Wednesday, August 09, 2006

Death, Taxes and... Staios fading?

The Oilers' blueline is a source of considerable concern for your average Oiler fan these days. Whether you're Chicken Little or Dennis, you have to be at least a little bit worried. Even Vic, our resident voice of reason, is probably more than a bit squeamish about the Oilers' pairings conceivably containing two of Matt Greene, Ladislav Smid or Tom Gilbert. On top of that, the rest of the Oilers' blueline has been experiencing a little more character assassination than usual. All of a sudden, Jason Smith is just a decent 2nd pairing defenseman and Staios can't possibly be expected to contribute any more than third pairing minutes. The main reason you'll hear around the intarweb for Staios' fall from grace is that he runs out of gas as the season progresses and the minutes pile up.

Ever since I started frequenting or the Oilers' board on Hockey's Future I can recall many an Oilers fan deriding Steve Staios "for falling apart when he gets too many minutes" or "fading down the stretch." He was heavily criticized for this in 2003-2004 and he was definitely accused of it this year too even while playing in Chris Pronger's considerable shadow. It's basically a given among Oiler fans: You can't play Steve Staios too much or his play deteriorates. The idea is posed as though it is almost as unavoidable as death and taxes for the hopeful Oiler fan.

I think that's bunk and here's why.

I think the best way to measure the performance of a defenseman is by EV+/-. The vast majority of them shouldn't even play on an NHL PP and the proper evaluation of individual skater contributions to penalty killing are not clear. (Although I'm inclined to suggest that defensemen may have their most profound impact in that situation.) That leaves us with even strength. Defenseman, particularly the more defensively inclined, generally play almost 70% of their icetime at ES. The good news is that we have EV+/- (GF and GA at 5V5 only with EN goals removed) as a fairly well defined and accepted method of evaluating players at even strength. It's obviously limited by opposition and linemates (especially for dmen, as we've seen that the forwards in front of them tend to drive the results) but those things tend to be reasonably constant throughout the year. ie. It's pretty rare that a dman like Staios sees his matchups or his partner change dramatically for significant periods of time in a given season.

Using this logic, I chose to approach the question of Staios' diminishing returns during a given season by breaking up his last three seasons into quarters: the first 20 games, the next 21 games up to the halfway mark, games 41-61 as the third quarter and finally the last 21 games of the regular season. Pretty simple and arbitrary cutoffs in my opinion.

Staios played 76 games in 2002-2003. His ES/PP/PK minutes breakdown was 1301/160/231 for 22:16 ATOI. As you can see from the EV+/- breakdown for this season, both Staios and the team were steady but unspectacular in the first half before Staios begins to slip up in the third quarter. It follows the script up to this point, but then Staios and the Oilers really shape up at 5V5 for the stretch drive. This data sure doesn't jive with the contention that feeble little Stevie Staios can't hack it as the minutes pile up.

In 2003-2004, Staios played a full 82 game season including a pretty significant minutes total with a ES/PP/PK breakdown of 1470/224/194 and 23:03 ATOI. Again the EV+/- breakdown contrasts the commonly accepted idea about Staios' ability to play late in the season. In fact, he actually thrived as the season progressed in 03/04.

This past season Staios again played 82 games but saw his minutes decline a bit as he was wisely removed from the 1st unit PP rotation. His ES/PP/PK minutes breakdown was 1294/149/269 for 20:53 ATOI. The shocking thing here is how awful the Oilers were at 5V5 this year. Sure you could see it on the ice and it's been discussed quantitatively around these parts and on MC79's site to a certain degree, but it's all too real when shown in black and white in this table. The Oilers had a -42 goal swing from 2003-2004 to 2005-2006 at 5V5 alone. Incredible. If it wasn't for the rather exceptional special teams, this team wouldn't have had a chance at a playoff spot despite the much-ballyhooed additions of Peca and Pronger, who were expected to really make an impact at ES. This has to be a major area of concern for Lowe and MacT as they absolutely must get their heads above water at ES for both short term and long term success... Anyway, back to Staios.

At first glance, this appears to be the first year that matches the Staios myth. It does until you see how abysmal the team was at 5V5 in the second half of the year - the entire season deficit of -15 was due to 2nd half performance. The wheels fell off the whole team at ES and Staios actually kept his head above water posting a -2 in each of the last two quarters despite the massive -15 total for the team. Again I don't feel this evidence supports the idea that Steve Staios can't be leaned upon for this team.

This argument seems similar to an earlier thread I posted about Pronger's Postseason myth. I think many fans around the league have developed an almost equally deluded idea about what Staios brings to the table late in the season.

It comes down to this: He's just fine at the end of the year, he's performed in the past, he's not that old, and he can do it again. Staios was referring to leadership back on July 7th, 2006 when the world was crashing down on Oiler fans and he took a moment out of his vacation to circle the wagons, but I think he himself said it best:

"So I don't think that will be a problem moving forward. I'm not concerned."

Ditto, Steve.


Anonymous namflashback said...

Thanks for the analysis Riv. I had generally remembered Staios to be pretty damn dependable in prior seasons so couldn't understand the reason he was being burned at the stake on the forums.

That said, it was hard to watch the Oilers down the stretch run -- and it is amazing how bad that 5v5 is. I'm glad that they looked nothing like that during the playoffs.

Without doing any analysis, my recollection is that they probably bested every opponent at 5v5. Throw in a great PK, and a passable PP in all but the SCF and the result was what we saw.

What the hell was going on during that last stretch to the season?

8/10/2006 3:43 am  
Anonymous lowetide said...

I think most Oilers fans would be thrilled with another Steve Staios on the blueline. In fact, that's probably about what they need and may in fact have in someone like Hejda.

