Have the New Jersey Devils Played Better Against Teams Better Than Them?

As we're past the halfway mark of the 2015-16 season, one of the comments I've heard and seen is how the New Jersey Devils do not play well against teams worse than them.  There are variations on that. Examples: The Devils play well against teams better than them. The Devils play to the level of their competition. The Devils show up for the good teams. The Devils don't show up for the bad ones. So on and so forth.  It's something that comes up when the Devils lose a game to, say, Columbus or beat a team like Dallas. Is it really true? Just like the thought that the Devils make backup goalies look good - which is not true - it's worth taking a closer look  Have the New Jersey Devils played better against teams better than them? Let's find out.

There are two questions within this question that I cannot fully and definitively answer.  What does it mean to be play better than an opponent? What is a team that's better than New Jersey? How those are answered will determine how to answer the initial question.  As far as the first question goes, I'll tabulate not just the record of the Devils in these match-ups but also the games' stats at Natural Stat Trick. That's Corsi for and against (all shooting attempts; abbreviations: CF and CA), Fenwick for and against (unblocked shooting attempts; abbreviations: FF and FA), shots for and against (abbreviations: SF and SA), and goals for and against (GF and GA).  It's not perfect but it'll give us a better idea of how the team performed.

The second question is a little more intriguing. From my understanding, the statement is directly related to the record of the opponent.  In Columbus' case, they were way below the Devils in the standings.  The opposite was true for Dallas. So to answer the second question, I'll be using the team's records.  If the opponent had a better record (more standings points) than the Devils, then they would be a better team.  If not, then not.  If there are any ties, then I'll use the NHL's tiebreakers. If they're still tied, then they'd automatically go into the not-better group since, well, the opponent wasn't better than New Jersey.  There is a bad but fast way of determining these groups.  There is a better, but a little more tedious way of determining these groups.  Let's do the bad way first.

The Bad Way

I put the numbers for this post together on the morning of Saturday, January 9.  At that time, the Devils were sitting fifteenth in the NHL.  Therefore, it was easy to identify who had a better record than the Devils and who had a worse record.  I just dumped the game stats from Natural Stat Trick and had to specify who was better or worse than the Devils.

As it turned out, the Devils played an even number of games between teams who have better records than them and those that do not.  They actually have picked up more points against the teams worse than them.  Their record isn't great, but 11-7-3 is better than 9-10-2.  Did they play better too? To a degree, yes.  The Devils have out-scored teams worse than them. They took more shots and had more unblocked attempts. However, they also conceded more attempts - all and unblocked - against than against better teams.  Furthermore, while they're better, they Devils have been out-attempted and out-shot by all opponents this season.  So while better than against the better teams, the Devils haven't exactly been all that good.  Except for the goals, of course.

Why is this a bad way? Because it's not accurate to what drives the question and/or complaint.  I just took the records of the Devils' opponents as of the morning of January 9, 2015.  The opponent may have a better or worse record now than what it was when they played the Devils.  A great example would be Friday's game against Boston.  By this method, Boston is part of that Better Teams group.  Yet, when they actually played the Devils, Boston was below New Jersey in the standings.  The big win for Boston moved them ahead of New Jersey.  That's not really holding true to the issue behind this canard.  Why would we do it the bad way?  Because it's easy and direct.  There's a better way.

The Better Way

The question at hand is whether the Devils play better (or worse) against teams better (or worse) than them.  This is tied into where the opponent stands when the Devils play them.  Instead of just looking at the standings on a given date and using that as our criteria, we need to look at the records when the Devils actually played them.  We put the records of the two teams in every game preview at this site.  So I pulled the information from there, calculated how many points the team had, and then determined what to do about any ties.  As an aside, there were three ties: one direct tie, one tie that the Devils won by tiebreakers, and one where I gave it to the opponent because 0-0-0 is better than 0-1-0.  Getting back to the issue at hand, I found out that the Devils have played many more games against opponents who were sitting in better spots than the Devils when they played.

