Monday, March 6, 2023

2022/23 Third Quarter NHL Betting Report

Welcome to my Third Quarterly Hockey Betting Report of the 2021/22 season. Unlike my weekly reports, the quarterly report will delve deeper into my team-by-team results and breakdown categories for the entire quarter. It should be noted that I’m not betting with real money. These are all fictional wagers in a spreadsheet. If you’re betting with real money, you should not be betting on every game, only the games you like the most. Whereas I’m betting on every game, every over/under, because it provides a complete dataset for macroeconomic analysis. Most of my lines were recorded from Draft Kings near noon Pacific time the day before games, and unlike last season, I’m recording alt pucklines for underdogs -1.5 goals and favorites +1.5.
 
To read my Second Quarterly report, click here. If you would like to read a comprehensive analysis of the last three years of hockey betting, my new book “The Hockey Economist’s Betting Prospectus” is available in the Amazon store. Just imagine this quarterly report but nearly 400 pages. You’ll get a profitability breakdown of the major categories since October 2019, and a chapter detailing the results for each team. Half of it discusses my own results, while the other half discusses what I should have been betting. To read more, visit the Amazon store.
 
My 3rd Quarter Profit: $5,134
My 2nd Quarter Profit: $5,527
My 1st Quarter Profit: $5,786

The third quarter of the NHL schedule (henceforth referred to as Q3) extended from Jan 10 to Feb 26, wrapping a few days before the NHL trade deadline. My Q3 kicked off in ominous fashion with a brutal week while compiling my Q2 Report. Evidently writing a 15,000 breakdown of everything that worked or failed in the previous 6 weeks was detrimental to my performance. The same thing happened in the last 7 days while compiling this report, but I’m blaming trade deadline roster volatility for that folly (read more here).
 
Given the strength at the top of the 2023 draft class, we saw an obscene amount of talent transferred from bad to good teams this deadline, as many managers did not want to risk lowering their number of ping pong balls in the draft lottery. This has me concerned that some of the analysis I’m presenting here won’t be transferable to Q4, better hindsight than foresight. The best example of this is the Chicago Blackhawks, who were the most profitable team to bet in Q3, but unloaded half their roster leading up to the trade deadline (most importantly Patrick Kane).
 
Chicago and Montreal went a combined 11-32 in the second quarter, with me generating $4,200 profit from their losses. Well they went 20-17 in Q3, costing me nearly -$3,000 when betting them to lose. Two weeks into Q3 my foot was firmly pressed on the brake pedal, but the damage was already done. The three best teams to bet on were Chicago, Detroit, and Montreal. I did manage $1,840 profit when betting those 3 teams to win, which of course did not pay for the money I lost betting them to lose at the beginning of the quarter.
 
A cold streak is always a good time to run some diagnostic tests on your methodology, or at least jolt you into making some changes/upgrades. Two weeks into Q3 the data that feeds my line value algorithms was modified to ignore results from the first quarter (pre-American Thanksgiving). That injected some heat into my cold streak, and reversed my trajectory. Granted, I’ll never know if my cold streak would have continued if that wasn’t done, but it’s extraordinarily self-satisfying when you make a change and immediately see a considerable improvement.

The actual best teams to bet against were not who casual fans might expect, at least those who rarely look at the NHL standings. There was no better team to bet against in Q3 than the Washington Capitals, which was also a major revenue source for my portfolio. They were followed by St. Louis, Winnipeg, Calgary, and Pittsburgh. I ran a nice profit betting all those teams to lose, except the Jets, who were among my worst teams overall. They took a big step back, and I failed to follow them down that new path. In fact Winnipeg was also among my worst teams to bet against, as they were only effective when my money was on their opponent.
 
My biggest revenue generator in the third quarter was over/under, which we’ll delve into deeper in the team and over/under sections. The primary driver of my success was Tampa and Vancouver overs, along with Minnesota, Columbus, and Dallas unders. My two best teams adding over and under results together were the Blue Jackets and Devils, while my two worst were the Red Wings and Blackhawks. This is my second season tracking over/unders, and Columbus has easily been my best team in that window. For whatever reason, they’re very predictable.
 

Over/Under

 


My 5 Best Over/Under Bets:                                Market’s 5 Best Over/Under Bets:

                                                                                ($100 wagers)

 

1) Tampa overs, (+$1,000)                                    1) Winnipeg unders, (+$1,111)

2) Minnesota unders, (+$988)                              2) Dallas unders, (+$1,107)

3) Columbus unders, (+$964)                              3) Minnesota unders, (+$968)

4) Dallas unders, (+$911)                                     4) Vancouver overs, (+$923)

5) Vancouver overs, (+$900)                                5) Edmonton overs, (+$758)

 

My 5 Worst Over/Under Bets:

 

1) Chicago unders, (-$646)

2) Winnipeg overs, (-$517)

3) Florida unders, (-$513)

4) Islanders overs, (-$505)

5) Calgary unders, (-$400)

 

Of my $5,000 Q3 profit, $3,000 of that was generated from over/under, bouncing back from a weak Q2. There was a 3-week window overlapping the end of Q2 and the start of Q3 when my over/under algorithm completely collapsed, losing -$1,832. It was a big enough loss to drive me into the red on my Q2 O/U revenue. I found myself in a similar situation last year around that time, when my current algorithm was born during the All-star break and crushed the remainder of the second half. I planned a new All-star deep dive in 2023.

 

Before the opportunity to test some new models had arrived, my previous formula was reduced to an advisory role with me increasingly investigating game logs, which reversed my negative trajectory. Granted, had my original formula stayed in operation for those 2 weeks before the break, it would have performed well too. The break presented an excellent opportunity to do some over/under model testing to try and improve my method. The idea of taking some time off for myself to relax and recharge never crossed my mind.


 

My existing algorithm performed very well post-AS break last season, but then again, overs performed so remarkably well in that sample that damned near anything you could have attempted would turn a profit. Those were unique circumstances on the tail-end of strict Covid protocols. History hasn’t been repeating itself. I tested 8 different models in this latest round, and every single one posted a profit on 2021/22 data; while only 3 generated a positive balance in this current season. The degree of difficult has increased.

 

My current algorithm did perform better than most of the others attempted, with the exception of one that produced better results in both seasons. Instead of average goals for both teams in their last 5 games, it takes the average from the last 8 games, but deletes the highest and lowest scores. If a high or low was duplicated, only one of them was erased. This was done to lessen the impact of outliers, which I’d observed causing me problems (most especially a Seattle 17-goal game followed by 4 consecutive unders).

 

The new method generated a respectable 4.4% return on games that were at least 0.25 goals above or below the betting total, which it was for 74% of games. That might have been case closed if not for the fact that my experiment parameters require a wager on every single game, which then left me to figure out what to do in the other 26%. In those games, unders went 102-76-12, cranking out more than $1,700 profit on $100 wagers. That would suggest it’s always good to take the under on close calls, but there’s no guarantee that sustains in the future.

 

It’s worth pointing out that my previous 5-game algorithm also had a higher rate of return when the average was 0.25 goals above or below the betting total. Had I only been betting those games, my performance would have been far better and I would not have been shamed into a deep dive. The closer calls were deflating my profit margin. Forcing a bet on every game lowered my rate of return. These two algorithms also generate the same recommendation in 75% of games. On the games when they disagreed, the two formulas had virtually identical success rates (the 5-gamer was right in 51%).

 

On those games when two algorithms disagreed, there was a predictable pattern to choosing the right side: Bet the under, which went 224-188-28 in the disagreement games. That also includes games from 2021/22 when overs were booming, and there’s more data from that season than the current one in the sample. This is more evidence to suggest that when in doubt, bet the under. Because of how well they performed in that 26%, all the permutations of the different formulas pointed towards unders and none of the combinations tested could consistently produce a positive number on overs.

 

My performance betting over/under was sensational for the two weeks leading into the All-star break, using the 5-game algorithm as the primary decision maker, but consulting how many of each team’s last 7 games went above or below the betting total. Making judgement calls when they disagreed, accepting 100% of the recommendations when they agreed. I thought perhaps I had stumbled onto a great new method, but the 7-game counter was among the worst of all the models at picking the correct outcome in 2022/23.

