Last week, we took a look at the centers. If you haven't seen it, click on that blue link (opens in a new window). In addition to being a shameless plug, the statistics contained there do make a good reference point for some of the winger analysis, and I will make comparisons across positions, so if you'd like to have the other stats handy, go for it.

This week was meant to be the left wingers. I thought that by splitting left and right, I could save the article from being a jillion pages long. However, it turns out that Ryan Malone is the only left winger on the roster right now, so I decided that maybe combining the wingers wasn't such a bad idea. It is a little longer, so I spent more time on the top four and combined the bottom four into two entries.

What did I find in the winger analysis? Martin St. Louis is the best at (almost) everything. But you knew that. What else? I think our memories (and offseason panics) might overstate the importance of Sean Bergenheim and Simon Gagne and understate the importance of Steve Downie and Edward Purcell (yeah, I know, his name is Teddy. You try finding his stats when they're organized by first name and you don't realize that "Teddy" is somehow short for "Edward.").

More specifics after the jump, and as always, if you're scared of numbers, you can skip to the end. (Note: playoff numbers are not included in the tables, but in some cases I found it worth mentioning in the bullet points).

• The first thing you may notice is that Marty gets more even strength time than anybody we have at forward. That's not surprising, but you wonder whether that will start to decline or whether he'll remain ageless. Next, there's the whole "numbers being impressive" part. Remember from the last post how Steven Stamkos had 2.71 points per 60 minutes of even strength time and Alex Ovechkin had 2.59? Marty has 2.74, so he wins at everything. (To be totally fair to Stammer, he has the edge in goals and Marty in assists).
• When Marty is on the ice, the team averages 3.36 goals per even 60, compared to 2.17 when he's off. While very slightly worse than Stamkos' splits, this is still 2nd best among Lightning forwards and tops among wingers. And his power play numbers, while also slightly worse than Stammer's, are still impressive. The Bolts score 7.12 per 60 five on four minutes with Marty, to just 4.32 without him.
• I know Marty had some fine moments in the playoffs, but by and large, the entire first line was shut down. The team only scored 1.8 goals per even strength 60 with St. Louis on the ice, compared to 3.28 without him. However, Marty did record 1.8 points per even strength 60, so whenever his line did something, he was involved.

• Steve Downie was better than you think last year (unless you think a whole lot of Steve Downie). At least in setting up chances he was. Downie led all forwards with 1.65 assists per even 60, and it really wasn't close, with Marty coming in second with 1.46. In fact, Downie had more assists per 60 minutes of even strength ice time than Vincent Lecavalier had points. His goal total was much less impressive at .69 (significantly behind the leaders, albeit closer to Dominic Moore than Nate Thompson), but the man sets things up.
• Continuing on that theme, the Bolts scored 3.38 goals per even 60 with Downie on the ice and just 2.42 without him. I know he spent some time on the first line, but I didn't recall him living there all season. But whether he did or not, the team had first line numbers when he was in the game. His power play numbers were a bit weaker, with the team scoring 6.95 per five on four 60 without him to just 4.41 with him.
• His playoff numbers (which certainly did not come with the struggling first line) were jaw-dropping. He had 3.53 points per even 60, including 2.88 assists, and the team scored 4.49 per even 60 with him on the ice, compared to 2.34 without him. Each of those statistics are the best on the team. To put the assist totals in perspective, Henrik Sedin, who led the NHL in assists during the regular season, averaged 2.29 per even 60, less than 80% of Downie's playoff pace. Of course, that also tells you that it won't be replicated. But credit where credit is due for a fantastic (and somewhat overlooked) postseason.

• Like Downie, Teddy Purcell was better than you think, even though we have a lot of Purcell fans here (myself among them). His .88 goals per even 60 were third among Lightning forwards, behind St. Louis and Stamkos, and his 2.02 points per even 60 were fourth. Per 60 minutes of five on five ice time, Purcell was almost .4 points better than Lecavalier, which is four times the gap between Lecavalier and Moore.
• While Purcell's on/off splits weren't buoyed by significant time with the top line, they were still solid, with the Bolts scoring 2.76 per even 60 with him and 2.45 without him. But the real surprise comes from his power play numbers. The Bolts scored 8.35 per 60 five on four minutes with Purcell and just 4.88 without him. That's better than St. Louis and better than Stamkos.
• To confirm what we remembered, Purcell turned in a top notch playoff run, with 3.41 points per even 60 on 1.42 goals and 1.99 assists, a point production second only to Downie. That the top two point producers were on different lines says a lot about how much both the second and third lines stepped up in the postseason.

