Toronto Marlies 2013-14 Analytical Year In Review

Josh Weissbock is a writer for NHL Numbers and Canucks Army who specializes in hockey analytics, specifically for the American Hockey League. Josh was gracious enough to provide some data and explain what it means on the 2013-14 Toronto Marlies. Josh’s thoughts are in regular text, while mine are in italics. Josh can be followed on Twitter.

First and foremost the most a single goalie played in was 48 games so we do not have enough data to say how they will play in the future but we can analyze how they have been this year. With the small sample size it is hard to say which way they will regress.

Drew MacIntyre had a phenomenal save percentage of .917, well above league average. His career average is .918 so it’s not hard to think he’s playing above what he is. What is a bit more concerning is some of his fancy stats. His shutouts are low but that’s not really his fault. His quality starts is closer to average/on the low end of “good”. He does have a low number of really bad starts, which is always good to see while his bailouts are zero, so his team is not inflating his wins.

MacIntyre’s low number of quality starts could be a reason to dial back expectations for him next year. He’s had a spectacular Calder Cup Playoffs run, but that’s an extremely small sample so should be taken with a large amount of salt. With that said, his strong season is reason for optimism. If he doesn’t win the backup job for the Toronto Maple Leafs next season, he could be given an opportunity with another NHL team.

Garret Sparks has played about half as many games as MacIntyre which means we can trust these results even less. This is his second AHL season and he posted a .915, which puts him at .913 on his career. He started in 18 games, had an average number of quality starts, a high percentage of really bad starts and zero bailouts.

Sparks improved quite a bit as the season went on. He was sent down to the Orlando Solar Bears of the East Coast Hockey League in November, and after early struggles he rebounded well, putting up a .915 in 10 games. He was called back up to the Marlies in December and with the exception of one short demotion back to the ECHL, played the rest of the way in the AHL. He basically stole the backup job from Christopher Gibson. Sparks turns 21 in 22 days.

Gibson has even less data and it is really hard to properly analyze his play. This is his first AHL season and he posted a .916. He is a rookie so the numbers are optimistic. He had an average number of quality starts, an average percentage of really bad starts and zero bailouts. He also had a really high save percentage that started to dip towards the end of his 10 games.

Gibson was kind of Bizarro Sparks as far as his season went. He started off hot, but faded, playing most of the season in Orlando. Probably the most worrisome thing about his year was his .892 save percentage in 21 games in the ECHL. This might be one of those weird instances where a goaltender plays much better at a higher level because the defense in front of him is better. As Josh mentioned, it was his rookie year and a very small sample, so we can’t read too much into it.

I am almost convinced that the Marlies have found a way at the AHL level to sustain high save and shooting percentages that I haven’t seen elsewhere. Every goalie that plays for them consistently has high numbers and I have only seen one instance (Gibson) with a game less than .780.

In the playoffs through eight games (at the time of this writing) they have a .955 save percentage and a 13.57 shooting percentage. This isn’t sustainable but they’ve seemingly been doing this for a while.

Toronto’s team possession has been league bottom, hitting almost 40 per cent at times. I am not sure if this is because of players who have left due to injury/call ups or not so this should be looked into.

As we can see from the graph, the Marlies big possession swoon came from about Game 28 to 52. Unsurprisingly, that coincides with the team missing key contributors in Trevor Smith, Peter Holland and Carter Ashton due to either injuries or call-ups. Still, three players don’t make a team.

On the year the Marlies averaged 48.59 per cent in possession when the score is close, putting them at 22nd in the league. Their PDO was slightly above average at 101.24 per cent with a 9.66 shooting percentage (higher than average) and a .9158 save percentage (higher than average). They also got some good efficiency out of their special teams.

One interesting thing is we can compare the points/82 for the AHL team vs. the NHL team over the years and usually a good AHL team predicts a good NHL team in a few years. We can sort of see that trend with the Marlies and Maple Leafs.

It’s disappointing the Leafs didn’t see much improvement in the years following 2007 despite the fantastic team the Marlies iced that year. Unfortunately, they didn’t have much young, top-end talent. With three straight seasons of 104 points per 82 games, you would hope the Leafs could improve as some of those players graduate. That will depend on how Ashton, Holland, MacIntyre, Smith, Jerry D’Amigo, T.J. Brennan, Stuart Percy, Greg McKegg, Josh Leivo, Petter Granberg, Andrew MacWilliam and others develop.

[table id=5 /]

When we look at the players’ data we can start identifying who stood out the most. I’d say there are three players with elite shot production, with a few more border lining it. Ashton and Smith were really good but in a short sample, but it’s hard to know if they would sustain that. Brennan had the best shots on goal per game rate over 76 games. Brennan had a normal shooting percentage and had a good individual point percentage value. His point production was high in the AHL. Players like Ashton, Smith, Holland and Leivo all had shooting percentages higher than normal and will regress so don’t trust their points per game. Players like Wade MacLeod will regress upwards.

[table id=3 /]

[table id=4 /]

At even strength, Sam Carrick was probably the best of all Marlies with a really high 20.19 even-strength relative goals-for percentage over 62 games. Spencer Abbott was right behind him. Granberg was one of the best defensemen. Brad Staubitz, Brad Ross, Kenny Ryan and Kevin Marshall were all very bad.

That’s a very encouraging sign for Carrick, a fifth-round pick in 2010 who turned some heads this season. We already know Abbott is an elite AHL scorer; now we know he some very good possession skills too. I’ve mentioned this many times but I’d give Abbott a long look at Leafs training camp and possibly even audition him for a second-line wing spot.

[table id=6 /]

The most surprising thing about the time-on-ice estimates was the low amount of minutes given to Troy Bodie and Jerred Smithson. Both these players played chunks of the season with the Leafs, while skilled players like McKegg and Abbott hardly sniffed the NHL. David Broll, another popular choice among many to make the jump to the big club, averaged less than nine minutes per game. Conclusion: Call up skilled players please.

[table id=7 /]

MacWilliam played against some of the toughest competition, as did D’Amigo and Granberg. Jamie Devane, Broll, Marshall and McKegg played against some of the weakest. Brennan was about average to slightly easy. Staubitz, Ross, Ryan and Smithson all had the weakest teammates, which would hurt their point production. Carrick, Abbott and Granberg were in cushy positions with strong teammates. Toughest deployment goes to MacWilliam.

I find it interesting that MacWilliam played by far the toughest minutes on the team. Understandably, he struggled quite a bit keeping the puck out of the defensive end. McKegg and Leivo didn’t do as well as you might expect considering their low quality of competition and high quality of teammates.

In conclusion, the Marlies played much better than expected this year, but given their overall weak possession game and high PDO, they could regress heavily next year. Still, the team has some nice pieces, especially at forward. I would expect a lot of players to make the jump to the NHL next season, specifically their two best players in Brennan and MacIntyre. The jury is still very much out on whether or not Sparks and Gibson can handle the load next year, but there are encouraging signs.

Again, thank you to Josh Weissbock for contributing the data and explaining what it all means. Be sure to check out his work at NHLN and CA and follow him on Twitter.