WOWY. Also known as With Or Without You. It’s not just a U2 love song, in fact it’s a pretty useful tool for looking at how players perform with certain teammates, what players are driving their lines, which ones are dragging their lines down, and which combinations seem to work at bringing the best out of each other. The basic data came from Stats.HockeyAnalysis.com, and boy do they have a lot of data. Within those pages you can find WOWY details for pretty much every player over the last few seasons, combined seasons, when the score is close, and much more. For this post, I used 5v5 data from the 2013-14 season only (to keep the sample size as large as possible). So I took all that information, and made a nice big fancy chart. It shows the Corsi percentage that each player is better (or worse) with each player. So the corresponding space in the chart of Johansen With (along the top of the chart) and Foligno (on the left side of the chart) shows the difference between Ryan Johansen‘s Corsi For % with Nick Foligno, and Johansen’s Corsi For % away from Foligno. Got it? Good. Let’s take a look at the chart, then go over a few things that need to be considered. After that is the good stuff: the players who come out best (and worst) in this analysis, the best combinations of players, and various ideal/terrible/etc. lineups.
First off is the issue of sample size. Not every player even played with every player. See those two blank spots in Horton’s row? He never set foot on the ice (at 5v5 at least) with Boll or Atkinson. Then you have situations like Matt Calvert. He played with everyone, but he only played significant minutes with two other forwards (Atkinson and Dubinsky), but he was pretty awesome playing with a whole bunch of other players.
The next major issue is accounting for usage. For example, Jack Johnson played almost exclusively tough minutes, against top players, with defensive zone starts. Ryan Murray played fairly easy minutes (comparatively speaking), so does the fact that he is worse with Johnson mean Johnson is terrible, or Murray played much tougher minutes when playing with Johnson (versus James Wisniewski, his usual partner).
There are a couple further issues regarding specific players. You’ll notice I only included players who crossed the twenty games threshold (in a CBJ uniform). So this didn’t end up including Jack Skille, Matt Frattin, Nick Schultz, Michael Chaput, and others. Two players it did include were Marian Gaborik and Corey Tropp. They were the only two Jackets players included here who also played with another team this past season. A quick look at their respective columns makes the issue pretty clear. The WOWY numbers include the entire season, so it includes Gaborik’s time in Los Angeles and Tropp’s time in Buffalo. So Gaborik was better off in LA pretty much across the board, meaning his numbers were worse off with pretty much every single Jacket. Buffalo is terrible, so Tropp’s is basically the exact opposite.
To recap: There are sample size issues, so take all of this with a grain of salt. There are usage issues, so it’s a little tough to completely differentiate whether it’s the player who is the problem (or boost), or the minutes the player plays. Finally, Tropp and Gaborik are a little messed up due to their time with LA and Buffalo.
First off, let’s look at the individual players. To check out an individual player, find them on the left side of the chart and follow them across. Lot’s of green in their row means good things. Lot’s of red means bad things. So let’s kick it off with by FAR the most impressive Jackets player.
James Wisniewski: Every player, with the exception of Gaborik, Prout, and Savard, is better when James Wisniewski is on the ice. That makes complete sense as well. Gaborik’s stats are a little screwed up by the Kings, while Prout and Savard are also right-handed defensemen. That means that someone was playing their wrong side when they were on the ice together, which makes the slight drop offs logical. Wisniewski also has the benefit of easier usage than most of the Jackets defense, which no doubt contributes here. However, Ryan Murray had slightly easier usage and didn’t fare even close to as well as Wisniewski.
Blake Comeau: Next behind Wisniewski in terms of most positive WOWY’s. Which makes is super disappointing that the Jackets aren’t interested in even trying to bring him back. Considering Comeau didn’t play particularly easy minutes, coming in this high here is pretty impressive. Looking at the chart, only Johansen and Horton were significantly worse with Comeau than without him. Which makes sense since Comeau is clearly not a top line forward so he drags down actual top line players. Beyond that, Comeau is a very versatile forward who can play anywhere in the bottom nine and do reasonably well.
Wisniewski and Comeau came in quite a bit above most of the roster, but I also want to point to Nick Foligno and Ryan Murray, followed by Marian Gaborik and Matt Calvert, as other players who most players were better off playing with than without. The other side of the coin is pretty much exactly as anyone who watched a handful of CBJ games would have guessed.
