With the regular season now over, it’s time for me to face the music on the 15 Fearless Forecasts I made right before the 2017-18 season kicked off. Rather than just listing the results, I’ll talk about lessons that can be learned from my correct and incorrect predictions and the logic I used in making them. That way all of us – myself included – can use the results to make better fantasy decisions going forward.
For each of the 15 forecasts, I’ll assign a grade of:
- HIT – the forecast turned out correct
- CLOSE CALL – I was almost correct and had sound logic
- MISS – I was incorrect but some aspects of my logic may have held true
- AIR BALL – I was wrong in my prediction plus with most – if not all – of my accompanying logic
Without further ado, let’s dive in!
1) At least 15 defensemen will score 50+ points
Result = HIT (19 tallied at least 50 points)
Seeing that 19 had 50+ points might make this seem like not such a fearless forecast; but remember, over the last six years there’d only been 9, 5, 9, 11, 12, and 9 who tallied 50+ points, so a jump to 15 was indeed a bold prediction. Plus, any “at least“ scoring forecast is always vulnerable to injuries.
So why the increase, and can we count on it being the new normal? The number of rearguards who averaged 3:00+ of PP Time didn’t rise (19 this season and last), but fewer averaged 2:30-2:59 (14, versus 17 in 2016-17). That suggests I was correct about more teams embracing the idea of using just 1D on PP1, and, as a result, that lone D factoring into scoring slightly more. But I think the key in this prediction was the peak age factor, as when you look at those who scored 50+ this season, 15 of the 19 were age 23 to 28, which is in the sweet spot in terms of production.
I was also correct about the number of 40-49 point d-men dropping (11, versus 14 in 2016-17 and none had 46-49 points) due to the same factors. I’d look for a similar pattern next season as well, so keep that in mind when making your keeper selections and assigning rankings for next year’s drafts.
2) No Sharks forward will score 60+ points
Not even a major injury to Joe Thornton and weak early season for Pavelski (24 points in his first 38 games) was enough to make this prediction come true, as Pavelski surged in the second half and Couture was a steady eddie all season, which was enough for both to finish above 60 points. Even still, Pavelski’s scoring rate dropped for the second straight campaign and this is Couture’s third straight season in which he’s scored at a lower rate than his prior four, so I was at least onto something.
Also, despite this being a miss, there were lessons to be learned, namely once again that age matters in fantasy (Pavelski and Couture are both post-peak), as does the loss of a long-tenured member of a team (in this case, Patrick Marleau). But my thinking that the departure of Marleau would adversely affect San Jose’s PP turned out to be incorrect, as its PP conversion rate improved year-to-year.
Result = AIR BALL (Radulov more than tripled Sharp’s meager output – 72 points to only 21)
All I can say is when you aim for the fences, sometimes you not only swing and miss but also split your pants in the process. I was way off both in thinking Sharp could be resurrected in Chicago and Radulov would be a flop in Dallas. Instead, Sharp looked completely washed up, and Radulov – fat wallet notwithstanding – played his heart out and meshed well with Tyler Seguin and Jamie Benn.
The lessons to be learned are that although some players can excel despite being older than 30, for the most part once an over 30 player starts to see his production wane it’s very difficult for it to bounce back. Exceptions can include a player who is plugged into a new, much different and reviving situation (like what’s occurred with Eric Staal in Minnesota) or one who is a truly generational talent (ala Jaromir Jagr for Florida in 2015-16).
With Radulov, chances are there won’t be too many directly analogous situations in the future, as normally players don’t head to the KHL for nearly a decade before returning to the NHL. But I suppose what I learned with him is you shouldn’t automatically assume players – even ones with enigmatic histories – are going to phone it in once they ink a lucrative deal.
There’s also a lesson here about coaches, as Ken Hitchcock had long been someone whose skaters underperformed due to his defense-first style. In fact, as I noted he had a previously unwavering track record of top players doing worse during his first year at the helm versus their previous season for the same team. However, I failed to give proper weight to the fact that in his later seasons in St. Louis he’d shown signs of refining his coaching style to embrace offense, plus he chose to come out of retirement to get back into coaching. So although a coach’s past tendencies merit consideration, don’t assume the worst, especially if there are signs to suggest the worst is no longer a likely outcome.