Problem is that they have too many question marks going into camp. IF Matt Greene can play elevated minutes with fewer penalties and IF Ladislav Smid can quickly establish himself in the NHL and IF Hejda makes a smooth transition the Oilers will be fine.

The odds that one of these things happen are quite good, but all three? The Oilers have a lot of question marks in one vital area, that's the concern.

Staios as a member of the top 4D for Edmonton shouldn't be a concern for anyone. In fact, I don't recall reading anyone saying it.

Having said that, your numbers are solid and it does let all the air out of the tires on something that had become prevailing logic (Staios fading).

Good stuff RQ.

8/10/2006 6:15 am  
Blogger Showerhead said...

Staios, to me, is a dependable guy who I'd want as the 4th best D-man on a championship team.

I was going to write a full response but now all I can think of is game 7. Fuck.

8/10/2006 8:18 am  
Blogger Black Dog Hates Skunks said...

Good post RQ. Interesting stuff.

LT - right on with your analysis - one of those will probably happen, maybe two, but all three - not likely

More and more I think they're going to camp to see what they have and if they think they are wanting then they will make their move.

8/10/2006 10:39 am  
Anonymous SmythSezNoQuestion said...

Rivers, it's like you can read my mind. I wanted to put together data on this same thing for HF till my PC tanked, and I can't get the game sheets happening in Excel on Mac.
Anyway, another thing I wanted to do (but am still figuring out the algorithm) is to tabulate, per game, how many minutes SS played before each +/- event.
Hypothetically, if he does wear down with more minutes, you'd expect the -s to come after 15-20 miutes of play, and the +s to be early. Personally, I'm betting that it'll be a relatively static distribution.

8/10/2006 11:25 am  
Blogger Dennis said...

I don't think he wears down as the season progresses...I think he's a better guy when he's playing 20 or less min then say 25. I dunno...maybe that's damn hard to prove/disprove but it's just something I've been clinging to for awhile;)

8/10/2006 4:43 pm  
Anonymous pdo said...

You cling to a lot Dennis ;).

But, I agree. I've never thought that Staios broke down as the season went down.. I did think he did when he started to play too many minutes though. Just seems as though if there's a stretch where Staios plays over 23:00 a night... he starts making blunders. I don't know if it's because of the quality of opposition (assuming he's usually on against the 2nd line opposed to the 1st) or what, just something that I've noticed over the years.

20:00 for Staios is perfect. Looks like he'll get 25:00 this season ;)

8/10/2006 5:01 pm  
Blogger Asiaoil said...

Yeah I never really bought into the gospel on Staios - the old "he wears down thing when given too many minutes" thing. Staios is just a good 2nd pair dman - simple as that and he will make a perfectly fine 2nd pair with Tjarnqvist. So Greene, Hejda, MAB and Smid can fight it out for the 5-7 slots.

On a related topic - everyone mentions the Pronger effect when an ordinary player suddenly becomes much better when he was paired with FCP - well how about the very very well documented Smith effect? Smith has a long history of improving his partner's numbers - and I would suggest that even FCP benefited from this effect as his playoff heroics just happened to coincide with his pairing with Smith. I'm not saying Smith is an elite dman - but he is clearly a #2 guy that fits well with a more offensive player and has been for a long time.

So for all the pissing and moaning about the defense - if we sign Sykora then this team has only ONE need to address - a single top pair dman to match with Smith. That's about 3 less critical needs than we had last year at this time when we needed a center, a RW, a dman and a goalie. This single need will be addressed as the right opportunity arises and the Oilers have proven time after time to be able to get the best out of their dmen.

8/10/2006 10:54 pm  
Blogger Black Dog Hates Skunks said...

that's an interesting point asiaoil - can anyone crunch numbers to back up his point re: Smith?

I'm with you on where this team is related to last season btw

8/11/2006 7:12 am  
Blogger Asiaoil said...

I think speeds or someone else did it for the period up to the lockout - and the positive impact of Smith on his partners was pronounced - he's a very good dman and FCP certainly benefited playing with him as we all saw.

8/11/2006 11:32 am  
Blogger oilswell said...

The problem I have with the "Staios wears down" proposition is that he's the D on the team that you'd expect to take on hard minutes if he's taking on more minutes per game. Simply put, if he's playing more minutes it seem likely the game difficulty has also increased, and this might be enough difference to explain any apparent shortfall in his game. A finer grained analysis of his ice time would be required. Props to RQ for doing that bit of legwork alone.

8/11/2006 12:26 pm  
Blogger Vic Ferrari said...

Terrific post riversq. I admit that I was one of the slobbering masses that bought into the whole "Staios wears down" thing. Clearly there is no evidence to support that.

I have a script that checks player's results by icetime, really for the Roger Neilson 'ten minutes of icetime' thing ... but used the same ditty to check on Staios and there is nothing that shows that he does worse or better with more icetime on a game by game basis either.

Of course if Huddy is throwing Staios out there even on 40 or 50 seconds of rest (real time, not clock time) then he must think that he is the best option. That a fairly tired Staios is still a better option than whoever the 5th or 6th options are for him there (Cross or Ulanov, anyone?).

The only thing of note: In games where Staios plays a whack of minutes ... the Oilers do worse when Staios is NOT on the ice. Which makes sense if you think of it from the root cause of the Staios icetime stat (which is of course the decision process of Charlie Huddy on the night.)

8/14/2006 11:40 pm  
Blogger RiversQ said...

Glad to see you were able to address oilswell's fine points, Vic.

It's interesting to see that result because I thought oilswell was onto something with the idea of more icetime increasing the number of tough minutes for a dman like Staios. That may still be the case, but he rises to the challenge nonetheless.

8/15/2006 1:38 am  

Post a Comment

<< Home