Jared Boll: The biggest drag on the team. Only four players were positive with Boll, likely due to very easy zone starts when playing with him. The Boll contract continues to be a massive mistake. He is a disaster is pretty much every way (he’s not even that good of a fighter!), and everyone who comes near him just plummets. Reminds me of this.
RJ Umberger: Next worse behind Boll. Only Boll is significantly better with Umberger than away from him. Basically every other player is worse with Umberger than without him (with Tropp being marginally better). There was a time when he was a pretty good player, but those days are likely behind him. I would not be surprised at all to see him the victim of a compliance buyout if he can’t be dealt.
Jack Johnson: Next up on the bottom of the list. MacKenzie and Tropp are the only players significantly better with Johnson than away from him. Johnson has been hammered enough all over the place for bringing everyone down with him, so I’ll go easy here. But check this out for more details on my thoughts on JJ.
The next step I took was to calculate the best forward combinations. I did the same thing for the defense, but that pretty much just gives you the best combinations for defense pairings, so we’ll get to that later. Below is a list of all forward combinations who played more than 150 minutes together (to root out some of the sample size issues). This gives us a good idea of forward combinations that work, and forward combinations that don’t. Below is the list, ordered from best to worst, with the positive combinations in green, the combinations that basically break even in yellow, and the negative combinations in red.
There are definitely some interesting nuggets to glean from this. The biggest takeaway I have though is regarding Ryan Johansen. Horton and Jenner were both terrible combinations with Johansen, while both of those players were great fits with Artem Anisimov. Meanwhile, the Johansen-Umberger duo looks pretty good. This is almost solely due to Johansen dragging Umberger up a ton, as Johansen was very slightly worse with Umberger than away from him. A combo I liked that also came out very well here was Johansen and Foligno. I love those two together and really hope they continue to see a lot of time together moving forward. I also thought it was quite interesting that the vaunted Atkinson-Calvert duo more or less broke even. Considering that Calvert was awesome with pretty much every player he played with, I’d really like to see him get some chances with some other linemates for next season.
Best/Worst/Actual Forward Combinations
Now for looking at line combinations. I did this the same way as above, basically taking the sums of the various combinations of WOWY’s. I did this with something of an eye to the future, so I made sure there was at least one possible center on each line (hence Boone Jenner as a center). So the best line came out as Calvert-Jenner-Gaborik, which breaks down as the Calvert-Jenner, Calvert- Gaborik, and Jenner-Gaborik having the highest combined WOWY ratings. After finding all 346 line combinations, I created an “ideal” lineup (consisting of the four best possible line combos), a realistic lineup (consisting of realistic lines comprised of players who finished the season with the Jackets), checked out the actual most common lines (per Left Wing Lock), and finally the worst possible line combinations (just for fun).
This lineup makes no sense obviously, although it would have been interesting to see more of the Calvert-Jenner-Gaborik group. Foligno makes sense pulling up usual fourth liners MacKenzie and Comeau. I have no idea how that third line happened. That fourth line actually looks like a pretty damn good line.
After putting together the first lineup set, I realized I needed to make something more realistic. So I removed Gaborik from the equation, and tried to put together something that made sense line-wise. The top line is pretty awesome, as they can play tough minutes, have two great faceoff guys, carries offensive pop, and keeps the Johansen-Foligno pairing together. The second line would be a more offensive minded group, as all three of these guys played well with the others, albeit in mostly limited roles.
Interesting note on Matt Calvert: Of the 346 possible line combinations, Calvert was a member of the top nine lines, and twelve of the top fourteen. Cam Atkinson was on zero of those lines. I like the Calvert-Atkinson duo, but, as I said earlier, I would really like to see Calvert moved around the lineup more. He’s a versatile player and could be incredibly useful playing with some other forwards.
Moving on to the 3rd line, we have another offensive minded group. As I mentioned earlier, I would prefer to keep Comeau and pair him with Anisimov and Atkinson. The fourth line is pretty much the fourth line. Note the absence of Jared Boll and RJ Umberger in this set.