4) More than half of all teams will have a goalie appear in 60+ games
Result = MISS (only 12 teams had a goalie play in 60+ games)
When the dust settled, the number was essentially in line with previous seasons (the average had been 11.5 in the past six campaigns, with a high of 13 and a low of eight). What I overlooked is goalie injuries are real and strike starters every season. Plus, despite there being no Olympics and each team getting a week off, squads these days are simply choosing not to overwork their netminders.
I also fell into the trap of looking at non-predictive stats in a predictive way, most notably that in each of the past two seasons nine of the top ten netminders in terms of number of appearances were on squads that made it to the playoffs. Although a fact that wasn’t something which influenced whether a team would choose to play one of its goalies in 60+ games. This is in contrast with my correct prediction of 15 or more d-men scoring 50+ points, since that was based on observable trends and peak ages. Be sure not to fall into the same trap as me – forecasts need to be tied to predictive stats or trends, rather than hunches or – even if factual – non-predictive data.
The small saving grace in terms of this forecast was my added note that poolies shouldn’t expect to see more than the usual 5-8 goalies appear in 65+ games, as indeed the number was five.
Result = HIT, but with an asterisk (Ehlers had more goals (29) than Schultz had points (27); however, had Schultz played in all 82 games and scored at the same pace his point total would’ve been 35)
I give myself credit for nailing that Schultz’s 2016-17 scoring pace would crater; however, I fell into the trap of putting Ehlers on an unrealistic fantasy pedestal. I figured if Ehlers performed as well as he did last season without any PP1 time, surely the team would opt to put him on PP1 and, between that and his natural career progression, the goals floodgate would open. But his PP1 time never materialized; and although he upped his goal total, it was only from 25 to 29.
So why is it that Schultz failed this season while Radulov succeeded, after both signed contracts in the offseason? For one, Radulov changed teams, which provides added motivation. Also, the concerning data I cited regarding Schultz in making this forecast (poor finish to last season; too much of his 2016-17 production tied to PPPts) was predictive, versus my “data” about Radulov being speculative. And although Kris Letang didn’t stand in the way of Schultz’s success as much as I figured, the fact that Letang was healthy for the entire season (and I’d noted that Letang had never failed to play 70+ games in two straight full seasons) did cast a shadow which hadn’t been there for much of 2016-17.
The major takeaway is that for every presumed breakout, only a few materialize. Also, no matter how talented a player – like Ehlers – might be on paper or in the minds of poolies, opportunity will have the most direct influence on whether (and, if so, the extent to which) a player actually breaks out.
6) Antti Raanta will finish in the top five in SV% among 40+ appearance netminders
Result = HIT (he finished first)
I’m especially proud of this forecast, not only because Raanta ended up being tops in the entire NHL among 40+ appearance goaltenders, but also because far more pundits jumped on the preseason bandwagon for Scott Darling, the other “graduating” back-up. Yet when the dust settled Darling was a major disappointment and Raanta was the best goalie in hockey once the calendar turned to 2018.
How did things go right for Raanta, but not Darling? For one, Raanta’s past data was culled from 94 career appearances (78 starts) spread over four seasons, compared to Darling’s 75 career games (64 started) in three seasons. Also, all of Darling’s games had been for the then powerhouse Blackhawks, who could make a netminder look better than he really is. Raanta, on the other hand, had played the past two seasons for the Rangers, who, although a playoff team, didn’t have as vaunted of a blueline as the Hawks did in the previous three seasons. Lastly, Darling walked into a situation where although it was presumed he’d be the #1 guy, Cam Ward was still lurking in the wings, whereas the Coyotes showed their commitment to Raanta by trading away Mike Smith and not re-signing Chad Johnson.