 

The new method worked very well in a limited 2-week window, but partially because of a high success rate on my judgement calls, which I figured would be impossible to sustain. I’m going to continue making judgement calls when the hybrid 8-game model isn’t 0.25 above or below the betting total, but will track my success on those wagers to see if that’s a quality method, versus just taking the under in 100% of those matches. Goal scoring did start creeping up post AS break, so betting every under when in doubt did not sustain.

 

I’m still consulting the 8 other models from my “deep dive”. During the break, my “Game Summary” worksheet where all my decisions are made was updated. Instead of just using 1 primary algorithm for over/under, I’m looking at what every model recommends for tomorrow’s game. You should check out my weekly reports to see how this all unfolds in the final quarter of the schedule. (note: my performance was terrible in week one of the fourth quarter, which I’m hoping is just a biproduct of trade deadline roster volatility.



 

Goalies

 

Goalies will be discussed in more detail in the team sections, so I’ll try not to ramble too much about the leaderboard to try and avoid repetition. It needs to be noted that when I’m discussing goalie win-loss records in the team sections, it’s entirely based on the team’s record when that goalie started. Similarly when discussing goalie over/under records, it’s the team’s record when that goalie starts. Also, in my world, overtime losses don’t get categorized separately. No bonus points from me, an L is an L, whether in the shootout or regulation.

 

Across all categories, my three best goalies (whether betting them to win, lose, over, or under) in Q3 were  1) Vitek Vanecek, 2) Linus Ullmark, 3) Andrei Vasilevskiy. Those were #1 goalies from 3 of the league’s best teams that I was betting to win often. But if you bet equal amounts moneyline and puckline on every goalie every game, there are some names you’d never expect to see, especially Jaxson Stauber, who more than doubled second place. He plays for Chicago and went 5-1 in his first 6 NHL starts, pulling off some lucrative upset victories.

 

Sadly the opportunity to participate in the Stauber profit has come and gone. They sent him to the minors (probably partially because he was doing too well) and then traded away most of their roster at the deadline, so don’t be tempted to bet him even if he’s recalled. It was also a good quarter to bet Ville Husso, but the Red Wings moved out some important pieces at the deadline that should hurt their Q4 output. Korpisalo cashed some longshot bets in Columbus, but has been traded to LA where he’ll likely play a supporting role. Jack Campbell had a few good weeks before falling apart again.

 

The goalies that cost me the most money (across all categories) were 1) Ilya Sorokin, 2) Sam Montembeault, 3) Petr Mrazek. Sorokin was good, producing a .923 SV%, but his goal support was insufficient, leading to a 6-9 record in 15 starts. They blew some games to tired teams that cost me some large wagers. I also struggled mightily when betting some big name goalies to win: Ilya Samsonov, Jake Oettinger (who should be a Vezina nominee), and Igor Shesterkin (the defending Vezina winner). You can read more about Shesterkin’s bad quarter in the Rangers section.

 

The two best goalies to bet against might come as a surprise if you spent the last 2 months on a deserted island. Darcy Kuemper and Jacob Markstrom were bad and their teams were bad. They were also my two best to short. If you bet Connor Hellebuyck to lose every start, you had a good quarter (sadly I was too bullish on the Jets, who regressed as a team). I burned nearly -$2,500 picking Monteambeault and Mrazek to lose, but I’m expecting to generate big bank from Mrazek losses in Q4 now that he’ll have an AHL team playing in front of him

 

 

Market’s 5 Best Goalies to Bet on:                     Market’s 5 Best Goalies to Bet Against:

($100 ML + $100 PL+1.5 + $100 PL-1.5)            ($100 ML + $100 PL+1.5 + $100 PL-1.5)

 

1) Jaxson Stauber, (+$3,026)                               1) Darcy Kuemper, (+$1,970)

2) Ville Husso, (+$1,317)                                      2) Jacob Markstrom, (+$1,890)

3) Alexandar Georgiev, (+$1,297)                       3) Charlie Lindgren, (+$1,662)

4) Jack Campbell, (+$1,057)                                4) Connor Hellebuyck, (+$1,104)

5) Joonas Korpisalo, (+$978)                               5) Thomas Greiss, (+$965)


My 5 Best Goalies to Bet on:                                My 5 Worst Goalies to Bet on:

(ML +PL)                                                               (ML + PL)

 

1) Linus Ullmark, (+$1,572)                                1) Ilya Samsonov, (-$1,102)

2) Vitek Vanecek, (+$1,551)                                 2) Ilya Sorokin, (-$1,082)

3) Frederik Andersen, (+$1,333)                         3) Pavel Francouz, (-$1,000)

4) Andrei Vasilevskiy, (+$1,127)                         4) Jake Oettinger, (-$975)

5) Ukko-Pekka Luukkonen, (+$973)                  5) Igor Shesterkin, (-$872)

 

My 5 Best Goalies to Bet Against:                      My 5 Worst Goalies to Bet Against:

(ML +PL)                                                              (ML + PL)

 

1) Jacob Markstrom, (+$1,003)                          1) Sam Montembeault, (-$1,500)

2) Darcy Kuemper, (+$968)                                2) Petr Mrazek, (-$956)

3) John Gibson, (+$840)                                      3) Alexandar Georgiev, (-$755)

4) Kaapo Kahkonen, (+$831)                             4) Dan Vladar, (-$609)

5) Felix Sandstrom, (+$496)                               5) Mackenzie Blackwood, (-$600)

 


My 3rd Quarter Results:

 

*Market Bets calculated by betting exactly $100 on every outcome this quarter*

 


Most betting categories will lose money in the long-term because we are playing a rigged game. I don’t mean that the league is literally fixing the outcomes of games to make more money for their betting partners, but rather that betting doesn’t offer “actuarily fair” odds. Two teams with an equal chance of winning aren’t both listed at +100, but rather -110, implying a 52.5% probability of victory. There is a 100% chance that one side or the other will win, but the probabilities add up to 105%. That is referred to as a “vig”, or the tax that oddsmakers charge to do business. This isn’t a charity.

 

That’s the explanation for how you can lose money whether you bet every single over or every under, and the longer the timeline you’re investigating, the more likely that both sides will incur a loss. That’s why of the 29 categories I’m tracking, only 6 are turning a profit on the whole season. The wider my window when reporting on category performance, the more likely it is to lose money. There is generally more optimism and good news to report in the 7-day window of a weekly report, with a lot more red ink in the quarterly version.

 

Of the few categories that posted a 3rd quarter profit, most of them are also generating positive returns for the full season. A big winner from Q2 that dropped considerably was road moneyline, which was also one of my best cats that quarter. While road moneyline reversed course, road dogs +1.5 goals and road favorites -1.5 goals performed very well, almost suspiciously well considering the declining MLs. Road winning percentage increased from 47% to 48.5% from Q1 to Q2, regressing to 47.9% in Q3. Small percentile changes can swing a category thousands of dollars from one quarter to the next.


 


It was noted in my weekly betting report roughly 14 days into the new quarter that road moneyline was starting to decline, but my suspicion was that there was simply multiple good teams on concurrent homestands (there was a higher than normal number of home favorites). My conclusion was “this isn’t necessarily the start of a new trend” when in reality it would be sustained in the coming weeks. The damage to my portfolio was minimal, since I was not consciously selecting road warriors, rather my line value algorithms just pointed me in that direction when there was a good opportunity.


 

Another explanation as to why trends are so unlikely to sustain in the long-term is that oddsmakers know exactly where they are losing money. If road teams are costing them disproportionately, they can just make line adjustments. Granted, that assumes that the entire public was aware of the trend. Perhaps most bettors just have a preference for picking home teams and the necessary adjustment for oddsmakers is offering a little more value on visitors to try and get an equal amount of money on both sides, guaranteeing a profit if either side wins. That’s how a winning angle can sustain longer term, if the public tends to bet the opposite outcome.


 

The anecdotal evidence that the public enjoys betting home teams a little more than they should is that they tend to be favored far more often than they actually win. Quite often when road teams perform well, so do underdogs. The covariance is large because they’re often the same thing. Underdogs started the season strong on the moneyline, but collapsed in the final week of the first quarter; proceeding to post a loss in 10 of the next 12 weeks. But road dogs were collapsing right around the time that road favorites really began heating up (having a fantastic December, cooling off in January).

