How the three-headed monster affected Islanders' stats

Editor's Note: More interesting work from BenHasna, who gives us some food for thought inside the numbers of guys like Kyle Okposo, Bruno Gervais, Blake Comeau and -- well pretty much the whole team...

In some of my recent comments about certain players I talked about how the goalies affected their individual stats. I found for example that Kyle Okposo's bad plus/minus was largely caused by the fact that Martin Biron struggled horribly when KO was on the ice. KO was only -3 combined with Dwayne Roloson and Rick DiPietro, but -15 with Biron, who stopped only 85.37% of the pucks when KO was on the ice.

There are a few similarly interesting cases and I thought some might want to take a look at all numbers, although we honestly can't draw too many conclusions about the performances of the last season and certainly can't make any predictions about the future based on these numbers. But it's fun and if anything shows that +/- numbers can be distorted quite a bit. The big table and further explanations follow after the jump.  

Before talking about what kind of conclusions we can take and we cannot take, let's look at the big table first. It shows every player except for those who played only a couple of games or never got significant ice-time, such as Klementyev or Rechlicz for example. The data is as always provided by the great sites of Vic Ferrari. It includes even-strength numbers of all 82 games, obviously excluding empty-net situations.

Thanks to particularly this work of Tyler Dellow, we know a bit about even-strength save-percentage or more precisely about PDO, which is EVSV% plus even-strength shooting-percentage, on the team level. The idea there is that any particularly high or low numbers are not sustainable. We don't know a whole lot about these numbers on the individual level, though. It's not unlikely that the range is a little wider and that a particularly bad or extremely talented player is able to influence these numbers. But it's definitely not something that stands out when looking at the EVSV% of all players. Excellent (two-way) players like Crosby, Kesler, Okposo, etc. are towards the bottom of that list, whereas a few mediocre players rank at the top.

Even Brendan Witt managed to put together a relatively healthy 90.28% this season. And if we expected someone to make life really hard for the goaltenders, it probably would be Witt, who definitely wasn't on NHL level anymore this season and thus would be expected to allow more quality chances than other defenders. He certainly hurt his team and meant additional work for the goalies, but judging from these numbers here, the chances they gave up with Witt on the ice were not necessarily of higher quality than with everyone else.

However, luck seems to be what mostly drives this number here and it's probably fair to attribute some bad luck to everyone with a particularly bad number. That of course also means that we can't fault anyone. Biron's horrible number with Okposo for example was just bad luck/fluke. Biron proved with other guys that he is able to stop 91%+ of the pucks and Okposo proved his play doesn't pull down the numbers with Roloson for example. But it obviously means that Okposo's final -18 was almost entirely caused by bad luck. He could easily have been at around evens with an average save percentage.

Some interesting results

That's what we can do with these numbers therefore - tell who probably suffered a bit from bad luck last season. I'll present a few of the most interesting results below, but the table is of course pretty straight forward. As mentioned, most of the numbers don't mean a whole lot, but it's fun to look at them and might be interesting to get back to them at some point during the next season.

Blake Comeau's final -6 was affected heavily by the time he spent together with DiPietro. He went -8 in those few matches with DP, who stopped only 82.98% of the pucks with Comeau on the ice. Biron and Roloson were slightly below average with Comeau, too, and thus Comeau could easily have finished as a plus player last season.

Andrew MacDonald benefited from an enormous 95.14% with Roloson (+9) and was overall lucky to finish as a plus-player, as he was -3, at exactly average save percentages, with Biron and DiPietro combined.

Besides Jesse Joensuu (98.04%), in of course a very small sample size, Trent Hunter had the highest EVSV% with overall 93.99%. All three goalies did particularly well with Hunter on the ice. But that as said doesn't necessarily mean Hunter played outsanding hockey. It probably was just luck, although Hunter led the team already in 2008-09. He finished towards the bottom in 2007-08, though.

Mark Streit and Frans Nielsen have similar numbers, although they didn't spend particularly much time together. Streit had the same amount of ice-time with the other centers. Both however did particularly well with Roloson (around 93%) and struggled with Biron (around 88%). Streit, Nielsen as well as Dylan Reese and Matt Moulson were the only players who finished as plus-players without benefiting from particularly high save percentages.

The numbers of Freddy Meyer are just the other way around. Roloson struggled with Freddy (88.35%, -7), but Biron was excellent (94.44%, +5) with him.

Bruno Gervais had quite consistent numbers with all three goalies - consistently bad numbers, actually. All three goalies had roughly average save percentages with Gervais, but Bruno still finished -19 overall.