Also, this is another example of the stats I used to help make this forecast (i.e., only seven netminders since 2013-14 had both a better GAA and SV% than Raanta, and five had finished in the top five in SV% at least once during that span; although Arizona gives up lots of shots, Raanta had been among the best in EV SV%) being predictive stats. It also didn’t hurt that going into the season Raanta was set to be a UFA this summer; and even though many a player in that same situation has laid an egg, quite a few have found that to be added motivation to play even better than they might normally.
Result = HIT (Pietrangelo tallied 54 points in 78 games, i.e., 17 more than Werenski’s 37 in 77 games)
This is another one of my prouder hits, what with Pietrangelo having outpointed Werenski by only one in 2016-17 and many going into the season touting Werenski as a young star on the rise. I saw things differently, and it turns out I was right both in terms of my prediction and my logic in making it.
What it boiled down to is Pietrangelo having more value than was evident from his 2016-17 season-long scoring, as he posted 18 points in 20 games once Mike Yeo stepped in as coach. Plus, although I noted above when referring to Ken Hitchcock that some coaches can and do indeed change their ways, Yeo’s MO of leaning heavily on one d-man is among the most well-established patterns in hockey. Between those two factors I saw bigger and better things from Pietrangelo, and sure enough he delivered despite a second and third quarter where he produced only 18 points in 35 games.
Meanwhile, working against Werenski was the cloud of the dreaded defenseman sophomore slump. But even beyond that there was the objective reality that Seth Jones was quietly emerging as “the guy” in Columbus in the last quarter of 2016-17. So just as Pietrangelo’s late season trends were key, so too were those happening in Columbus.
The major takeaway here is to be careful not to value players based solely on season-long totals. Also, don’t be sucked into thinking every upstart player is on a fast track to becoming a phenom, especially if they have other at least equally talented players breathing down their necks.
Result = MISS (they raised their combined total from 177 in 2016-17 to 178, although Draisaitl did see a decline in production)
My reasoning in making this forecast was Edmonton needing to focus on being a more all-around team in order to mount a serious challenge for the Cup. But a funny thing happened – the Oilers didn’t come close to qualifying for the playoffs. So once the season was a lost cause, McDavid, who actually stood at a not so jaw dropping 69 points in 58 games, went into beast mode over the last quarter of the season (39 points in his final 24 games).
But in truth, my basing the forecast on the Oilers being a lock for the playoffs is less of an excuse for me missing this prediction and more of a flaw in the prediction itself, since yet again it was an example of a forecast being made on an assumption rather than on sound, predictive logic. There was some solace to be found in that Draisaitl’s production fell, in part because the Oilers did indeed separate the two more so than last season. Even still, Draisaitl showed he can produce in his own right, and had the team’s PP clicked at anything close to a normal rate this would’ve been a huge miss.
Thus, again the takeaway is to base forecasts on predictive factors, not assumptions. And especially not multilayered and convoluted assumptions, as doing so amounts fantasy FPS, with the FPS standing for “fancy play syndrome.” Although in competitive leagues you do have to outthink your opponents in order to prevail, what that means is thinking smarter, not necessarily thinking fancier or using layer upon layer of non-predictive logic to reach a flawed conclusion.
Result = AIR BALL (Vas started 64 games, tied for fourth most among all NHL netminders)
Technically Budaj got hurt and missed three months; but even a healthy Budaj would’ve done nothing to derail Vasilevskiy’s superb season. I still think my reasoning wasn’t entirely flawed – I just mistook the Bolts bringing in Budaj as a lack of confidence in Vas, when in truth it was as an insurance policy. What this also underscored (as did St. Louis continuing to turn to Jake Allen, even amid a superb season from Cater Hutton) is golden boy goalies will get every chance to succeed before being pushed out of the picture even temporarily. Or to put it another way, there are just some situations where a back-up will always stay in that role no matter how well he plays and/or how poorly the starter fares.
