Using Arbitrage for Perry/Gallagher, Drafting James Neal, and Joe Thornton’s Decline


An important part of drafting in any fantasy sport is not paying a premium for an asset. Drafting at a likely peak is the first step to low/no profitability. What can compound this problem is that paying a premium for an asset can also forego paying a discount for a similar asset.

Originating in the finance world, arbitrage in fantasy, for our purposes, refers to drafting a player at a later round that should provide similar production to a player early on. This is why drafting someone like Duncan Keith last year in the fifth or sixth round made little sense when a guy like John Carlson could have been had in the seventh or eighth, and John Klingberg in the ninth or tenth. Sure, the latter two names have less of a track record, and carried some sort of inherent risk because of that, but with their situations – top-end offensive defencemen playing on likely top-end offensive teams – there was little upside to drafting Keith, and much more in drafting Carlson or Klingberg.

As mentioned, sometimes this requires a leap of faith because the discounted player is discounted for a reason – lack of a track record, uncertainty about health, a poor previous season – but that’s where research and projection play significant roles. And if fantasy owners aren’t projecting what a player can/should be, rather than what they were, they’re playing a losing game.

With that in mind, here is one case of arbitrage that I think will work out very well this year.

Player to avoid: Corey Perry

Over the last few seasons, Perry has seen a decline in point/game rates (1.01 to 0.82 to 0.76), and a similar decline in shot/game rates (3.46 in 2013-2014 to 2.62 last season). It would be fair to say that this is a result of decline in ice time, and it has in part, but he’s also declined in per-minute rates at five-on-five as well. Perry is not a player still in the midst of his prime, either, as he turned 31 back in May.

One wrinkle here is Randy Carlyle being introduced as head coach of the Ducks. How does he manage the ice time of Perry? One area that would boost his totals is an additional couple of minutes of ice time per game like he had a couple of years ago when he had 82 points in 81 games. The problem therein lies with predicting what a coach who hasn’t been in the league for a year and a half will do. If anyone has some insight to provide beyond Carlyle’s tendencies at the end of his tenure with the Leafs, I’m all ears.

A final point here is that Perry isn’t really the bruiser that he had been in the past; from 2007-2012, Perry topped 100 penalty minutes in every season. In the last three 82-game seasons, though, Perry hasn’t topped 70 penalty minutes (albeit he may have in 2014-2015 had he not been injured, but we’ll never know). So while Perry may be good for 60-70 penalty minutes, and that is helpful, it’s not like that PIM total is irreplaceable. The fact is, unless Perry is a 40-goal, 80-point guy, the peripherals may not be enough for him to be worthy of being drafted in the first three rounds, which is where he is likely to go.

Player to draft: Brendan Gallagher

Out of 168 forwards since the start of the lockout-shortened season with at least 3000 minutes played at five-on-five, Gallagher is tied for 33rd in points per 60 minutes at five-on-five. From Hockey Analysis, this is the list of players ranked below Gallagher, and inside the top-50, in points per 60 minutes. Count how many players will probably be drafted ahead of him this year:

In fact, over his three 82-game seasons, Gallagher has increased his goal, assist, point, and shot per game rates in each year. Going into his Age 24 season, he is also coming into his prime. These are all good signs.

Had Gallagher not been injured last year, he may have cracked 30 goals and 60 points, as he managed 19 and 40 in just 53 games. For our purposes this year in fantasy hockey, though, that may be a good thing because his price should come at a discount.

Despite not having eye-popping raw totals, Gallagher’s five-on-five numbers are very solid. Now that he should be on a true top line with Max Pacioretty and Alex Galchenyuk, there’s a real possibility that, barring injury, this is his breakout season. In a full year on the top line, with true top line minutes (at least 18 a game), Gallagher is very capable of getting to 30 goals and 60 points with over three shots a game, and at least 40 penalty minutes. With Carey Price healthy again, that plus/minus should be solid again as well. My one concern would be power play points, as if the Habs run Markov/Weber as the defence pairing, it seems possible that Alex Radulov eventually supplants Gallagher on the top unit. I wouldn’t do it, and don’t think it’s likely, but it’s possible. All the same, 15 power play points is well within reach.

