I guess I am still struggling finding the balance between compiler and short peak. I mean, many old-time players dont have much outside of 7 years, but this board seems cool with that. Why is Marchand's best 7 not treated the same? His ESVsX wont get any worse after these 7 seasons, it can only go up. If this year holds, he'll be at a 69.4! That is a damn fine score, especially considering his usage during his first couple years.
Nothing here is an exact science. Here’s one way to put it, I suppose – Imagine the 7-year exercise to be a way of judging what a player was like at their best. But over the course of an ATD season, a player is not always at their best. You could consider pre-and post-prime play (or longevity) as a way of approximating how much the dropoff is going to be from their best games to their off-games and how consistently they’ll play game-to-game.
We are drafting players for their whole careers, and I think that the larger sample we have on a player, the more sure we are about what we are getting. Two players with an identical “best 7 years†are only equal in that regard, but it says nothing about their whole careers. If one player only played 7 seasons in total and another had 7 more seasons as a useful player, I’m much more confident in his ability to deliver a consistent performance.
You mentioned Sandford later on, and I don’t recall calling him a “top notch 4th linerâ€. Someone else may have, I don’t remember. I have my own concerns about his career length too. But one thing I’ll say is that career length standards are different from era to era. 7 years is often used as a judge of a player’s prime (as it’s the length of time we can expect most good players from most eras to have achieved), but it’s also not the magic mark where a player has finally played enough for consideration. Sometimes it’s more, sometimes less.
And no, I dont have the spreadsheet. I tried opening it earlier, but for some reason my computer doesn't like it, and it wont open.
I think I had that issue at first, but I changed the extension and it opened. If you have excel, it should work. Want me to resend it?
Hey rmartin, I'm not a voter but I like the Marchand pick. I think he's made his mark on history already with his big game performances and his unique style and skill set. Most hockey fans may hate him...but they'll remember him.
That’s a little too subjective. Just about any player selected at this time is still “remembered†by a segment of hockey fans. You could say as of the summer of 2011 that Marchand will be “remembered†and you’d be right, but at that point he’d clearly be a bad ATD pick. That aspect shouldn’t be considered a reason to draft a player on its own. There’s still got to be a point where a player’s significance for their skill set is properly backed up by a sample of on-ice play to make this draft.
I’m more of a fan of GMs timidly bumping players up 100 picks at a time and seeing how well-received they are as they go, and how they compare to the players selected around them. I looked forward to that opportunity over the next 4 years with Marchand in the lower drafts. He shouldn’t have skipped the MLD outright.
For those knocking his PK time, try looking at his TOI as well as the numbers in my spreadsheet. I have a feeling that he and Bergeron are underrated by the metric based on GA, possibly because they've been so freaking good on the PK. Without checking the numbers I'm pretty sure they were on the ice for only 4 PPGA in the shortened 2013 season while playing 2 minutes/game on the PK.
You are right; playing with Bergeron (and being quite good himself) could contribute to a “SH%†being lower than his actual PK time. In his first 6 seasons, here are the Bruin forwards with at least 200 SH minutes and how much they had per game:
Other 2.74
Other 2.01
Bergeron 1.95
Other 1.59
Marchand 1.54
Other 1.05
Other 0.85
Using the situational play data from nhl.com, I was able to calculate that the Bruins have been shorthanded approximately 5.35 minutes per game during his career. So his career usage before this season sits at 29%, a tad higher than your sheet has him at. (that takes his 24 missed games into account)
I like your sheet as a quick shorthand, but for post-2008 players I recommend anyone who thinks they have a superb penalty killer whose numbers might break the SH% formula run the calculations themselves to see if maybe they deserve to be described with a more appropriate number. I am happy to see that even a player whose team results have been especially good, and who has played with Bergeron, is not that far off from what your sheet calculates.
Some of us probably wouldnt take him as high as he did..but hes also been in the league for almost double the time...has a better peak and is having a historically impressive season (which naturally will make others rate him higher than he probably should be).
Pretty much.
What he is doing this year should be judged as what it is right now, not what it could potentially look like.
If what you’re saying is, his season isn’t done yet, I agree, but the fact is even if he got injured right now, the events of this season help his case in a HUGE way.
I like Subban now more than Burns when he got picked, I'll tell you that.
I'd really be interested to hear what seventies has for VsXD for both Burns and Subban. I wouldn't be surprised if it's close.
Subban = 571 including a 56 for this incomplete season.
Burns = 614 including a 118 for this incomplete season. That excludes 2014.
Is VsXD calculated against the forward benchmarks or are new benchmarks done for defensemen?
In any case, I'm still not totally convinced VsX is a good system to compare the offense of defensemen.The way they accumulate points is a bit different than forwards.I have no better alternative though.
Yes, pretty much. Here’s how I’d put it:
- If a player has a good VsXD score, then they were among the higher producing defensemen of their time. Does that in and of itself mean they can be in an offensive role here or on the PP? No, but it is the first prerequisite.
- If a player does not have a good VsXD score, they were not used offensively in real life or they were not strong offensively at all, and you should not try to use them that way at a higher level.
Everyone is on a 4 Hour Clock for the Rest of the Draft
Why is this necessary?
Just curious, is this something we all actually want, or is this just “the way it’s always been†and has never been revisited?
Ehh...I remember looking him over in a past draft, and if I recall correctly, Oliver's post-expansion track record as a penalty-killer is largely fool's gold. I'm also not sure he's really a wing; think he's a LW/C sort of like Don McKenney is, and is in many ways a lesser version of McKenney. If memory serves, he didn't top 30% on any PK until he ended up on an expansion team, and while he killed a bunch of penalties for that team (forget which one it was), the team's PK was terrible when Oliver was one of their top guys and actually better in the season or two when he didn't have a big role.