What factors can exist which suggest a starter might be vulnerable to losing his gig as the #1 guy, other than via injury? Contract status, including seasons he’s signed for and money he earns, is probably the #1 factor. There’s also the past glory factor, which some could argue led to Cam Ward staying the starter or at least remaining in the starter picture for as long as he has in Carolina, despite his Stanley Cup winning season now being more than ten seasons ago. But there’s also the flip side to that, namely the “young phenom” factor, where a young anointed starter stays in that role even if/as he falters. Another factor is playoff experience, since if a back-up has never even tasted playoff action or not met with success, it’ll be more difficult for him to supplant a starter despite stellar play. The converse also holds true, where a starter who has a superb playoff record is more likely to keep his gig even if he’s being outplayed by a back-up in the regular season, to help him work out any kinks before the second season.
10) The Winnipeg Jets will score the most goals of any NHL team
Result = CLOSE CALL (their goals improved by more than 10% from 246 to 273, finishing second in the league, up from seventh last season but nevertheless behind league leading Tampa Bay)
My logic was sound, but in the end Tampa gelled even more. What are the takeaways? If a team – like the Jets – has a mix of veterans who always produce, players already in or just now entering their peak, plus young talent who could break out, chances are they’ll be a success offensively, especially if – as I noted – their previous season’s PP clicked at what looked to be a lower rate that it should’ve.
I was also correct they’d pass all six teams who finished ahead of them in goals last season. What that means it it’s equally important to examine which teams might see their offense falter and why (whether due to too many older player, having overachieved the prior season, trying to become a more balanced team in order to be built for the playoffs, etc.).
11) A Carolina forward will score 75+ points
Result = MISS (Sebastian Aho scored at a 68 point pace – highest among any Hurricane player)
I think I bought into offseason hype, as nearly all pundits had Carolina as a breakout team in the East. In doing so, I failed to account for the fact that no Hurricane forward last season topped even 2:16 per game on the PP or 17:05 non-shorthanded minutes under the same coach who’d be at the helm for 2017-18. Simply put, those are not normally the kind of conditions which will lead to a player topping 75 points, no matter how skilled he might be.
So this wasn’t a case of using non-predictive stats in a predictive way, or fantasy FPS. It simply boiled down to ignoring reality and deployment, and the limitations those create on scoring for even the most already talented or up and coming hockey players. Whether in making predictions like these or simply valuing players in general, you have to look under the hood at teams/coaches to see if their ice time philosophy is one that will make it plausible for 75+ point scoring to occur.
12) Milan Lucic will score fewer than 40 points
Result = HIT (he finished with a mere 34 points)
While it wasn’t exactly rocket science to predict that Lucic might not shine for the Oilers this season, to have predicted a drop below 40 points was indeed bold, as not only did he finish with 50 last season he’d only once been below the 50 point mark since 2008-09 and had topped 55 in three of the previous four full campaigns. He also had a big fat contract that seemingly guaranteed he’d get prime minutes with top tier linemates.
But the key in making my prediction is sometimes those things simply aren’t enough to ensure a player can produce. Beyond that, some rough-and-tumble, fire-in-the-belly players seem more vulnerable to slowing down even in their 20s, especially after signing big contracts. That’s not always the case, but it’s something that occurs often enough to warrant concern and attention when drafting and valuing those types of players.
13) Alexander Wennberg will finish within the top three among NHL forwards in assists
Result = AIR BALL (Wennberg’s production dropped like a stone, and he had fewer assists on a per game basis than he did not only last year but two seasons ago)
My mistake here was in looking at stats in a vacuum and failing to fully consider both circumstances and players, plus putting too much weight on the “magical fourth year” factor. Wennberg did have strong numbers trending in a direction that suggested he would do even better for 2017-18; however, much of those numbers were due to him being a #1 center more by default than by design, and also by virtue of him having been deployed with less talented players, ala Brandon Saad and Nick Foligno. Essentially, Wennberg had been functioning as a placeholding top line pivot and, once displaced from that role by Pierre-Luc Dubois, saw his stats come back to earth with a crashing thud.