In terms of pure upside, Perry still has the edge here. Even in a great year, seeing Gallagher top 40 goals and 75 points is unrealistic, but no one would be really surprised if Perry managed that. All the same, the decline in Perry’s ice time, combined with being split from Ryan Getzlaf, has me nervous. Maybe Carlyle goes back to the duo and their 20 minutes of ice time a game, there’s just no way to know. Given that Gallagher could come at a discount of six or seven rounds, though, I’d rather bet on Gallagher out-performing his draft position than Perry’s, and in the end, their production may not be too dissimilar.


Thirty goals? Nearly 60 points? An exquisite plus/minus? Over three shots a game? James Neal is back, baby!

Let’s hold on to that thought for a second.

Not for nothing, but coming off a conversation about why drafting Brendan Gallagher in rounds 9-10 rather than Corey Perry in rounds 2-3 is probably a smart move, I feel the same way about Gallagher and Neal.

Remember that this is the same James Neal that wasn’t healthy for three straight seasons. It was the first time in his career he played all 82 games, and just the second time he had played at least 80. Going back three years from 2012-2015, Neal played in just over 78-percent of his games (78.3-percent, to be exact). That is a pace that would see a player play 64 out of every 82 games. Fantasy owners need to decide is they can rely on Neal for 80-plus games. I have my doubts.

Also, the goaltending behind Neal was sublime. Over the course of the season, Nashville goaltending saved 94.7-percent of five-on-five shots (per Hockey Analysis). As a team, Nashville was a .921 overall. So either Neal, and line mates, are among the best defensive quintet in NHL history, or he got pretty lucky with his on-ice save percentage. Bet on the latter.  

Finally, over his 149 games with Nashville, Neal has just 18, count ‘em, 18, power play points to his name. It seems impossible, but Sam Reinhart had more power play goals last year (8), than Neal has had in his tenure with the Predators (7). Bad luck? Maybe. But the Predators’ power play runs through the points, and with Filip Forsberg emerging as an elite offensive talent, I’m not sure how many power play goals Neal could add to his total from last year (4) in a full healthy season (which, as mentioned, is another question mark).

I think Neal will be over-drafted this year. There is no question that he was a top-tier option last year, fulfilling what he could be in fantasy. There is also no chance he repeats close to his plus/minus, will probably not post top-end power play numbers, and the health questions linger. We will have to see where his ADP sits, but I can’t imagine he slides out of the top-50 this year, and I think there’s too much inherent risk. 


One question I can’t help but ponder over the off-season is as to when, exactly, the decline of Joe Thornton will begin. I say that because his late career production is off-the-charts good (from Hockey Reference's Play Index):

That is a list of names since 1990 with more points than Thornton in their Age 34-36 seasons, or Thornton’s last three years. That’s four Hall of Fame players, plus two more that have a case to be made for the HoF. By any definition, Thornton’s production these last three seasons stack up against some of the best to play the game over the last 25 years. In other words, elite.

So when does the decline come?

Well, Thornton’s minutes per game hit their lowest point since he was a teenager, so that is one factor. This led to his lowest per game shot rate since he was 18 years old at just 1.48 shots per game. Even without a decline, an “unlucky” season could see him not even crack double-digits in the goal column. We saw similar sharp, one season declines from the likes of Marian Hossa and Ryan Getzlaf just this past season. Thornton could not decline in actual on-ice talent one bit, but a couple inches here and there would leave him with 8-10 goals. It’s not hyperbole, it’s just a reality of his shooting rate.

Thornton also saw two peripheral spikes last year, one putting up over 50 penalty minutes for the first time in five 82-game seasons, the other was posting a plus-25, the highest for him since 2005-2006. Can he post that PIM total, and can San Jose goaltenders save nearly 94-percent of the shots with Thornton on the ice again, like last year, when Thornton had a .939 save percentage behind him at five-on-five? I wouldn’t bet on either.

Finally, the Sharks shot 10.7-percent with Thornton on the ice at five-on-five last year, which ranked third in the NHL. Good team finds the net often with an elite centre on the ice, right? Well considering the team shot 8.2-percent with him on the ice the five years prior, and he had never exceeded 9.2-percent in any of those years, repeating that 10.7-percent number is unlikely.

There are a lot of reasons to draft Thornton, but drafting a 37-year old with unreliable peripheral stats, and outlier underlying percentages, seems very risky. Those who drafted him last year turned a tidy profit. Those who draft him this year, I think, will be disappointed.

*Stats from Hockey Reference and Hockey Analysis