The Conundrum of Ilya Kovalchuk's 5-on-4 On-Ice Impact in Last 3 Seasons

Yesterday, I explored the even strength on-ice stats and on-ice impact, the change in those stats, of Ilya Kovalchuk from the last three seasons.  While he showed some improvement, he was only among the best NHL forwards in terms of on-ice shooting percentage, goals for per 60 minutes (GF/60), GF/60 impact (change in GF/60 when he's off the ice minus when he's on the ice), and for 2009-10, PDO.    In other stats like goals against per 60 or shots for per 60, Kovalchuk not only didn't rank all that well but didn't even provide a positive on-ice impact when on the ice.    While Kovalchuk has definitely been a big contributor, he's not at all like, say, Pavel Datsyuk or Zach Parise, where the team performs so much better when the stepped on the ice.

When I was pulling those numbers from Behind the Net, I thought to myself that perhaps Kovalchuk would shine in on-ice numbers and on-ice impact for the power play. After all, NHL coaches has given him bucket-loads of PP time and he has 117 power play goals.  34.6% of his career total of goals scored came on the man advantage. Surely, he'd stand out, right?  

Behind the Net keeps track of 5-on-4 stats as well, so I figured I do what I did for Paul Martin a few weeks ago and look at his special teams performances.  Only the power play, as Kovalchuk's penalty killing time has been minimal.  What I found befuddled me.  Please continue on after the jump so I can explain what I found.

Introductory Explanation

I pulled the 5-on-4 on-ice/off-ice numbers from Behind the Net from the last three seasons.  I filtered out all NHL forwards who didn't play 30 games and had a time on ice per 60 minutes of 2.  This way I looked at forwards who played significant roles on their team's power play. 190 forwards met this criteria in 2007-08; 193 in 2008-09; and 183 in 2009-10. 

I've focused on four stats, two on-ice stats and two for on-ice impact.  The on-ice stats are goals for per 60 (GF/60) and shots for per 60 (SF/60).  Since these are 5-on-4 situations, I want to know how the team's scoring and shooting rate were when the player was on the ice. The impact stats are GFON IMPACT and SFON IMPACT, which are the off-ice stats for GF/60 and SF/60 subtracted by the respective on-ice stats.  This was the methodology I've been using across all of these posts. For the impact numbers, the lower the negative number, the better.

I did not take on-ice shooting percentage, PDO, and adjusted CORSI/60 into account because I'm more interested in how the player affects his team's offense in offensive situations.  On-ice shooting percentage won't say much, PDO is pointless since the on-ice save percentage is generally going to be incredibly high, and the team with the advantage usually attempts a lot more shots so everyone's going to have a high value there.    Likewise, I'm not interested in shots against or goals against. 

Since power plays usually feature the team's best offensive players against penalty killers, who are the best defenders not necessarily great players at both ends, QUALCOMP was generally low and QUALTEAM was generally quite high.  So for context, I'm using TOI/60, as being out for more time subjects a player to less success which will hurt the stats more so than someone at the minimum TOI/60, and total power play goals for on the ice, as the best power play performers should be present for a significant amount of goals.

Note: I'd provide links to the tables for all of the NHL forwards themselves, so you can confirm this for yourself and check out how other players performed. Yet, they are so large that they are even harder to read than the ones from yesterday's post.  Until I find a better solution, please bear with me.

Ilya Kovalchuk's 5-on-4 On-ice and On-ice Impact Stats and Ranks

Here's where the conundrum hits.  Despite being on the ice for a lot of power play goals, Kovalchuk actually didn't rank very high in these stats.

In 2007-08, Kovalchuk's on-ice stats for GF/60 and SF/60 weren't good but he provided a significant enough positive impact to rank among the best in the group analyzed.  That was understandable and what only struck me as a little odd was that he was on the ice for only 39 goals.  Not that's bad or anything, the closest Devil was Zach Parise at 26. 

As for 2008-09, well, I don't know what to say.  Below half among all 199 forwards in the group analyzed in terms of ranking isn't good.  The GF/60 and SF/60 actually got worse when he stepped on the ice.  Yet, amazingly, he was on the ice for 47 goals.  The put him in a tie with teammate Todd White and Evgeni Malkin, and only 4 forwards were on the ice for more 5-on-4 goals than that.  Productive, yes; but a major impact for the Atlanta power play, not so much.

2009-10 had him on the ice for 6 fewer goals (12 forwards; but his rankings improved across the board, his on-ice impact for both goals and shots were positive, and his on-ice GF/60 and SF/60 remained steady.   While his on-ice impact for SF/60 was good; he still remained below half in all else. 

