Goaltender usage on back-to-back games

How NHL coaches handle their goaltending breakout has long fascinated me. And, not to say a certain strategy is right, or that it is wrong, but it's very clear that current head coach Mike Johnston and former coach Dan Bylsma had two very different strategies forming for how to handle back-to-backs.

But first, to understand why this is potentially important, a little information on how games in 2 straight days tend to effect the hockey team. NHL.com (back in November) had this interesting data:

Overall, teams are just 37-45-8 in the tail-end game of a back-to-back. That includes a 12-17-0 record at home and a 25-28-8 record on the road. Of those 90 tail-end games, 51 of them have seen the tired team allow more shots compared to the previous night. That means 56.7 percent of tail-end games will see the goalie face more shots and likely more scoring chances behind a tired team.

Overall save percentages in tail-end games are way below the "quality start" category. Starting goalies have a cumulative .904 save percentage in tail-end games, stopping 2,474 of 2,736 shots. That number does not include the workload of replacement goalies thrown into these games after the starter was pulled due to performance or injury. This has happened 10 times out of the 90 total tail-end games.

The most alarming back-to-back statistic is the 4-13-0 record in tail-end games when a goalie starts both games.

And there is other research out there (that isn't readily available now) that shows very clearly teams suffer on the second game of a back-to-back series. And, that's obvious- you're tired, usually there's some travel involved, routines like the morning skate and practice are mostly skipped, it's tough to do.

That said, how would you deploy your goalies knowing it's tougher sledding on the second night makes personnel decisions of which goalie to start all the tougher, and more interesting.

I made the chart below. The @ symbol indicates (as you may have guessed) a road game for the Pens. The "better opponent" column points out which team finished higher in the standings for 2013-14, or is higher in the standings as of right now for 2014-15.

Of the 17 back-to-back games, Bylsma pretty clearly was shielding Jeff Zatkoff. Only twice was Zatkoff challenged with having to play the tougher team and 2 times choosing to play Marc-Andre Fleury in BOTH games (which almost always proves to be a very bad decision, NHL goalies are not good having to play 2 days in a row). And, indeed, Fleury has had to been pulled in each of his past 2 games where he starts 2 days in a row.

Further showing Bylsma's protecting of the team, in 11 out of the 17 instances (65% of the time), he elected to start Fleury on day #2 of the b-2-b, when the team is more worn down and likely to have a tougher time.

Let's contrast that with this season so far:

Can you see the subtle difference so far? Fleury has played the first game of a b-2-b 6 out of 10 times (60%) this season, that number was only 8/17 (47%) last season. This year, Fleury only seems to play the back game when it's a clear important matchup (like NYR on 11/15 or CHI on 1/21) or getting a break the first day after the all-star break.

Other than that, including the games that begin tonight, Johnston has gone with Fleury in the first game more frequently. He has also tossed the toughest team Thomas Greiss's way 4 times (including 3 times on the road on the second night) for tough situations. Greiss has already had tougher games in these situations than Zatkoff did all of last season.

---

I will be interested to see what the consensus is about the way Bylsma and Johnston have handled their decisions on goalie starts. On one hand, I think Bylsma had a very smart system to protect the team by frequently playing MAF on the second night, and almost always playing his best goalie against the best team they would play.

On the other hand, Johnston seems to favor playing his best goalie on the night the team has the best chance of winning, statistically speaking, which would be the first night. It demands more of the backup goalie, but Greiss now is a lot more established (and frankly, better) than a fresh Zatkoff with 0 NHL experience last season.

Matchups don't always dictate the results of wins and losses, and naturally it's the 18 skaters that tend to get tired and have to pull their weight on these nights as well to help whomever is in net out. There's no set way of how to skin this cat, but the differences between the two coaches are starting to standout.