I was surprised to see that the, yes, the Devils have put up a better record against teams with records superior to New Jersey at that time.  In contrast, the Devils have fallen on their faces more often than not against teams with worse records.  It means there's some evidence behind the thought that the 2015-16 Devils played better against better teams, with better defined as having superior records than them at that time.   The more accurate statement may be that the Devils get better results against teams better than them. But since many equate results with performance (e.g. you win because you play better), I get why it's said in both the positive and negative fashion.

Why is this case? Let's go back to the first quarter of the season. The Devils started off the season with three straight losses and it turned out that most of their opponents did not similarly fall flat on their faces.  It wasn't until their eighth game of the season that the Devils played a team who had a worse record at them when they played their game (Buffalo, 10/24).  Through October and November, the Devils played six games where their opponent did not have a better record than them - and one of them was a exact tie: Chicago and New Jersey were both 8-6-1 on 11/12.  The Devils did start moving up in the standings in those months, so their record against better teams got better.  They never had a long hot streak or any real losing streak beyond the first three games.   By luck of the schedule and opponents not having a slow start, that kept them just behind plenty of opponents.  As the season goes on and now that New Jersey is slumping, the disparity in games may not necessarily change so there's a shot this could continue if the Devils get out of their slump (hopefully soon).  Still, that first quarter of the season was heavy on the opponents with better records and as the Devils kept succeeding, they put up the better results.

Now, let's go back a bit.  I don't believe that results are necessarily equal to performance. Not in a sport like hockey. So let's look at those other numbers. Have the Devils played better against the better teams? It's not so clear. As with the bad way of doing this, the Devils have been out-attempted and out-shot by their opponents, better or not better of them.  The Devils have been better in terms of attempts against the worse teams than the better teams.  That may be the result of the Devils playing better, though being down in games should yield more attempts for the losing team - the Devils in this case - since they are down.   Shots favor the games against better teams slightly, though; and the Devils have out-scored those better teams.  Both of these stat lines gives further notion that the 2015-16 Devils may not be all that and a bag of chips, but it's not readily apparent that the Devils have performed better against worse or better teams.

I will admit that the numbers I've used so far from Natural Stat Trick come from all situations play in a game.  Corsi, Fenwick, and shots can be influenced by the amount of special teams in a game. So let's look at the two records with 5-on-5 stats only.  5-on-5 play is the most common in hockey, after all.  If a team is playing better than another, then this is where we should see it.  This filter may give us a better idea of whether the Devils have played better against teams who had worse records than the Devils when they played or against teams who had better records than the Devils when they played.

OK, it's as clear as mud .  The attempts have remained in favor of the games against worse teams, but when we get away from Corsi, the percentages rise for the games against better teams.  I'd like to think that there's some kind of score effect in play seeing that the Devils remain positive in goal differential in the games against better teams than the games against worse teams.  Do I know how to account for that? No.  Not yet, at least.  Again, between both groups, the Devils are still being out-shot and out-attempted in general.  That does not bode well for the currently-good 15-7-3 record against teams who were better than New Jersey when they played nor the 5-10-2 record against teams who were behind New Jersey in the standings when they played.

Your Turn

My answer of the original question is yes, there's something to it.  In a results-oriented business like hockey, the Devils have the better results against teams who were ahead of New Jersey when they played. They have worse results against teams behind them in the standings.  However, I'd be careful on using the phrase "play better" because it isn't that clear.  From a pure goal scoring standpoint, yes, the Devils have been better against the better teams. When you start going deeper and looking at attempts, it isn't so clear.  With both groups, the Devils are still being out-done by their opponents in general.  So even if you see these numbers conclude, "To me, they're better," better isn't all that good.  Because of that, we could be saying that the Devils may be better against the better teams but they still have losing records as the season goes on.

I hope this was somewhat instructional like the post last week about goalie stats. And I especially hope you know that the better way to do this really is the more accurate way than the bad way.  Now I want to know what you think about this post. Were you surprised (like me) to learn that the Devils have a better record against teams with better records when they faced off?  How would you define a game to have been one where the Devils played better in beyond whether or not they won?  Please leave your answers and other thoughts about this post in the comments. Thank you for reading.