Hello everyone and welcome in for another article, the long-awaited #ZeroG update based on the 2022-2023 season. Before I get into that, I want to take a second to invite you to join the Apples & Ginos community on Discord and the Apples & Ginos Patreon where I would be more than happy to answer any questions you have about #ZeroG or other fantasy hockey strategy topics while we count down the days til the next fantasy season.
What You Need to Know
Before getting too far into the analysis, you should have the baseline to understand the topic at hand. If you haven’t already read my original article on the #ZeroG draft strategy, you should probably take a moment and catch up on that one. While you’re at it, you could take a second moment and read last year’s update, which will bring you fully up to date on what’s become a yearly (and heavily anticipated) article for me here at Apples & Ginos. If you prefer to get your info in audio format, you can always check out the recent podcast we put on this topic as well.
How Did #ZeroG Fare in 2022-23?
The basic analysis I perform each year is to take the top 14 goalies by Yahoo! ADP (average draft position) and find their average season finish rank. Then I take the next 14 goalies (drafted on average 15-28 at the position) and find their average season finish rank. I do this by total fantasy points and average fantasy points per game using the KKUPFL points scoring system. For reference, I’ve included the last three years’ data in this table:
| Goalies by ADP Rank | 2020-21 Average Rank | 2021-22 Average Rank | 2022-23 Average Rank | |||
| Total FPts | FPts/GP | Total FPts | FPts/GP | Total FPts | FPts/GP | |
| 1-14 | 25.6 | 27.2 | 14.9 | 15.1 | 19.6 | 24.5 |
| 15-28 | 22.5 | 19.1 | 24.8 | 23.7 | 27.4 | 26.7 |
As you can see, in all three years the difference between total fantasy points and fantasy points per game (also known as “average fantasy points”) is such that the gap between goaltenders 1-14 and 15-28 by ADP is lessened. In 2022-23, the difference in average fantasy point per game rank from goalies 1-14 and goalies 15-28 was nearly negligible. So while you were gaining some volume from the goalies drafted earlier, you were not gaining much of anything on a per-game basis.
Of some interest as well is that we have had some very significant differences year-to-year in terms of these ranks. The COVID-shortened 2020-21 season may be something of an outlier, but it’s hard to say that even if that season had been 82 games that we would have seen such a regression over the last part of the season that it would have changed those numbers to the point that they were more favourable to the 1-14 goalies than this past season was. All told I think the data accumulated shows that (a) some years are significantly kinder to the expected top goalies than others and (b) we are still not as good as we think we are at predicting goalie performance and it’s still an unreliable bet to make early in your fantasy drafts (ask anyone who drafted Jacob Markstrom last year).
Of course, these numbers still don’t take into account the opportunity cost of drafting early goaltenders versus the very real potential of unearthing a Linus Ullmark or Stuart Skinner or Filip Gustavsson later on. This is the core component of the #ZeroG philosophy; spending on goalies in your draft is a bet on a historically unreliable asset with huge opportunity cost when you have a very strong possibility of equaling or beating that production with a late-round selection or waiver wire pickup later on.
#ZeroG FAQ
When preparing for the #ZeroG in 2023 podcast, my co-host Blake and I put together a list of #ZeroG FAQ to discuss. I figured I may as well tackle some of them here as well, since there are some more granular topics that we can dig down to that will actually be of considerable use to #ZeroG truthers when attacking their drafts come September.
- How Does #ZeroG Change in Categories vs. Points Leagues?
- My take on cats leagues versus points leagues is that oftentimes cats leagues are much more weighted towards “rate stats”. Rate stats are statistics like goals against average (GAA) or save percentage (SV%) which are much more dependent on a goaltender playing well or playing for a good team than they are on sheer opportunity. Points leagues are inherently weighted moreso towards “volume stats”, which are counting stats like wins and saves. These categories are obviously much more tied to how many starts a goaltender is getting rather than exclusively how well they personally are playing.
