Will giving Troy Brouwer more offensive zone time work?

Updated: January 11, 2018 at 2:13 am by christian tiberi

Of all forwards who played 500 minutes of 5v5 this past year, Troy Brouwer received the 66th hardest zone starts. On the Blues, he had the sixth roughest time by raw numbers (26.71 OZS%) and relative numbers (-11.73% ZSRrel% [zone start ratio: OZS/(OZS+DZS)]). If he played on the Flames last season, he would have the fifth worst starts by raw and fourth by relative numbers.

Brouwer was placed in situations where he would get his head kicked in. As a result, he got his head kicked in (49.56 CF%, -3.32 CFrel%). Among the fanbase, there are those who would say that Brouwer is doomed to fail due to his history of poor possession numbers (among other things). Others would say that Brouwer’s consistency with regards to scoring prove that he is an offensive stalwart whose poor possession stats have been the result of his zone starts.

This season, Brouwer is going to test that hypothesis for us. He is (probably) going to get first line duties with Sean Monahan (36.71 OZS%, +8.43 ZSRrel%) and Johnny Gaudreau (37.24 OZS%, +9.94 ZSRrel), and hopefully prove to be a solid offensive addition. Brouwer hasn’t seen starts like that since his final season in Chicago. 

Using stats, we’re going to try and predict how it’s going to go.

All data in this article is from corsica.hockey and Puckalytics.

The simple solutions

The quick way of solving this debate would be to adjust for zone starts. In that case, Brouwer goes from -3.32 CFrel% to -2.35 CFrel%. All things being equal, he is still a possession drag. Thank you for reading my hockey article.

However, adjusting for zone starts could perhaps slight Brouwer. Considering his history of being in the trenches, his stats have already taken enough hits that, even in an attempt to make all things fair, Brouwer still doesn’t get the benefit of the doubt.

The second thing to consider are his WOWY stats. To say Brouwer is a passenger is to put it lightly.

If he isn’t dragging players down when together, he is noticeably bad away from linemates. While he will be playing with Calgary’s best, it’s still not great if he has to be held up by Johnny Gaudreau to look serviceable.

To argue the case for Brouwer, we can look at what he did historically. The RW has only had two seasons when he has had a positive corsi rel, both with Chicago. Those are also the only two seasons when he had relative zone starts like Monahan and Gaudreau (16.93 ZSRrel% and 1.02 CFrel% in 08/09, 11.8% and 3.2% in 10/11). Although he has added years since, perhaps the more important thing is that he stays away from the defensive zone.

Zone starts and trendlines

If we’re assuming that zone starts and corsi go hand-in-hand, we can build a model for Brouwer. We’re taking every game he’s played since the beginning of the 2007-08 season, comparing the zone starts (x-axis) to the CF% (y-axis) and putting in a line of best fit. Here’s the chart:


Pretty simple stuff. I’ve made a few observations after diving into the data.

  1. If you follow the line of best fit and match it up to where Monahan and Gaudreau usually start (together, 37.5%), you could expect Brouwer to have a CF% of 51.95%. That is actually very good, considering that Monahan and Gaudreau only managed a CF% of 50.1% together. Theoretically, Brouwer’s impact should actually boost the two youngsters.

    The problem with that theory is that it doesn’t account for relative stats. Due to Corsica not including ZSR rel in their individual game stats, I couldn’t measure the relationship between ZSRrel% and CFrel% as I originally intended to. The raw numbers work fine as a substitute, but a bit more context is always nice.

  2. The r-squared number, 6.05% for Brouwer, suggests a few things.
    a) The standard interpretation would be that if Brouwer did receive the same zone starts as Monahan and Gaudreau, there would only be a 6.05% chance of him actually achieving that 51.95CF%. Not very good.

    b) Let’s take it one step further. Statistically speaking, that r^2 number can just barely be recognized as a relationship. Starting him in the offensive zone is likely to be beneficial, but that won’t solve his possession woes. If 6.05% of his corsi is explained by zone starts (mathematically spitballing), than it’s more important to pay attention to what makes up the 93.95%.

    If we assume individual input is the biggest chunk of that 93.95%, then Brouwer is in trouble. His most common linemates throughout his career have been Patrick Kane (1,045:59), Jonathan Toews (1,021:37), Evgeny Kuznetsov (601:24), and Alexander Ovechkin (572:28). His linemates are a topic for another day, but for now Brouwer’s poor corsi performances seem to be completely on him.

  3. I did some digging around and chose to see whether or not the r^2 values actually meant anything for other players; i.e, would an elite player have a higher number? Do grinders have lower numbers? I searched for forwards who have played more games than Brouwer since 07-08 (for sample size and practicality purposes) and compared the OZS% vs CF% data.

    Generally speaking, the relationship between the two is very weak. R^2 numbers rarely went up past 8%, with the highest being 12% (belonging to Jeff Carter). 7% was about average for my sample.

    Compare r^2 to players’ careers, and the numbers are more random. Alex Ovechkin barely cracks 5%. Iginla is slightly above Brouwer at 6.15%. Henrik Sedin is at 8.15% while his brother is just above 5%. Generally speaking, OZS% and corsi rarely have anything to do with each other. All players’ trendlines were positive, but as a whole the data suggests that putting players in the offensive zone to boost their possession does not guarantee results.

Putting it all together

Brouwer is the ideal fit with Monahan and Gaudreau. He’s a sharpshooter with size and the tendency to drive the net and get the dirty goals. He could be the final piece that the top line was missing last year.

Ideal isn’t reality. Reality is that Brouwer has never been an effective driver, which is what the Flames really could use. It isn’t enough to just slot in a guy on the first line; there should be someone who will be keeping pace instead of always catching up. The data here reasserts that point. Just because someone starts a lot in the defensive zone and suffers does not mean that they will flourish with offensive zone starts. That’s not a caveat to a rule, the data suggests that that is the rule.

I don’t think Brouwer will be as bad as some people expect him to be. On the first line, he will be an inoffensive choice and will probably put up 40 points again. If the Flames are wise, they can maximize their return by keeping him in the offensive zone. I don’t think it will turn him into a whole new player at age 31, but I believe that keeping him away from the defensive zone is best.