So how can one avoid making the mistake(s) I did? Focus on depth charts and see if a player’s success is occurring because of a situation which is on tap to change, rather than due to his own pure talent.
14) No player from the Metropolitan Division will have more goals than Anders Lee
Result = MISS (Lee did finish with the third most goals in the Metropolitan Division, but a total of nine behind Alex Ovechkin)
Although Lee finished well behind Ovi, I still feel pretty good about this forecast. After all, Lee improved from fourth to third in the division, and his goals per game rate increased from an already high 0.42 to an even greater 0.487. Not bad for someone many had pegged for regression this season due to his lack of proven track record and 17.8% personal shooting percentage in 2016-17.
My logic was sound too, in surmising that Lee would see more PP Time, plus I correctly noted that his chemistry with John Tavares would lead to those two staying together. Those, plus Tavares’ motivation to score both in general and due to his impending UFA payday, would only help Lee’s role as a sniper.
Result = HIT (Kane finished with 54 points in 78 games, versus 48 points in 82 games for Guentzel and 30 points in 79 games for Sheary)
I nailed this one both in terms of result and reasoning. Say what you want about Kane – and there’s plenty you could say – but when push came to shove in terms of playing for his NHL future he appeared in a career high 77 games and posted his second highest point total. Coincidence? Nope. Let this be another lesson to never underestimate how well one can suddenly play when a paycheck is at stake.
As for Guentzel and Sheary, my key logic was that neither would find room on PP1, as Patric Hornqvist might not be as “sexy” of a fantasy name these days, but he’s an all-important “mule” for a Pens top unit, leaving the more diminutive Sheary and Guentzel with PP2 scraps. Couple that with the reality of many productive rookies – including on the Pens – not living up to the hype when it came to their sophomore season, and it shows that poolies should not ignore the dynamics of the squad a skater plays for and longstanding team dynamics that have met with success and thus unlikely to be rethought.
* * *
The final tally was six Hits, one Close Call, and three Air Balls, which, considering these were “fearless” forecasts, is a pretty solid result and an improvement over last season. How did you fare in your voting? The five “Hits” finished second, third, fourth, tied for fifth, also tied for fifth, and 12th in your votes, with the Close Call finishing first. So hopefully you followed your instincts, since they were pretty solid!
I hope you enjoyed reading the original forecasts and this look back in review, and as had as much fun as I did along the way. But let’s not lose sight of key fantasy lessons we learned from (or were underscored by) these forecasts. With a stick tap to Tom Collins and his weekly column, here’s a rundown of the top 10 (in no particular order) takeaways:
1) When making forecasts, you’ll tend to succeed more (or fail less) if you base them on predictive data as opposed to speculative or coincidental data.
2) Don’t underestimate the motivation for players to suddenly step up when they need to earn a UFA contract, and by the same token for players without a solid track record to come back to earth with a crashing thud once they get paid.
3) Age matters in fantasy hockey, and peak ages for forwards, defensemen, and goalies are important. But there can and will be variations, with some staying productive for atypically longer but others (especially rough and tumble players) not only entering their post peak earlier but seeing their scoring drop more precipitously.
4) Beware of players who could be doing situationally – and, thus, temporarily – well, rather than as a result of pure talent.
5) Tandem chemistry should always be factored into player value, especially the lesser value player.
6) Winning in fantasy is indeed based on using sound logic and outthinking your opponent; but beware of falling into the trap of “fancy play syndrome,” where you get too clever for your own good and end up making unsound decisions.
7) Late-season trends should be given added weight over season-long trends provided that the late season trends better represent what will likely be the “new normal” for the affected player(s).
8) Buying into hype is dangerous, since it often means paying an inflated price and then having to hope the hype is real just to “break even” in terms of value given what you paid.
9) Making predictions for goalies is extra tough because of the team influence factor; about the only sure bets are goalie injuries will always happen and “golden boys” will get a much longer leash.
10) When teams are jumping aboard trends (like 1D on a PP1), factor that into your ratings but don’t go too overboard since you never know when winds might change or shift direction.