- So the strategy in these instances then becomes weighted moreso towards goaltenders we expect to have good team situations and/or are playing well at the current moment. I’ve already stated in many places that I’m not in the business of predicting goalie performance, but team performance is much more reliable and teams like Boston and Colorado have made a business out of maximizing goalie performances with their team’s defensive ability. In both cases you did not have to overpay in your drafts to acquire their goaltenders; both Linus Ullmark and Jeremy Swayman were very available late in drafts this past season and Alexandar Georgiev was #ZeroG pick I touted on several occasions that turned into a massive hit due to extreme volume married with top-end performance.
- Of course, if you miss in the draft you may be on the lookout in-season for waiver wire pickups, but players like Pyotr Kochetkov and Filip Gustavsson were world-beaters for month-long stretches at a minimum. You’ll need to be vigilant to catch these players in the wild, but you can narrow your search in-season by looking at teams with strong defensive structures (like Carolina and Minnesota above) and jumping on goaltenders that see an increase in opportunity. I think this formula for finding waiver wire goaltenders is much simpler and easier to execute than many make it out to be – it just takes some continuous attention to goaltending situations across the league throughout the season.
- How Does #ZeroG Change in Different Points Scoring Setups?
- As I mentioned above, most points systems are inherently volume-centric and the #ZeroG focus is on finding cheap volume in the draft and on the waiver wire. However I have seen some points setups that have an outlandish emphasis on wins rather than saves, for example, and this is a clear deviation from more standard points setups. Andrei Vasilevskiy received 68 points for wins and 276.3 points for saves minus goals against last year in KKUPFL scoring, or roughly a 4-to-1 ratio of saves/GA to wins. If your league has a points setup where this ration is something like 2-to-1, then wins have a massive import and you should be much more focused on volume goaltenders on very good teams. Georgiev and Vitek Vanecek would have both been very good options in a points setup like this that were projected volume starters on their respective squads.
- What Round Is It OK to Draft a Goalie?
- One of the most common questions I get is “When can I finally draft my first goalie?” or “What round is the first round you’d finally consider taking a goalie?”. The answer is never so black-and-white that I can give a specific round number and say “you have the green light in Round 10 or later”. There are league settings considerations and in-draft opportunity cost considerations to be evaluated that simply cannot be generalized into a hard and fast rule. I participated in one league last season in which nearly everyone faded goaltenders to an extreme degree, resulting in Connor Hellebuyck being available in the 9th round. At that point I took Hellebuyck because he represented the best value over replacement selection in my mind. So focus less on which round and more on avoiding spending a lot of meaningful draft capital on goaltenders. “a lot” is intentionally vague 🙂
- In What Situation Would Taking a Goalie Earlier Make Sense?
- There aren’t a lot of situations in which it makes sense, quite honestly. Perhaps there are some leagues out there with a strong points weight on volume stats for goaltenders and an overweighting of goalie points relative to skater points that make grabbing a clear workhorse a near necessity, but I would simply choose not to play in those leagues. In my case from the previous question, I would consider my selection of Hellebuyck in the 9th round to be “early” for me, but it was still based on opportunity cost and my relative lack of conviction about any of the skaters available to be selected at that point in the draft.
- How Has #ZeroG Changed Since Its Inception?
- Quite honestly, the main thesis of #ZeroG has not changed much since the original article in 2021. Goalies are unreliable and replaceable, therefore don’t use a lot of precious draft capital on them. Some goaltenders seem to have the goods to buck the overall trend of unreliability like a Vasilevskiy or Hellebuyck, but there is always the possibility that one of those goes full Bobrovsky and self-detonates (before miraculously returning to all-world form in the playoffs?!?!) in any given season. On the whole fantasy managers are paying much more than they can reasonably expect to return from their goaltenders in fantasy drafts, and I anticipate that this will continue to be the single greatest edge in fantasy drafts until the wider fantasy hockey community fully embraces #ZeroG.
- What Statistics Can We Use to Determine Goalie Targets?