He clearly could kill penalties, but I don't think he's good at the ATD level, even on a second unit. More of a third unit PKer, imo. Still one of the most talented all-around forwards available when he was picked, but not really much of a PK ringer.
I don’t know man, you just finished saying I’m being too hard on guys with my expectations for PK numbers and then you give Oliver the seventies treatment.
It’s absolutely true that he was only a 21% penalty killer for 4 of his 8 post-expansion seasons (for teams 9% better than average), and then he was used for an ungodly 62% (for teams 26% worse than average).
But, there are two things we must remember about this. The first is that many players that we draft and then tout their career average PK numbers have breakdowns like this. It should be clear when you draft a guy who’s at 30% on his career, it probably means at his peak he was 45%, and other times was barely used at all. And if he was on teams that were 10% better than average, it means they were sometimes below average, sometimes average, sometimes excellent. Since the quality number is weighted by usage, you’re still ending up with two averages that do a good job of approximating what kind of penalty killer they were.
The second for Oliver, is that these are just his 597 post-expansion games. He had 530 more before that, that aren’t captured in these numbers. During this time he had only 2 SHP, so he was not likely a frequent penalty killer, but also SHP show up much more infrequently in that era (and I’ve found a few that aren’t captured in the EV_PP_SH spreadsheet from the HSP, too). Just knowing what we know about him immediately following expansion, I’d call him perhaps a 15% penalty killer those years, which would make him about a 30% on his career. Boston had a below average PK in the years leading up to expansion, but also a below average PK among original six teams would likely still be a tad above average in the years just following it.
If I was to throw out two numbers that best represent Oliver’s usage and success, keeping in mind that it’s difficult because his career straddled two statistical eras, I’d say 30% for teams 6% worse than average. I’d call him a passable 2nd unit player.
Yes and no. There were not shutdown lines as we know them today, but there absolutely were players who shadowed a star at the expense of offense as they would do now. In those instances, their linemates took on the brunt of the scoring.
Skinner is that rare breed who appears to have been both. His playing weight was often mentioned in contemporary accounts, but for all the flak it earned him, that fact didn't seem to impede his ability to skate or check at a high level. On top of that, he was a consistent scorer.
On the subject of players who should go 300 spots higher, I can see Skinner being a third liner here. One of the things I was surprised to learn about him was how adept he was at scooping up rebounds and being what we'd consider a "net presence". He was noted to be the best of his time at it.
If Skinner is as good defensively as you say, absolutely he can be a 3rd liner, because certain special players can play that high in the lineup here on the basis of defense alone.
As far as offense is concerned, Skinner is pretty unimpressive. He’s a good example of a player from the pre-merger era that looks good based on raw rankings but then you start to look at what percent of the leaders he scored, and it’s nothing special. I’ve got him at 71, 59, 48, 42, 38, 33, 30. That puts him about 10% behind a guy like Rejean Houle, a fairly well documented checker who nonetheless will be playing on a 4th line or as a spare.
The Invincibles select
Red Berenson,
When I took Rolston, I was seriously considering Berenson too. I needed a guy who’d play LW at ES and C on the PK2 unit. I ultimately went with Rolston because of his PK success (as opposed to Berenson’s higher usage)… but there’s lots to like about Red.
One thing that’s not great is his ES offense. It’s ok, but it’s not what you’d think from looking at his raw numbers. He did get more PP time on his teams than he probably would have in a more normal situation.
To replace Giroux at RW, I'll pick RW Harry Oliver. He is good defensively, has a vsx of around 75 according to seventies' pre-1927 vsx method (which would put him in the top-100 seven year vsx scores), and he was one of the premier playoff performers of his era.
His speed, good defence and his ability to work in "combination" plays makes him an ideal RW to play with Colville and Nash.
Offensively, I think it’s a stellar pick. He’s right there with Kessel as the best RW “by the numbers†at the time you picked him.
…but he’s good defensively? That’s news to me and I had him 2 years ago in the MLD. Looking forward to this.
OK I'll take
Cully Wilson, RW.
If he were around today he'd probably look something like this:
Why were you upset to lose O’Reilly when Wilson was on the board? I thought you considered him just an O’Reilly with better offense.
I haven't had time to look further into Sturm's comments about Murray Oliver's PKing, so this guy might end up as our starting 4th line center or might be a spare. A guy who has a good PKing record for good PK teams, and somewhat surprisingly about the same ES scoring as Ryan Kesler (although this is more of a knock against Kesler than a plus for Sutter).
The Chicago Shamrocks select Brent Sutter, C
He was already back on my radar in the early 400s. I liked that he was a little of everything – decent scoring, ES defense, PK defense, and grit. Then I saw Kesler and thought, what makes him any better than Kesler? In the end I couldn’t find any reason he was better. They were practically equal in all areas, with only Kesler’s far stronger selke recognition being far the tiebreaker.
Selecting Alex Pietrangelo, D.
Am I crazy, or has he been a better all-around player day in day out, than Burns, Subban and Letang? Whole careers considered.
I considered him when I took Patrick. I hate him as a player, but nobody can deny the offensive effectiveness.
Agree. Super strong numbers and has been good every time he’s been in the playoffs too.
seventies is skipped.
rmartin selected Ray Getliffe, LW/C.
This was another guy I was strongly considering when I took Rolston. If I could have found PK quotes about him, I might have taken him.