 

Road dogs were unequivocally bad in December across every category, but improvements occurred in January and February. Their winning percentage climbed from 36.1% up to 38.1% to 39.7%. Each passing month the rate of return on the moneyline got higher. In January specifically, betting $100 on every road dog moneyline would have produced a -$469 loss, but betting every puckline +1.5 goals yielded $695 of profit in those same games: ergo, there was an improbable number of 1-goal losses in that month. That also means by osmosis that home favorites -1.5 goals was a big loser.


 


As previously mentioned, the large majority of favorites are home teams. If we’re talking about favorites -1.5 goals, there was a remarkable difference between the profitability of road versus home. First, there were nearly twice as many home favorites, which generally means you had to be extra good to be favored as a visitor, or at least there had to be a wider gap. Considering home teams won 55%, there were far too many teams favored at home who should have been underdogs. You don’t get that on the road side. There’s a higher bar to clear.

 

That’s also why it’s very important to draw the distinction when discussing favorites -1.5 goals. For me, who was burned for some big losses on the puckline courtesy of Chicago and Montreal near the beginning of Q3, my total PL investment dramatically decreased in the weeks that followed. The reason being, I was generally only betting puckline against the league’s worst teams (especially Chicago in December), and when those hit a roadblock, my foot slammed on the brakes. Towards the end of the quarter, I made a few extra PL wagers because favorite moneylines were growing increasingly expensive.

I’m generally fascinated with the dynamic between faves and dogs, with many of the categories I’m tracking breaking down across that delineation, also by home/road and moneyline/puckline. The prices on favorites certainly seemed to be rising as Q3 was approaching its end. We only saw 3 lines open beyond -400 from Oct 7 to Feb 12, and there was 3 in week 18. There have been 15 lines close beyond -400, and those teams are 14-1, generating a positive return if you bet them all (obviously). Problem is, if they went 12-3, you would only have broken even.

It might be time to re-evaluate my position on -400 moneylines, as they were a big net loser in 2021/22, but are dominating in 2022/23…(insert dramatic pause to investigate how I actually bet those 15 games)…okay, so in 10 of those 15 games, my money was on the favorite puckline, which they covered 13 times, generating $2,222 of profit. Puckline -1.5 did perform substantially better than the moneyline, so my inclination to avoid ML and bet PL was a big net positive. But I went 1-4 betting the underdogs in those games (the only successful bet was +1.5 goals).

What’s strange here (which you may have noticed if you looked at category results) is that longshots of +200 or higher on the moneyline are one of the best demographics on the season. Yet teams that closed beyond -400 went 14-1. Ergo: the longshots facing a -400 favorite went 1-14. That means that underdogs of +200 to +300 are doing very well (a -400 favorite tends to have a +320 opponent). Note that I have not recorded a single -400 moneyline from a road team this entire season. Every single one of those are at home.


Where this investigation gets interesting is that the +200 to +300 teams that are performing very well are not road teams. Yes, road dog ML is significantly outperforming home dog ML, except in the +200 to +300 range. Road underdogs are doing better, but that’s exclusively in the +100 to +200 group. Whereas home teams in the +100 to +200 range are getting destroyed, the least likely winners beyond +200 are producing fantastic returns. Betting $100 on every Home ML >=+200 would bank more than $1,600 profit on the season. Whereas +100 to +200 would create a -$3,700 loss.


In theory, the longshots should be worse teams with a lower probability of victory, so the fact that they are generating so much more profit than the better teams below them is paradoxical. Perhaps that’s just random variance that’s not consistent with previous seasons and is not replicable in the future. Fortunately for me, I recorded all the betting lines from the previous 3 seasons (which is broken down in my book), so I can actually investigate whether or not this phenomenon is completely new, or has occurred in recent history. Note to self.

The other category that needs to be applauded for its awesomeness is betting against back-to-backs. On the whole it’s been among my best categories following a bad first quarter.  There was an 11-week period where if you bet $100 on the moneyline for each rested team to defeat every tired opponent, you banked $2,045 on the moneyline and $3,505 on the puckline -1.5 goals (both dogs and favorites). Clearly I need to start shifting some of my btb moneyline investment to the puckline. It also needs to be said those returns are on the opening line, with a significantly lower return on the closing lines. The public loves betting these too.

 

My Best Categories:                                             Market’s Best Categories:

(all wagers)                                                             ($100 wagers)

 

1) Longshots moneyline, (+$2,600)                     1) Longshots moneyline, (+$2,455)

2) Overs, (+$2,056)                                               2) Road underdogs +1.5 goals, (+$2,193)

3) Road moneyline, (+$1,496)                              3) Road favorites -1.5 goals, (+$1,583)

 

My Worst Categories:                                          Market’s Worst Categories:

(all wagers)                                                             ($100 wagers)

 

1) Shorting back-to-backs -1.5, (-$360)              1) Home favorites -1.5 goals, (-$4,588)

2) Underdogs -1.5 goals, (-$200)                          2) Underdogs -1.5 goals, (-$2,765)

3) Home underdogs moneyline, (-$157)              3) Heavy favorites -1.5 goals, (-$2,635)

 

Market Best Moneyline Bets:                              Market Best Teams to Bet Against ML:

($100 wagers)                                                           ($100 wagers)

 

1) Chicago Blackhawks, (+$1,615)                     1) Dallas Stars, (+$849)

2) Montreal Canadiens, (+$702)                         2) Washington Capitals, (+$838)

3) Detroit Red Wings, (+$668)                            3) Calgary Flames, (+$777)

 

Market Best Bets +1.5 Goals:                         Market Best Teams to Bet Against +1.5 Goals:

($100 wagers)                                                        ($100 wagers)

 

1) Chicago Blackhawks, (+$465)                        1) Minnesota Wild, (+$630)

2) Edmonton Oilers, (+$418)                              2) Washington Capitals, (+$548)

3) Colorado Avalanche, (+$314)                         3) Pittsburgh Penguins, (+$399)

 

Market Best Bets -1.5 Goals:                          Market Best Teams to Bet Against -1.5 Goals:

($100 wagers)                                                        ($100 wagers)

 

1) Chicago Blackhawks, (+$725)                        1) Washington Capitals, (+$2,245)

2) Detroit Red Wings, (+$720)                            2) St. Louis Blues, (+$1,180)

3) Edmonton Oilers, (+$630)                              3) Winnipeg Jets, (+$1,080)


My 5 Best Teams to Bet on:                                Market’s 5 Best Teams to Bet on:

(ML +PL)                                                              ($100 ML + $100 PL+1.5 + $100 PL-1.5)

 

1) Boston Bruins, (+$2,244)                                1) Chicago Blackhawks, (+$2,805)

2) Carolina Hurricanes, (+$1,611)                     2) Detroit Red Wings, (+$1,566)

3) New Jersey Devils, (+$1,341)                         3) Montreal Canadiens, (+$1,543)

4) Chicago Blackhawks, (+$1,175)                     4) Colorado Avalanche, (+$1,336)

5) Tampa Bay Lightning, (+$1,027)                   5) Edmonton Oilers, (+$1,096)

 

My 5 Worst Teams to Bet on:

(ML +PL)

 

1) Dallas Stars, (-$1,325)

2) Winnipeg Jets, (-$1,068)

3) New York Islanders, (-$995)

4) Colorado Avalanche, (-$868)

5) Toronto Maple Leafs, (-$693)

 

My 5 Best Teams to Bet Against:                       Market’s 5 Best Teams to Bet Against:

(ML +PL)                                                              ($100 ML + $100 PL+1.5 + $100 PL-1.5)

 

1) Washington Capitals, (+$1,357)                     1) Washington Capitals, (+$3,631)

2) San Jose Sharks, (+$1,286)                             2) St. Louis Blues, (+$1,807)

3) Philadelphia Flyers, (+$1,239)                        3) Winnipeg Jets, (+$1,422)

4) Anaheim Ducks, (+$859)                                 4) Calgary Flames, (+$1,229)

5) Vancouver Canucks, (+$x)                              5) Pittsburgh Penguins, (+$969)

 

My 5 Worst Teams To Bet Against:

(ML +PL)

 

1) Montreal Canadiens, (-$1,493)

2) Chicago Blackhawks, (-$1,456)

3) Ottawa Senators, (-$905)

4) Colorado Avalanche, (-$835)

5) Winnipeg Jets, (-$705) 


 
Team By Team Power Rankings
 
The team-by-team gambling Power Rankings are ordered by the sum of all my bets on each team to win or lose, over or under for the entire season. They are my own personal power rankings, reflecting my own success picking the outcome of their games. These aren’t necessarily the best teams to bet on, as some were swung by a few instances of good luck or bad judgement. You’ll have to read the team summaries for a deeper understanding of the replicability. If you are going to be betting on hockey in the near future, it may help you to read about my own personal success and failure each quarter. I’ll also list the results of betting $100 on every outcome for each team.
 