- I want to briefly address a statistic that I have seen some people hold up as the single best statistic for evaluating a goaltender’s ability, and that is 5v5 delta Fenwick save percentage (5v5 dFSv% for short). From BlueshirtsBreakaway.com we have this definition for dFSv%: “The difference between a goalie’s (or team’s) actual Fenwick save percentage and his expected Fenwick save percentage. The formula is dFSv% = FSv% – xFSv%.” Fenwick attempts are all unblocked shot attempts (whereas Corsi includes blocked shot attempts), and so Fenwick save percentage is the save percentage on all unblocked shot attempts (rather than traditional save percentage which is based on actual shots on goal). Expected Fenwick save percentage is based on an expected goals model that takes into account where all the shot attempts are located on the ice and assigns each one a probability of becoming a goal based on where the shot attempt originated from. With this model, the expected FSv% is the save percentage we would expect a statistically average goaltender to have given the expected goals from the shot attempts that particular goalie faced. Clear as mud?
- With definitions out of the way, I do believe that there is some value in ascribing more credence to 5v5 goaltender performance than penalty kill performance. On the penalty kill the goaltender is much more at the mercy of the performance of the penalty killers in front of him and the effectiveness of the PK strategy as a whole. Many terrific goaltenders would post terrible PK save percentages relative to league average if the team in front of them was simply terrible at defending penalties for any reason. However, I have issue with ascribing predictive power to 5v5 dFSv% as a statistic when I view it as simply a descriptive statistic.
- All statistics are inherently descriptive in nature: they quantify events that have happened in the past. But whereas for skaters we know that statistics like shots/60 and individual scoring chances for/60 have predictive power to tell us which players are likely to score more goals in the future, we cannot have that certainty with 5v5 dFSv% or any other goaltender metric I’ve investigated. Simply put, last season’s 5v5 dFSv% mark does not correlate with the following season’s fantasy points or fantasy points per game. There is a lot of noise in this correlation due to player movement, coaching changes, etc. but unless you’re willing to go through each individual situation and ascribe some sort of numeric value to offseason changes, I don’t see how you can confidently use 5v5 dFSv% or any other mark to predict goaltender performance for fantasy.
- I am much more interested in team-level statistics. Teams like the aforementioned Boston and Colorado or even Toronto and Carolina have demonstrated the ability to produce strong goaltending performances from multiple different goaltenders throughout the last number of years. Therefore, teams with both strong Corsi Against/Scoring Chance Against rates as well as stable coaching situations seem like very strong bets to reproduce usable fantasy goaltender seasons.
- Any Early Predictions on #ZeroG Goalies for 2023-24?
- I view this question as an extension of the question above, where I’ll be looking moreso at teams with strong situations where the market is unsure about a new goaltender joining the team. Alexandar Georgiev and Ilya Samsonov are two obvious fits for this category from the previous season, and there will undoubtedly be more this season. I’m specifically looking at teams like Buffalo, Ottawa, and Los Angeles as squads that may surprise in 2023-24 with added defensive acumen and as a result become a terrific situation for a lowly-regarded goaltender to emerge as a league-winning option. We will clearly need to revisit this topic once the summer’s goaltending moves have taken place and the tandems are set, but if you’re in a dynasty league looking for an early edge I’d be looking at any goaltender slated to see volume in any of those situations.
So in summary, it’s full steam ahead for me and the #ZeroG train moving into 2023-24. I have no qualms about recommending the strategy to anyone who is willing to be vigilant on the waiver wire and take shots on goaltenders throughout the season. I picked up Filip Gustavsson in all three of my redraft leagues this offseason; with a little bit of elbow grease and an ear to the ground I don’t think it’s a stretch to say that you’ll be able to pull off something very similar in 2023-24.
That’s all for this year’s edition of the #ZeroG strategy update folks, I hope you had as much fun reading as I had writing it! There will be plenty of content all offseason so don’t be a stranger and stay plugged in as we break down player movement, playoff performances, and more. Make sure you follow Apples & Ginos on Twitter and join the Apples & Ginos Discord server for more content and to ask any fantasy hockey questions you may have.
Much love,
Nate
Apples and Ginos 