These actual lines came out a little worse than the realistic lines above (and about half of the “ideal” lines). However, the lines were greatly in the positive. For perspective, the “ideal” came to a combined sum of 253, while the realistic and actual lines came out to 116.9 and 90.9 respectively. This shows that Todd Richards did a pretty good job of putting lines together. It’s also WAY better than the worst lineup below.
It’s hard not to laugh out loud at this. Combined -399.3. Damn. That is terrible. These lines make no sense either. Comeau isn’t skilled enough to play with Joey and Horton. Umberger can’t even dream of skating with Gaborik. Tropp and Atkinson on one line is a huge skill disparity. Any line with Boll is clearly the fourth line. Foligno and Arty do not belong on fourth lines.
Best/Worst/Actual Defense Combinations
Fortunately the same kind of analysis for the defense is quite a bit easier than for the forwards. We can just look at the best combinations of players, and project the pairings from there. The defense has a clearer delineation in their roles compared to the forwards as well, which makes this even more beneficial.
These pairs are both the ideal and are actually realistic. The top pair would be offense-centric (or pretty much what the Murray-Wiz pair was this past season). If Nikitin isn’t brought back, Tim Erixon would be a solid fit on an offense first pair, or a free agent pick up could do the trick. Murray and Tyutin would make a great two-way pairing. They can play tougher minutes than the Wisniewski pairing, and take some of the defensive burden from the other pairing. Johnson and Prout were a net negative, but they were less negative than many of Johnson’s defense partners, and those minutes came with some tough usage, which would continue here. That negative though is overcome by the other two pairings, which just so happen to be the best two combinations of defensemen, and two of the four best combinations of players (of those that played significant minutes together). Prout barely beats out Savard, not by really being better than him, but just by being a better fit alongside Johnson. This isn’t really a 1st pair, 2nd pair, 3 pair kind of defense corps, but I think that is a better fit with the Jackets personnel. It would be great to have a clear number one pairing, but until Ryan Murray gets to the point of being the clear number one defenseman, they won’t have that. An even split in ice time between these three pairs, with the usage splits mentioned above would make for better results than this past season.
Like the above grouping, this pair is also listed best to worst. These pairings just don’t make quite as much sense as the previously mentioned pairing. The Murray-Wiz pair was good offensively, but not quite as good as Nikitin-Wiz. Tyutin-Johnson was slightly better than Johnson-Prout, but Tyutin just isn’t a good fit in that kind of role. He’s better off on a two-way pairing, as is Murray, and those two were pretty awesome together. Nikitin-Prout is a clear third pairing, with Savard again missing the cut.
Murray doesn’t appear to be quite ready for the type of minutes and usage Johnson plays (although maybe Johnson isn’t ready for that yet either). I went over Savard and Wisniewski as a pair earlier as well. Here is the oddest result of anything that came about in this research: Nikitin and Tyutin were not only the worst defense pairing the Jackets iced this year, but they were the worst pairing of players, and by a significant margin too. Both players were at least 23% better away from the other one. That is just an insane drop for a pair that was raved about a couple years ago.
So what does this all mean? Well a lot, but how significant it actually is can be debated given issues I mentioned earlier (sample size, etc.) There are a few takeways, however. The primary thoughts I’m left with after spending far too much time looking at this are as follows: James Wisniewski is awesome, Nathan Horton and Ryan Johansen probably aren’t a good fit together, I hope Matt Calvert gets a few chances to play with some other linemates moving forward, the defensive pairings could be shuffled a little to great positive effect, and good riddance to RJ Umberger.
WOWY is an interesting thing to keep an eye on, and something that always needs to be considered when it comes to looking at team based stats like Corsi. Hockey is a fluid sport, and sometimes two particular players are able to make each other better (think a great playmaker helping a solid shooter get more scoring chances, etc.). Sometimes players make each other worse (think two guys who like to head to the front of the net, if two guys are there, you’re in trouble, and their constantly thinking about who is going to go where). Finally, you can also have players who make another guy better while making themselves worse (think a grinding in the corners guy playing with a power forward type, as the grinder then frees the other guy to do more). Those are a few off the cuff examples, but the point remains. Being successful has almost as much to do with finding the best fit for players as it does assembling the best team on paper. Talent helps, but if that talent doesn’t fit or isn’t used to the best of their ability, then it isn’t much use.