 
1)  San Jose Sharks, ($4,204):
 

The San Jose Sharks lost 70% of their Q2 games, then started Q3 by losing 7 of their first 10. Being predictably bad is very profitable, and my money was on their opponent in 9 of those 10, helping them supplant Buffalo for top spot in my Power Rankings. They maintained that top spot right through the entire quarter, gradually increasing their lead on 2nd place. Technically they were my 4th best team of Q3, as my money was almost exclusively on their opponents. They did trade their best forward Timo Meier at the end of the quarter, so they are unlikely to get better anytime soon.
 
The Sharks did show some improvement later in the quarter after Kaapo Kahkonen took control of the net and registered some strong performances. Kaapo received that opportunity after an injury to James Reimer, who did not play well in Q3, going 2-6 with an .894 SV% (while Kaapo went 4-6 with a .907 SV%). Kahlonen ranked as my 5th best goaltender across all categories, posting a profit when betting him to win, lose, over, and under. I really had the “Midas touch” when the youngster got the nod.
 
The Sharks were involved in a lot of high scoring games in the first half which made their overs a great investment, but when Kahkonen got hot, suddenly their unders started cashing, finishing Q3 with a 10-9 record. Fortunately my algorithms were able to navigate that transition seamlessly, generating a really nice profit on both sides. They were not among the teams who shamed me into upgrading my method. Losing Meier will surely have a very negative impact on their overs, so be weary if you’ve been hitting that a lot this season.
 
 
2)  Tampa Bay Lightning, ($3,397):
 

The Tampa Bay Lightning dropped all the way down to #24 in my week 12 Power Rankings, then climbed all the way back up to #2 by week 17. A large majority of that profit was attained by betting Tampa to win and from over/under (also producing a decent profit betting their opponents). Once again they were a dominant team on home ice, and less effective on the road (their home win % did drop from 90% to 78% from Q2 to Q3). I found myself betting the home team in 12 of their first 14 Q3 games, mostly because there was too much juice on their moneylines as visitors.
 
That “bet the home team” blueprint has also been a successful strategy for betting Tampa games in previous seasons, so this wasn’t something that I recently stumbled upon. Their effectiveness overall was negatively affected by the diminishing play of Andrei Vasilevskiy, who was sensational in Q2 but came back down to Earth (relatively speaking) in Q3. He still finished the quarter with a respectable .918 SV%, but also had 5 starts in which the team allowed 5 or more goals. Thankfully that was offset by 9 games allowing 2 or less, but there were at least a few turds on his game log.
 
Their unders went 11-7-1 in Q2, but this was another team that completely reversed course with overs going 13-8-0 in Q3. Though looking at their game log, early in Q3 they did have an unusually high number of opponents good at scoring goals and bad at preventing them. Back-up Brian Elliot only started 4 games, going 1-3 with an .894 SV%. It actually didn’t matter which goalie was in net when betting their over/under, but Vasilevskiy overs had a better return on the closing line. Looks like the public likes betting the under when he’s confirmed as starter, but you would have lost money doing that regularly.
 
 
3)  Buffalo Sabres, ($3,379):
 

The Buffalo Sabres went 12-7 in Q2 to help them climb to the top of my Power Rankings, but hit a wall losing 4 of their first 5 third quarter games (ceding their #1 rank to San Jose). My ticket aboard their bandwagon was purchased in October, so while their mini-slump was mentioned in my game notes, I also wrote “I’m not ready to quit them yet”. That loyalty cost me a big wager when they lost to Chicago, but my foot stayed on the gas pedal, as they won their next 5 in a row, recouping my loss. Keeping my faith led to a nice reward.
 
One characteristic that carried over from one quarter to the next was their strength on the road, where I bet them to win nearly every opportunity (except one game they were a -200 favorites vs Anaheim, which Sabres won 7-3). The biggest difference between Q2 and Q3 was their performance on home ice completely collapsed, winning only 33%. Despite those struggles, I still managed a small profit betting their home wins (they only won 3 home games, but those were my largest bets, 2 vs back-to-back, 1 vs Anaheim). Regardless of venue, my money was on Buffalo 17 times in 20 Q3 games for $883 profit.
 
The team continued to get subpar goaltending, with Ukko-Pekka Luukkonen starting 60% of their games, winning 50% with an .895 SV%. Craig Anderson was their best goalie in limited action with a .915 SV%, but unlikely it would have been that high with a heavy workload. You might expect the goaltending to help their overs, but they only went 11-9. Their goal scoring did regress to the mean, and their totals tended to be very high (reaching 7 on the closing line 6 times). Oddsmakers were just smart about setting the prices, nerfing a potential opportunity.
 
 
4) New Jersey Devils, ($3,326):
 

The New Jersey Devils continued their outstanding play on the road in the third quarter, which began with a road-heavy schedule. The big difference was they became much more effective on home ice, which has been a problem for them in past seasons. I was a net loser betting their Q2 games, but only because of a catastrophic week when they went 0-4 costing me -$2,000. If that one week never happened, they would easily be #1 in my Power Rankings. My faith did waiver for a few weeks, but it didn’t take many wins before I was back on their bandwagon.
 
At one point they were without Jack Hughes for a few games due to injury, compelling me to bet their opponents. He returned after 4 games, and the Devils went 2-2 in his absence (with me going 2-2 on my bets). Otherwise two thirds of my money was on Devils to win in Q3, but did have 2 large wagers on Dallas and Los Angeles to win when New Jersey was on a back-to-back, which they won despite the fatigue. Note to self: Jersey is 4-1 on back-to-backs against rested opponents this season, which cost me -$900 in Q3, accounting for a large share of my wagers on the opposition.
 
The Devils were my second best teams for over/under wagers, as their overs went 11-7-1. Their goal scoring rebounded after a dip in Q2, and their goaltending also improved. Vitek Vanecek went 10-2 with a .912 SV% and accounted for all my profit betting New Jersey. His back-up MacKenzie Blackwood had a higher SV% but a worse goals against average. There wasn’t much difference between them in terms of over/under, but Blackwood only went 3-3 in his 6 starts. I lost -$600 betting Blackwood to win, and another -$600 betting him to lose.


5) Anaheim Mighty Ducks, ($3,184):
 

The Ducks were not very mighty in the first half, winning only 29% of their games. Their futility inspired me to predict on New Year’s Eve that they would be the eventual winning of the Connor Bedard sweepstakes. But somebody forgot to inform their roster that this season is a lost cause, as their play improved early in Q3. They went on a 5-2 run at the beginning of the quarter, but that was short-lived, embarking on a 6-game losing streak shortly after the All-star break. Granted, Colorado was the only good team they beat in those 5 wins, otherwise they beat Arizona twice, Columbus and Chicago. That’s not exactly murderer’s row.
 
That little mini-hot streak did help convince me to bet them to win more often, but generally only when they were extreme longshots like +320, +280, +280, +320, +215, +210, +170, etc. This was not a conscious strategy on my part, it was just a function of my line value algorithms complaining about the price on their opponents. It became harder to justify betting Anaheim when the new cold streak took hold, as they were blown out by 3 or more goals 5 times in 6 games. Yet I did manage nearly $400 profit betting the Ducks to win/cover, which is impressive considering they only won 37% of their games.
 
There was an even split in their Q2 over/under, but their overs caught fire in Q3, going 13-4-2 as both goal scoring and goal allowing increased. It didn’t matter which goalie was in net when making over/under picks, as all 3 had similar splits. John Gibson was decent with a .908 SV%, but he got peppered with shots in the process, sporting a GAA of 4.00. He accounted for nearly all of my profit both when betting them to win and lose. Back-up Anthony Stolarz went 3-1 with a .920 SV%, sadly I bet their opponents in all 4 of those games, leading to a net loss in his starts.
 
