues... the advice that I typed out here is nothing but easily criticized.

[mister_mcgoo]

Weighting is a neat concept.
But... "weighting" is usually created by a commissioner and most weighted leagues have different weightings and thus can not be discussed the same way.
In other words, this thread really only applies to non-weighted standard category roto (or H2H) pools.

There can be no generic thread that captures all differently weighted fantasy leagues.
The only solution to evaluating players in a weighted league is to go to FHG and input settings.

So... yes, for weighted leagues... the advice that I typed out here is nothing but easily criticized.

Well, yes and no. Your core points still have value in a weighted system as well, you just have to do more math is all.
A simple analyis would be to compare categories for rarity and contrast that with their relative values (according to the league weighting) and see what anomalies you can idenitfy.

So for example, lets use Goals and SOG again for simplicty. Last season there were 6520 Goals scored on 73321 SOG, that means that a SOG is:

73321 / 6520 = 11.25 times more common than a G

Now if your league happens to score G=10 and SOG=1 then you would be wise to target players who shoot a lot whether they score or not since SOG are slightly overvalued. Mind you this is a simplistic interpretation because SOG and Gs tend to correlate (i.e. players who shoot more do so because they're good at scoring goals so if you look at the top SOG leaders each year it will also feature most of the league leaders in Gs as well). BUT where this info can be very useful is with targeting undervalued players. For example a smart tactic would be to identify a player who regularly puts a lot of pucks on net but is coming off a season with a low shooting %. If you look at Vrbata last season he's a good example...consistently around 250 SOG every season but only shot 7.6% last year.

[Canucks_fan18]

Pengwin7 excellent work as always but the numbers are so objective. Its just data.. and there's no analysis of data being completed. What do the numbers mean? Yes there rarer and might be better in a spot start but if you dig deeper a high SOG defencemen vs Block head play two vastly different roles on their team. And that factor needs to be considered because the block guy with low sog will by and large be just a block guy perhaps he touches other cats like HIT/PIMS but point production is low. A High SOG dmen usally translates into good point production the only exception I seen at the top of last year were Garrison 30+ points and Muzzin 25 points( a little snake bitten perhaps)

[Pengwin7]

Pengwin7 excellent work as always but the numbers are so objective. Its just data.. and there's no analysis of data being completed. What do the numbers mean? Yes there rarer and might be better in a spot start but if you dig deeper a high SOG defencemen vs Block head play two vastly different roles on their team. And that factor needs to be considered because the block guy with low sog will by and large be just a block guy perhaps he touches other cats like HIT/PIMS but point production is low. A High SOG dmen usally translates into good point production the only exception I seen at the top of last year were Garrison 30+ points and Muzzin 25 points( a little snake bitten perhaps)

Yes, but I gave you the data for your choices.
We aren't talking about Garrison or Muzzin here.
I understand your thinking... because normally guys that are HIGH-SOG play a lot... and by default... their other stats will be solid.

But, in the example I gave we are talking about four guys with some combination of 5-5-5-20 in four categories.
Truth is... any "High SOG" defenseman isn't really on the waiver wire.
There is simply a "highest SOG" defenseman of the group.

Waiver wires are going to have these kinds of guys (with 2013-2014 stats):

a) Ryan Ellis. PIM=24. SOG=123. HIT=57. BLK=97.
or
b) Carl Gunnarson. PIM=34. SOG=48. HIT=166. BLK=176.

Most people in category leagues are probably going to take Ryan Ellis... because the SOG look so good.
Most GMs are very offense-centric in their thinking in multi-cat.

But I don't think Ellis is the right choice.
SOG is the most plentiful of all these stats.
And while Ellis' extra 75 SOG seems like the massive boost... if SOG is plentiful... it's really not THAT incredible an advantage over Gunner's other cat advantages.

By my math:
Ellis. SOG+HIT = 180. 2x(PIM+BLK) = 242. Value = 422.
Gunner. SOG+HIT = 214. 2x(PIM+BLK) = 420. Value = 634.

[Canucks_fan18]

But, in the example I gave we are talking about four guys with some combination of 5-5-5-20 in four categories.
Truth is... any "High SOG" defenseman isn't really on the waiver wire.
There is simply a "highest SOG" defenseman of the group.

Waiver wires are going to have these kinds of guys (with 2013-2014 stats):

a) Ryan Ellis. PIM=24. SOG=123. HIT=57. BLK=97.
or
b) Carl Gunnarson. PIM=34. SOG=48. HIT=166. BLK=176.

Most people in category leagues are probably going to take Ryan Ellis... because the SOG look so good.
Most GMs are very offense-centric in their thinking in multi-cat.

But I don't think Ellis is the right choice.
SOG is the most plentiful of all these stats.
And while Ellis' extra 75 SOG seems like the massive boost... if SOG is plentiful... it's really not THAT incredible an advantage over Gunner's other cat advantages.

Yup and this makes sense I would take the Gunner in this situation. Last year after Stammer went down in one of my leagues I employed an all grinder squad in H2H to hopefully slip into the playoffs and I was dam close. I routinely had Brendan Dillion and Gudas on my squad.