• Ryan Malone's goal-scoring numbers leave something to be desired, but his overall point production remained strong. His .38 goals per even 60 can be compared to Tom Pyatt, Ryan Shannon, or Adam Hall, but his 1.42 assists are just .04 behind St. Louis for second among forwards, giving him a respectable total of 1.80 points per even 60, slightly better than last year's numbers from Joe Thornton (I'd compare him to a winger, but I don't have a list of per game averages for wingers. I looked up a few, and he's well behind Alex Tanguay and Thomas Vanek but well ahead of Dany Heatley).
• With Malone on the ice, the Bolts scored 2.94 goals per even 60, compared to just 2.33 without him, better splits than everyone except St. Louis, Stamkos, and Downie. However, it's his power play numbers that are truly eye-popping, as the Lightning scored 9.36 goals per five on four 60 with Malone and just 3.48 without him, far and away the best on the team.
• Malone's playoff numbers, however, were near the bottom, as he had just 1.03 points per even 60. The team scored 2.57 per even 60 with him and 3.12 without him, so his line was only slightly below average, but his direct contributions to the offense took a major hit. In the 2012 playoff run that we all hope to see, Downie and Purcell are sure to drop from their 2011 numbers, but by that same token, you can't expect Ryan Malone's production to be this bad again.

• Hey, everybody's it's finally time for the guys that everyone's afraid to lose! (okay, John, I know you're not afraid. It's called exaggeration). Gagne and Bergenheim both produced roughly 1.7 points per even 60 last year, with Bergenehim being heavier on the goals with .8 (and .86 assists), and Gagne on the assists with 1.11 (and .59 goals). You might worry when I tell you that those numbers are better than Lecavalier's. But you should stop when I tell you that they're worse than Downie, Purcell, and Malone, who are among the players who will be charged with picking up some of the slack.
• During the regular season, the Lightning offense worked slightly better without Gagne and Bergenheim, scoring 2.51 per even 60 with Gagne to 2.56 without him, and 2.32 with Isberg to 2.63 without him, numbers that are decent but not top six.
• Combining these two players into one analysis might make the chart a little more difficult, as those special teams numbers for Gagne are power play goals for (6.22 per five on four 60 with him, 6.06 without) and for Bergenheim are penalty kill goals against (4.27 per four on five 60 with him, 5.07 without him). So Gagne's power play numbers are slightly above average, but nothing eye-popping, as are Bergenehim's penalty kill numbers (which are below Nate Thompson and Dana Tyrell but ahead of Dominic Moore and Adam Hall). Together, both sets of numbers show solid special teams play, but neither suggest anything irreplaceable.
• Yes, Bergenheim was awesome in the playoffs. He averaged more than a goal per period of ice team (3.03 per even 60). Yes, that's ridiculous. Yes, major props. No, he won't do it again. However, his playoff point production was still below that of Downie and Purcell.

• Again, I'm combining two players because there are way more wingers than centers, I don't want this to get too long to read, and their stats say similar things. Both produced about 1.15 points per even 60, numbers that are the worst of the wingers and compare roughly with Tom Pyatt at center.
• When Hall was on the ice, the Bolts only scored 1.63 per even 60, compared to 2.87 without him, again very poor offensive numbers, comparable to Dana Tyrell. Shannon was a bit better, with the Senators scoring 1.85 per even 60 with him and 2.09 without. While that on ice output is low, it is important to note that the difference between Shannon being on ice and him being off is not large. In fact, it is similar to the Lightning's production with and without Sean Bergenheim. The difference is that the Sens as a team don't score nearly as many goals as the Bolts.
• Adam Hall's penalty kill numbers were below average, as the Bolts allowed 5.16 per four on five 60 with him and 4.62 without. Ryan Shannon's were above average, as the Sens allowed 4.38 per four on five 60 with him and 5.34 without. Again, the Senators as a team are worse on the kill than the Bolts, but the on/off splits are similar to what Tampa got from Nate Thompson and a hair better than Sean Bergenheim.
• I didn't include faceoff stats because non-centers don't try them as much, but Adam Hall is a faceoff beast, winning 55% of his 655 attempts, best on the team.

IF YOU SKIPPED THE BULLETS, START READING HERE

Conclusions? Take home message? First, Marty St. Louis continues to be the best (or near the best) at everything.

Second, Downie, Purcell, and Malone all have very strong assist numbers and on/off splits, but their lower goal totals may cause them to be overlooked. Well, Downie and Malone's low goal totals may cause that. Purcell's third on the team in goals, so I don't know what his excuse is. But I do know the debates on whether to keep Purcell or Bergenheim were no debates at all (I know most, including our JM, picked Purcell, but it really shouldn't have been a question).

Third, Purcell and Malone both seem to have skill sets suited for the man advantage. Precise passing and good anticipation are huge when set up in the offensive zone, and their numbers showed it to a shocking degree, suggesting that they were more important for the power play than St. Louis and Stamkos. Of course, you can't have a power play with all passers on the ice, so they aren't exactly making Marty and Stammer obsolete, but their contributions are again underrated.

Fourth, losing Bergenheim and Gagne won't hurt as much as some feared. Both players put up either strong third line or weak second line numbers, and losing a couple productive third-liners shouldn't be cause for too much alarm.

Fifth, Ryan Shannon's personal numbers don't look very good, but his on/off splits are more encouraging. His performance relative to other Lightning players would put him at or near the bottom, but his numbers relative to the Senators suggest that he was their Sean Bergenheim, without the playoff numbers (which were hurt by the Senators not playing in the playoffs). The hope here is that Boucher can turn him into our Sean Bergenheim, now that our Sean Bergenheim has run off with a Panther.

Finally, Adam Hall is good at faceoffs. Like, really good at faceoffs.