 
6) Boston Bruins, ($3,120):
 

The Boston Bruins continue to lead the NHL standings by a considerable margin, and I had a strong first half when betting them to win. Unfortunately, I burned -$1,375 on their over/under as my algorithm struggled at recommending the right outcome. That did eventually correct itself, converting the Bruins into my most profitable Q3 team (across all categories). They ranked #27 in my week 11 Power Rankings, climbing up to #6 by the end of week 18. Their winning percentage remained the same from Q2 to Q3, equally good at home as the road.
 
Good as Boston was, if you bet $100 on all their Q3 moneylines, you only walked away with $201. There was a tax on their lines, especially on home ice. You could get slightly better value on their road ML, but not much. That $201 profit breaks down into $2 at home and $199 on the road; whereas you could have generated twice as much profit from their pucklines -1.5 goals. The small amount of money I invested in Boston opponents actually generated a nice return, thanks to Seattle, Carolina, and Tampa. It required a very strong opponent for me to even consider betting Bruins to lose.
 
While my performance on their over/under did improve, the profit was only $145. That’s still much better than a sizeable loss, so I’m chalking it up as a victory. Both their goalies were outstanding, with Ullmark sporting a .937 SV% and Swayman .934. In terms of betting their wins and losses, it didn’t matter which goalie was in net. Ullmark unders went 8-3 while Swayman unders only went 3-2-2. Yet I lost -$134 on Ullmark O/U, with $279 profit from Swayman. Ullmark was arguably more predictable, yet led to a worse result (mostly due to blown overs).
 
 
7) Minnesota Wild, ($2,238):
 

The Minnesota Wild kicked off the third quarter by playing 8 of 11 games on the road, which hurt their performance as a home dominant team. Fortunately, 8 of their last 10 were home, and I bet them to win all but 1 of those home games with mixed results (they won 5 of 8). There was very little change in their home winning percentage from Q2 to Q3, but their road win % was cut in half. The biggest issue facing the team was a dramatic decrease in goal scoring, dropping from 3.7 in Q2 all the way down to 2.3 in Q3.
 
The only problem with betting Minnesota to win at home is that oddsmakers are also aware of their proficiency and will often make the lines prohibitively expensive. This did lead me to bet some of those opponents, like Arizona +300 and Buffalo +155, both games Minnesota won, but only by a goal. I actually picked the Wild to win more often the road, but was mostly swayed by generous line offerings. Granted, most of those bets were losers, so maybe the line was priced right and my algorithms were wrong.
 
Perhaps the biggest change from quarter to quarter was their over/under. In Q2 overs went 11-8-2, but that flipped to 4-12-2 in Q3. That transition cost me a few bets at the beginning of the quarter, but went 9-2-1 in their final dozen games. The big news was the breakout of goaltender Filip Gustavsson, who missed a few games with injury, but firmly established himself as the #1 starter upon his return, finishing the quarter with an outstanding .934 SV%. Marc-Andre Fleury wasn’t bad (.912 SV%, but had a losing record). Both goalies had nearly identical over/under splits.
 
 
8) Carolina Hurricanes, ($2,207):
 

The Bruins might have finished Q3 atop the NHL standings, but the highest winning percentage of the third quarter actually belonged to the Carolina Hurricanes (thanks to Carolina beating them in their head-to-head meeting). The only reason I lost money betting Carolina to win in the second quarter was because of 3 upset losses to close Q2 that cost me -$1,600. In the course of one week, they dropped from #8 to #25 in my Power Rankings. That proved to be a blip in the radar not the beginning of a new trend, as their otherwise dominant play continued in the third quarter, winning 78% of their games.
 
Most of my money (82%) was invested in Carolina victories. Those bets made on Carolina opponents tended to be for my minimum amount when there was good value on the underdog. I managed a $258 profit when betting Canes to lose, but otherwise that was not a wager you wanted to be making often. The Canes played a home heavy schedule in Q3, but went 5 for 5 on the road. I actually generated a decent return betting Carolina opponents, but only because I took Anaheim as +350 dogs when the Canes were on the second leg of a back-to-back.
 
This was shaping into a strong quarter for Hurricane overs, at least until their unders cashed in 5 of their last 6 Q3 games, finishing with even 9-9 split. Goal scoring was up while goal allowing was down. Frederick Andersen was finally healthy, going 10-2 with a .914 SV%, while Antti Raanta went 4-0 with an astonishing .953 (though his unders only went 2-2). Across all categories, Freddy was my 4th best goalie to bet, with most of my profit coming from his wins. The public liked betting Hurricane opponents when Raanta was named starter, but that was a mistake.
 
 
9) Washington Capitals, ($2,013):
 

The Washington Capitals were among the NHL’s best teams in the second quarter, but alas that proved to be woefully unsustainable (dropping from a 70% Q2 win % to 32% in Q3). Their goal scoring dropped considerably (from 3.7 to 2.5), while their goal allowing increased (from 2.2 to 3.6) as both goalies began to struggle. All of that despite Nick Backstrom and Tom Wilson returning from injury. My stake was split almost 50-50 between Caps and their opponents, pulling a nice profit on both sides (albeit twice as much from their losses).
 
My results betting Caps to win vastly exceeded expectations thanks to going “all in” on Washington to beat Arizona (which in retrospect was riskier than I appreciated in the moment), which they won 4-0 and another victory against the struggling Penguins. If you deleted those games, the results get more bleak. They’ve been without their best defenseman John Carlson for several weeks, and that loss eventually became insurmountable. Goaltender Darcy Kuemper dropped from a .937 SV% down to .889, and no NHL gatekeeper was more profitable if you bet them to lose every start.
 
Wins and losses notwithstanding, my algorithm continued to struggle with their over/under, though not nearly as badly as Q2. Inconsistent goaltending was a major contributor. Their unders went 9-7-3, and my money had the appropriate split, but just not on the right games. Kuemper started 15 of 19 games, as Lindgren posted an .879 SV% in limited action. Kuemper unders went 8-6-1 despite allowing more goals, as the offense provided inadequate support. The Caps started selling off assets at the end of the quarter, but it’s entirely plausible they’re better in Q4. They can’t get much worse.
 
 
10) Columbus Blue Jackets, ($1,934):
 

The Columbus Blue Jackets were a bad team in the second quarter and my profit betting them to lose was very disappointing.  The problem was, there were a few games where I had especially large wagers riding on their opponents, and the BJs pulled off upsets. As a result, the amount I was laying on their opponents decreased, but my profit betting that outcome increased. My foot eased off the gas pedal, they improved slightly, leading me to a strong Q3. The key to betting them successfully was in the goalies, bet Joonas Korpisalo to win (.912 SV%) and Elvis Merlikins to lose (.894 SV%).
 
I actually didn’t even notice until beginning to compile this report that there was a stretch where my money was on Columbus for 12 consecutive games. Reviewing my notes from those matches, nowhere am I professing a new love for the Blue Jackets or an eagerness to join their booster club, every single note is complaining about the line price on their opponent. My consulting algorithms kept telling me the lines were off. I walked away from those 12 games with a small profit and would not have been better off betting the other side.
 
Columbus unders went 15-6 in Q2, but their overs went 6-2 to start Q3, at which point it was noted in my weekly report that their overs might be resurgent. Then that 6-2 run was followed by 6 games of alternating back and forth, then closing the quarter with unders going 6-1. That level of trend shifting should have caused problems for my algorithms, yet my O/U results were outstanding. Those shifts were actually entirely predictable. What’s strange is that Korpisalo was the better goalie, but was better to bet over, while Elvis was worse, but was better to bet under (which had to be a function of goal support).
 
 
11) Arizona Coyotes, ($1,752):
 

The Arizona Coyotes taking a month-long road trip in November provided them with a home-heavy schedule for the remainder of the season, and they sustained their strong play at Mullet Arena. They continued to suck as visitors, but the problem was the line prices on their opponents were absurdly expensive when they played good teams. Bad as they were, there were a lot of 1-goal games. They went 1-7 on the visitor moneyline, but 7-1 on the puckline +1.5 goals. With each passing quarter, my stake in Coyote opponents has grown larger, but their winning percentage has barely changed.
 
Arizona overs went 14-8 in Q2, but that flipped in the third quarter, and I went 0-3 on their overs while that shift was occurring. Eventually my algorithm did get on their unders, which went 11-8-1. My final results were neither good or bad, but I’ll chalk that up as a victory given the size of the shift from one side to another. That could have been much worse than it was. One of the big changes was substantially improved play from back-up goaltender Connor Ingram, who moved into a nearly equal time-share with Vejmelka. Both goal scoring and goal allowing diminished, boosting their unders.