After three seasons, I was astonished that Kovalchuk wasn't providing impact or great rates for goals for or shots for on the power play.  It's clear that he scores power play goals and he's present for so many of them.  How could this be?  It'd be one thing if he ranked outside of the top 30 in most of these stats but still above the halfway mark; but this can't be rationalized.

I thought that it could be a result of Kovalchuk's ridiculously high TOI/60 for 5-on-4 situations.  As we've seen in New Jersey, Kovalchuk can and has been out there for nearly the entire power play. Therefore, these numbers are the result of playing with the top and the secondary unit. Whereas most players are committed to a single unit, having to endure a drop-off from the first to the second unit may hurt his numbers.  Who honestly plays as much on a power play as Kovalchuk?   After all, since these stats are per 60, so unless Kovalchuk's present for a ton of 5-on-4 goals (e.g. Alex Ovechkin's 74 in 2008-09), more time without scoring will make them look worse than they really are.

No forward broke 5 TOI/60 in 2009-10 other than Kovalchuk, so let's go to 2008-09 for examples of someone who did.  One of them was Alex Ovechkin in 2008-09, who not only ranked in the top 30 in all 4 stats, but had the very best on-ice impact on goals for (GF/60 went up by 7.63) and shots for (SF/60 went up by 30.8).  He didn't play as much (only 4.74 TOI/60), but still ranked in the top 30 in three out of the four stats (6 players did a little better in on-ice SF/60).   So there's one player who disproves my thought.  Another big-minute PP player has been Evgeni Malkin.  Again, back in 2008-09 when he also broke the 5 TOI/60 mark, he had on-ice numbers that ranked just above half, but his impact on those numbers were among the league's best (GF/60 increase by 3.9; SF/60 increase by 14.7).  Another player whose performance doesn't mesh with my theory on Kovalchuk.

Moreover, even if Kovalchuk was on the ice for more goals, that doesn't necessarily mean the SF/60 would change much.  Plus, his TOI/60 may go down a little bit because 5-on-4 time would end sooner and thus end the situation. So my original thought of TOI/60 being the big factor doesn't really hold up based on what I found.  Not to mention that he actually had a significantly positive impact on SF/60 and GF/60 in 2007-08.

Maybe the explanation is more relevant to his even strength on-ice and on-ice impact numbers?  I'm referring to the larger point that he didn't provide much of a positive impact on most of those 5-on-5 stats.   The comparison, I think, meshes well in terms of SF/60. If the shooting rate suffers or isn't all that high, then the GF/60 can suffer.  Basically, if he wasn't bringing a lot of positive impact when he stepped on the ice in terms of scoring and shooting rates; then I don't think it's too much of a leap to suggest the same follows for the 5-on-4 situations.

However, he didn't provide a big impact on 5-on-4 GF/60 as he did on 5-on-5 GF/60.  Plus, he did have a big impact in 07-08, just not in 08-09 or 09-10.  So that's not a very good explanation.

I'd love to hear what other people may think about this occurrence. Maybe it's something out of Kovalchuk's hands entirely?  Maybe it's a function of the power play (doubtful, as you'll see at the end)? Maybe there's something I'm missing?   Maybe I have to ask why he got worse in 2008-09 and didn't do as much in 2009-10. Please let me know in the comments.

A Couple of Concluding Thoughts

Between these surprising findings and yesterday's findings at 5-on-5 hockey, I'm growing more convinced that Kovalchuk's value isn't in providing an impact when he comes on the ice.  He's an excellent shooter, he can and has played a lot of minutes at forward, he's not a sieve on defense, and he contributed a lot, as per GVT (2009-10 example here).   Those are all good functions.  However, there doesn't seem to be a whole lot of evidence to the argument that Kovalchuk's presence makes everyone better.

This includes the power play. However, because of his talents and his contributions (it can't be coincidence he's on the ice for so many PPGs), coaches continue to give him a ton of ice time, as shown in the last three seasons.   I openly suggest to whoever that will be coaching to consider not playing him across two units.  It's not like Kovalchuk will forget how to shoot; but consistently playing him with the same personnel will allow his impact to be clearly observed. 

For example, look at Zach Parise:

Not as many goals, mind, but again, when he comes on the ice, positive things happen - even on the power play.  Something Kovalchuk can't claim.

Thanks for reading.  Please leave your thoughts in the comments about Kovalchuk's on-ice and on-ice impact on 5-on-4 situations from the past 3 seasons.   Surprised to see he didn't have much of an impact, despite how much he's scored and been present for power play goals?  Just as surprised to see that Zach Parise had much better numbers despite not being on as many power play goals?