Karel Vejmelka was decidedly mediocre, posting a .903 SV% while Ingram was all the way up to .928. My rates of return were very similar for both goalies across all categories, as it didn’t make a big difference to me who started. Vejmelka did have a slightly higher win % despite the lower SV%, and even posted a nice profit if you bet him to cover -1.5 every opportunity (only hitting 3, but +475, +390, and +300) while Ingram didn’t cover -1.5 in any game. Both were good bets +1.5 goals.
 
 
12) Edmonton Oilers, ($1,585):
 

After the Oilers struggled in the second quarter, Connor McDavid put the team on his shoulders and carried them to a 61% Q3 winning percentage. McDavid is having the most dominant season we’ve seen in the salary cap era, and this team would be competing for top spot in the draft lottery if he wasn’t on the ice. This was my best quarter betting their games, mostly from picking them to win in combination with overs. If you were consistently betting Oilers to win and parlaying it with the over, you had a great quarter, especially when Jack Campbell started (he went 8-4 and his overs went 12-0).
 
What initially lured me back onto their bandwagon was the return of Evander Kane, whose injury made me very pessimistic about betting Edmonton. They are a much better team with him in the line-up. The Oilers continued strong play on the road, and I often found myself gravitating to the road team in many of their games. Edmonton opponents only solicited 16% of my total money wagered, which did generate a nice return. It was actually better to bet against Stuart Skinner than Campbell, despite Skinner being the better goalie.
 
Those aforementioned overs went 14-4-0 (after going 15-7-1 in Q2), producing even greater profit thanks to a 1.1 goals-per-game increase in scoring. Their goaltending was largely erratic, as both Campbell and Skinner had hot (or at least decent) and cold streaks. They did move into a time share for a few weeks, at least until Campbell collapsed again. Great as their overs were, their unders actually went 4-2 when Skinner was in goal. He might have won a lower percentage of his starts, but there was less scoring overall when he played.
 
 
13) St. Louis Blues, ($1,375):
 
 
The optimal St. Louis strategy in the second quarter was just betting the visitor every game, as the Blues were brutal at home and good on the road. That trend did sustain itself when they began the quarter with a 7-game homestand, going 3-4. Then they went on a road trip and went 0-3, before returning to St. Louis for another homestand. They would play 13 of their 18 Q3 games on home ice, winning just under half of them, but their road record completely collapsed, going 0-5.
 
Near the end of Q3 the Blues traded two very important pieces in Vladimir Tarasenko and Ryan O’Reilly, leading oddsmakers to start charging a significant tax on their opponents. My results were strong when betting them to lose (accounting for 63% of my money wagered), while posting a small loss when betting them to win. My feelings about this team were very pessimistic, but frankly the value being offered on their lines was very seductive. All my Blues bets they were either a significant underdog, or facing an opponent who played yesterday. They won 3 in a row after the Tarasenko trade, then lost 6 in a row after the O’Reilly trade.
 
The problem for me wasn’t betting their wins and losses, but rather over/under. Their overs went 17-6 in Q2, but that shifted to 7-9-2 in Q3. Jordan Binnington started 15 of 18 games and did improve his play after a terrible .874 Q2 SV%, up to a decent .903 in Q3. His back-up Thomas Greiss saw very little action, probably due to his .862 SV%. Frankly this team should be tanking for draft position, but the coach Craig Berube seems too competitive to lose hockey games, so management needs to sell the tools at his disposal.

 
14) Florida Panthers, ($1,336):
 

The second quarter was not kind to the Florida Panthers, losing 41% of their games and falling behind in the playoff race. I was fortunate to invest in their opponents, generating a nice return betting them to lose. Well the Cats were able to stop the bleeding in Q3 and ceased being a good team to short. It took me a few weeks to completely ease my foot off the gas pedal, but the reversal didn’t prove too costly. Despite the Panthers winning 55% of their Q3 games and 83% of my money wagered on their opponents, I still managed a small profit on those bets. Betting them to win at home and lose on the road was the winning formula (which we’ve seen from them before).
 
Florida unders went 12-9-1 in Q2, but that trend swung hard in the opposite direction, with their overs going 14-5-1 in Q3. Goals against did not significantly change, but the offense caught fire, climbing from 3.1 goals per game up to 3.9. What’s surprising about that is they ran into injury issues in goal and were forced to start Alex Lyon for 5 games (going 2-3 with an .887 SV%). Spencer Knight missed most of Q3 with some form of ailment and was eventually placed in the NHLPA Player Assistance Program.
 
Fortunately for Florida, Sergei Bobrovsky stepped up his play, going from a .902 Q2 SV% up to .914 in Q3. Bobrovsky overs only went 8-5, while the other two keepers went 6-0. Across all categories, I lost -$533 in Bobrovsky starts (most of that loss coming from blown unders), profiting $632 when the back-ups got the nod (most of that coming from overs). They were still outside the playoffs when the quarter ended, but have good expected goal numbers so still attract more public money than their record warrants.
 
 
15) Toronto Maple Leafs, ($1,249):
 

My sensational second quarter betting Toronto to win proved too good to be true, but it wasn’t entirely unexpected as their best player Auston Matthews missed a chunk of Q3. My investment in their opponents increased, but the Leafs performed disturbingly well in those games specifically. They also saved their worst performances for when I picked them to win, as 2 losses to Ottawa and Montreal cost me a combined -$1,100. One was a bad beat, making a giant wager on Leafs to win and finding out Matthews was injured the next day. Shit happens.
 
They continued to be a dominant team on home ice, but their road winning percentage collapsed. The good news for me was that I rarely picked them to win on the road and they played a large majority of their games on home ice. 67% of my total investment was on Leafs to win, leading to a -$693 loss on those games. More often than not my money was on Toronto, but their line prices got increasingly expensive even as their winning percentage declined. Most of the games that I bet their opponent, my game notes complained about the line price.
 
Ilya Samsonov was the starting goaltender for every penny that I lost betting Toronto to win. He posted a strong performance with a .919 SV%, but was out-dueled by the opposing netminder on the games that cost me the most money (I lost -$1,000 betting him to win, and he went 9-5). Though I did perform well when betting Samsonov over/under (his unders went 7-6-1), going 0-5 on their O/U when Woll or Murray got the nod. Matt Murray spent most of the quarter on injured reserve and was bad in his 3 starts, posting an .880 SV%. That trade hasn’t worked out for Toronto, but the Samsonov acquisition has paid dividends.

 
16) Pittsburgh Penguins, ($879):
 

After a mediocre first half, the Pittsburgh Penguins hit a wall in Q3 as an injury to Tristan Jarry forced them to lean far too heavily on Casey DeSmith, who was not equal to the task. Despite losing one of their most valuable players, oddsmakers continued pricing their lines like they were an elite team. When Jarry returned from injury, he was not at an elite level, only starting 5 times in the quarter and posting a .911 SV% (down from .930 in Q2). My performance picking their wins and losses would have been much better had I just bet them to lose every single game.
 
Reviewing my game notes, there was a lot of confusion about why the price was so high for a struggling team. That skepticism led to a profit betting Pittsburgh opponents, but there were two games when I made big bets on Pens to win because their opponent played the night before, but Winnipeg and San Jose managed to pull off upset victories as +145 and +195 underdogs. Here’s a fun coincidence: I bet the Pens to win 5 times (moneyline) in Q3 resulting in a -$528 loss. If you bet $100 moneyline on the Pens to win all 20 Q3 games, you lost exactly -$528. I had to run a diagnostic to make sure that wasn’t an error.
 
Pittsburgh over/unders were an enigma that confused my algorithms, which was attributable to their erratic goaltending. Their goals scored per game barely changed, but their goals allowed increased substantially thanks to the increased load on DeSmith, as overs went 10-9 (after going 6-13-1 in the second quarter when Jarry started most of their games). Casey did have a winning record in Q3 (7-6) with an .899 SV%, but if you bet $100 on him to win every game, you lost -$497 because the line prices were often expensive, relatively speaking.
 
 
17) New York Islanders, ($738):
 

The New York Islanders won 60% of their games in the first quarter, 48% in Q2, then lost 8 of their first 9 in Q3 (with 6 of those on home ice). The descent grew so untenable for the fan base that fans were chanting for Lou Lamoriello to be fired. Feeling the pressure, Lou made a big trade to acquire Bo Horvat trying to improve their biggest deficiency, scoring goals. But for those of us who were crushing their unders in January and February, the prospect of more Isles goals was anything but tantalizing. Their unders were on an 8-3 run when the trade was made, then the over hit in 4 of 6 games after the trade, which cost me -$600.
 
They might have lost 52% of their Q2 games, but they won 67% at home. That led me to regularly bet the home team in most Islander games, but it was not a very profitable strategy because they got worse on home ice. As a result, I did well betting the Islanders to lose on the road, but performed badly betting them to win at home. The two that did the most damage were both on a rested Isles team to beat Vancouver and Ottawa who played the night before, the latter with a goalie playing his first NHL game (which was after the Horvat trade when I grew more bullish on NYI).
 
Not long after that Ottawa loss, news broke that Matt Barzal suffered a serious lower body injury and might be lost for the season. They were already struggling, which encouraged me to increase my investment in their opponents, and they went 3-0 the following week without their best offensive player. Their biggest strength was their goaltending, as both Varlamov and Sorokin were outstanding. That’s a big reason why their unders went 14-7-1. Most of my O/U stake was on the under, only generating $23 profit (thanks to the reason mentioned above). Across all categories, I banked $463 in Varlamov starts and lost -$1,593 on Sorokin starts (which was the most money I lost on any goalie in Q3).
 

18) Vancouver Canucks, ($693):
 

The Vancouver Canucks fired their head coach in the second quarter, bringing in the widely respected Rick Tocchet to right the ship and take a run at a wildcard spot (at least I’m assuming that’s what ownership had in mind). Sadly for the owner, the team got worse (winning % dropped from 55% down to 30%), which actually served to boost my performance betting their games. The biggest mistake that I made was assuming that they might get a “new coach bump” that never quite materialized, resulting in a -$475 loss. Their biggest problem was not getting NHL quality goaltending.

  
Spencer Martin and Collin Delia started 16 of their 20 Q3 games, posting save percentages of .837 and .863 respectively. Delia still managed a 4-5 record while Spencer Martin went 0-7 and lost his job to Arturs Silovs, who actually posted a respectable .908 SV%. I might need to buy a membership in the Silovs fan club, going 4 for 4 picking the winner of his 4 starts. Unfortunately, I lost money on his over/under because there was an even split when my algorithms were “all in” on Van City overs thanks to their otherwise porous goaltending and decent goal scoring.
 
One trend that firmly took hold was their overs becoming a fantastic wager (going 16-4-0). I ran a nice profit betting that outcome, but ran into issues when they played low scoring teams. My algorithm did incorrectly direct me to a few under wagers in those cases, when everyone was high scoring in Canuck games. One thing that I neglected to check in my over/under study; if one team is significantly above the betting total and the other is well below, which outcome is more likely? That may need to go on my “To-Do list”.


19) Dallas Stars, ($654):
 

The Dallas Stars took a big step backwards in the third quarter, as their winning percentage plummeted (from 59% to 39%), with most of the damage happening on home ice. I’m a big believer in the core of Jason Robertson, Miro Heiskanen, and Jake Oettinger, so my natural inclination is to bet them to win, but often found myself turned off by the line price. Jake Oettinger was outstanding, posting a .933 SV%, but that was only good enough to win 6 of 14 starts. If you bet $100 on Jake to cover every moneyline, you lost -$366; but you could have profited $276 by betting him on the puckline +1.5 goals every opportunity.
 
My money was split relatively evenly between the Stars and their opponents in the third quarter, with the team on the other side of my wager winning 75%. Dallas lost games to bad teams when I picked them to win, and won difficult games where I bet their opponent. The two that did the most damage was a win (on a back-to-back with Wedgewood in goal) against LA and a loss to San Jose that cost me a combined -$1,000. I had bet Dallas to win 4 consecutive times when they lost to San Jose, which fueled my gloom, leading me to bet their opponent in 5 of the next 6. They also had a 4-3 loss to Chicago who had played the night before, costing me -$500.
 
Picking their wins and losses might have been a perilous endeavor, but there was an opportunity to make a lot of money betting their unders, which went 15-3. All these low scoring games helped mitigate my losses. Both goalies produced profit on the under, so you didn’t need to wait for starter confirmation before placing your wager. Wedgewood only had 4 starts in Q3, but posted a respectable .919 SV%. If you bet $100 on Oettinger to lose every game by at least 1.5 goals, you lost -$980; but the same wager when Wedgewood got the nod, you banked $540.

  
 
20) Ottawa Senators, ($619):
 

The Ottawa Senators maintained their strong play from the second quarter, and my biggest problem was not betting them to win often enough. There was an acknowledgement in my game notes that the Sens had been doing well, but I was often discouraged from betting them because the line prices were too expensive, at least for their games against inferior teams. I was never thrilled about buying Ottawa as a significant favorite, especially after they ran into injury issues in goal. Cam Talbot missed some time, then not long later Anton Forsberg suffered a season ending injury.
 
While I wish Anton a speedy recovery, it was actually good news for my betting portfolio when he went down, as he was costing me several wagers (totaling -$792 while posting $643 profit when Talbot or Sogaard started). I mostly avoided making large wagers on Ottawa opponents, except for one game against Toronto that the Sens won 6-2, which inflicted the most damage on my Sens portfolio in Q3. The bet was registered 28 hours before puck-drop, and we found out next morning Austin Matthews was injured. Shit happens.
 
Ottawa unders went 14-6-2 in Q2, but their overs came storming back to go 10-8 in Q3. There was an increase in both goal scoring and goal allowing, thanks to both goalies suffering injuries. They had a 5-game stretch with Mads Sogaard and Kevin Mandalorian protecting the net, but they managed to win 3 of those games, one of them particularly expensive when they were on a back-to-back and beat the rested Islanders. I managed to pull a profit from both their overs and unders despite Talbot unders going 4-1, Forsberg overs going 5-2, Sogaard overs going 4-0, and Mandalorian unders going 2-0.
 
 
21) Seattle Kraken, ($289):
 

The Seattle Kraken were among the head-turning surprises of the first half, improving dramatically on a disappointing inaugural campaign. Sadly for Kraken fans, their play did start to slip in Q3, and oddly my performance betting their games improved. My splits of money wagered on Kraken (69%) vs their opponents was very similar from Q2 to Q3, but my rate of return on their wins got worse while my return on their losses improved. They continued to be a strong road team, and even when they were struggling, the value on their road lines was often very appealing. 
 
My week 14 Betting Report noted Seattle had won 8 in a row (7 of those on the road), but by week 17, there was an expressed concern about the state of the Kraken (due to only 4 wins in their previous 11 games). The bus came to a crashing halt very quickly. They were also upset 3 times by teams who played the night before. One interesting aspect of their schedule was how often one of the teams involved played the night before, with 9 back-to-backs in 20 games. This heavily influenced my bet selection, as my foot was firmly pressed on the gas pedal shorting tired teams in Q3.
 
The real reason for my big improvement betting their games was over/under, which for me was an absolute trainwreck in Q2. They had an 11-game stretch alternating back and forth from over to under. Then unders got hot, then overs had a good week, then it went back to alternating. Through it all, my new method navigated the madness admirably. Philipp Grubauer started 11 of 20 games and was the better goaltender, posting a .905 SV% compared to Martin Jones .877. Gubauer unders went 7-4 while Jones overs went 6-3. It paid off to figure out who was starting and betting accordingly.

  
 
22) Colorado Avalanche, ($169):
 

The Colorado Avalanche were brutal in the second quarter, at least by the standards of a Stanley Cup champion, but the team had a legitimate excuse being decimated by injuries. I certainly wasn’t complaining, banking big profit betting them to lose. Well eventually Nathan MacKinnon did return to the line-up, they got healthier, and a new hot streak was born. This did not translate into gambling profits for me because of their loss to Chicago that cost me -$650 (with no Patrick Kane). That loss really compounded my Avs pessimism, but it proved to be the start of a 6-game winning streak.
 

For the second consecutive quarter, the Avalanche had nearly identical winning percentages at home and on the road, the difference being they were much better in Q3. Even after MacKinnon and Nichushkin returned, Cale Makar was in and out of the line-up with concussion issues. So while their performance was improving, there were still news items to feed my pessimism. The end result was a significant loss on both sides, losing -$1,000 betting Pavel Francouz to win (3-3 with a .929 SV% in Q3) and -$755 betting Alexandar Georgiev to lose (9-3 with a .926 SV%).
 
While my overall performance betting Avs games significantly declined, one area where I improved considerably was over/under with my algorithms struggling in Q2, but good in Q3. They scored more goals and allowed fewer, with their unders going 10-8-1. My algorithms leaned a little too heavily on the overs, but still walked away with a profit. With the talent exodus from west to east leading up to the trade deadline, we might see the best western teams crushing the competition in the fourth quarter. Granted, good luck getting a fair price on Colorado vs Chicago or Anaheim.

 
23) Philadelphia Flyers, ($60):
 

The Flyers were among my worst teams in the second quarter, all stemming from big upset victories against really good teams (a majority of those after losing Carter Hart to injury). I did learn my lesson from those errors, and pumped the brakes on betting Philly to lose (they had won 7 of their previous 8 games at the beginning of the quarter). That inspired me to bet them to win more often, as they solicited 39% of my investment (up from 8% in Q2), but that turned out to be a net negative. It would have been a better strategy to bet against them every game, but hindsight is 20-20.
 
My performance betting Flyer opponents jumped from -$1,687 in Q2 to +$1,239 in Q3. One area where the team did get worse was on the puckline +1.5 goals. Carter Hart was decent, posting a .907 SV%, but you would have done well betting him to lose every game. He’s always a risk to steal a game against a superior opponent, so there’s always a danger going too big on the other side. In Q3 Carter was just as likely to give up 4 or more goals (including vs Montreal & Vancouver) as allowing 2 or less goals (including against Edmonton and Winnipeg). It’s hard to trust that either way.
 
Rookie Samuel Ersson was my worst goalie in Q2, costing me -$1,631 in 5 starts, but that improved considerably in Q3, with $1,002 profit in only 3 starts (going 6 for 6 on my bets). Ersson did take a big step backwards, as his SV% dropped from .924 in Q2 to .869 in Q3. Evidently John Tortorella isn’t interested in maximizing ping pong balls in the draft lottery, leaning heavily on his best goalie.
 
 
24) Montreal Canadiens, (-$10):
 

The Montreal Canadiens ranked #2 in my Q1 Power Rankings, dropping to #10 in Q2. Two weeks into the third quarter, they fell all the way to #24. What happened? There were 3 big upsets against the Rangers, Jets, and Maple Leafs when the Habs were +235, +150, and +230 dogs that cost me nearly -$1,500. Goaltender Sam Montembeault shouldered most of the blame, sporting a .961 SV% in those 3 games, out-duelling a pair of former Vezina trophy winners. All this inspired me to bet Montreal when they had easier opponents, but they lost to Detroit and Ottawa twice. I tried to tail the Montembeault hot streak just as it came to an end.
 
The good news was breaking even on the rest of their third quarter. They were actually a good team, winning 50% but were frequently priced as massive underdogs. My line value algorithms were frequently encouraging me to bet the Habs, and it would have been smart on my part to take that advice more often. Had you bet $100 on every Montreal moneyline in Q3, you would have banked $702 (that would be $231 on the puckline +1.5 goals, and $610 -1.5 goals). They began soliciting more of my action towards the end of the quarter, but I’m nervous about their remaining schedule after this Eastern Conference arms race to the deadline.
 
The aforementioned Montembeault started 12 of 18 games with a .910 SV%. Jake Allen had an identical winning percentage (50%) with a .906 SV%, so there wasn’t much difference based on which goalie would be playing. I posted a net loss on their over/under, but only because of Jake Allen.  Montreal unders went 13-7-2 in Q2, but it swung back towards the overs in Q3, which went 9-7. What’s strange about that is that midway through Q3, the Habs lost their best goal scorer Cole Caufield to injury, yet continued delivering overs.
 
 
25) Vegas Golden Knights, (-$348):
 

The Vegas Golden Knights were dealt a devastating blow in the third quarter as Mark Stone was lost to another back injury. They had won 4 of 5 games at the time of the injury, and proceeded to lose 7 of their next 8. He’s such a key component that they were unable to make the playoffs without him last season, and once again find themselves in danger of falling out of the playoff race. They were later dealt another blow when goaltender Logan Thompson was lost week-to-week, but fortunately Adin Hill was able to effectively fill that role in his absence.
 
Adin Hill was actually outplayed Thompson, as both started 7 games but Hill went 5-2 with a .925 SV% while Thompson went 2-5 with a .916. Hard to say that Hill getting the net is what turned them around, but that was the moment that ended their skid. That good news for Knights fans was actually bad news for me, as I was succeeding when betting Thompson to lose in the absence of Stone, and only pressed my foot down harder on the gas pedal after the Thompason injury.
 
Who was the starting goalie had no impact on their over/under, as all three had the same 2:1 ratio of under to over. Laurent Brossoit returned from injury, starting 3 games, losing 2, but posting a sensational .936 SV%. It helped their unders that the goal scoring remained subpar, as it actually remained mostly unchanged from Q2 when Stone was making a contribution. They added Ivan Barbashev near the end of the quarter, who is a useful role player but no substitute for Stone.


26) Winnipeg Jets, (-$380):
 

The Winnipeg Jets were a difficult team to figure out in the third quarter, after winning 65% in Q2. They had a really easy schedule for their first dozen Q3 games, facing only 2 opponents sitting in a playoff spot, going 5-5 against those lesser adversaries. They had wins against the Rangers, Sabres, Kraken with losses against the Canadiens, Blue Jackets, Predators, and Flyers (though, they had substantially more losses to non-playoff teams than wins against playoff teams). The biggest difference between Q2 and Q3 was diminished proficiency on the road, where they have excelled in past seasons.
 
Difficult as it was to get a handle on their wins and losses, there was one very profitable angle to bet the Jets, and that was unders (which went 13-9-1 in Q2 and 15-3-2 in Q3). Their goals against remained virtually unchanged thanks to Vezina-level goaltending from Hellebuyck, but their goal scoring did drop by a goal per game, increasing the rate of return on an already profitable under wager. I had a 12-game stretch betting every under, banking $728 without a single losing wager (pushed twice). I just lost -$617 on the other 8 games.
 
In past seasons we’ve seen the Jets post a good return +1.5 goals (in part because of Hellebuyck), but that was far from the case in Q3. Had you laid down $100 on every Jets opponent to cover -1.5, you banked $1,080. Whereas I did not lay a penny on that wager, posting an embarrassingly large loss whether picking them to win or lose. Back-up Dave Rittich continued his solid play, posting a .923 SV%, while Helley saw his SV% drop from .923 to .912, his win-loss record from 12-5 to 6-9.
 

27) Chicago Blackhawks, (-$553):
 

The Chicago Blackhawks were my worst team to bet in the first quarter, my 2nd best in Q2, then back to the bottom in Q3; keeping in mind 83% of my Chicago stake this season has been wagered on their opponents. There were 3 games specifically in Q3 that inflicted most of the damage, upset wins against Colorado, Buffalo, and Philadelphia. Making a large wager on Colorado was defensible with Patrick Kane out of the line-up. But I have nobody to blame but myself for the other two. Buffalo was on a back-to-back and I got too excited about a Flyers mini-hot streak.
 
That was the moment my foot slammed on the brakes and I started betting Chicago to win more often. It would probably surprise you to know that no NHL team was more profitable than Chicago on the moneyline or both pucklines. That’s right, the tanking Hawks were the best team in all 3 categories, sweeping the majors. In total they won 58% of their games, up from 19% in Q2. Their goals scored per game climbed from 2 to 3.2 while their goals against dropped from 3.9 to 3.4. The Hawks spent their first 10 games of Q3 alternating back and forth from over to under (with one exception), but their overs eventually finished on top.
What’s even crazier is that rookie Jaxson Stauber played the first 6 games of his NHL career, going 5-1 and was the most profitable goalie league-wide in Q3. If you bet $100 on him to win all his starts on the moneyline, you banked $1,140, with an astounding $1,500 on the puckline -1.5 goals (including +750 vs Calgary, +500 vs St. Louis, +550 vs Toronto). Petr Mrazek was also decent, posting a .904 SV% in 13 starts (which wasn’t good for me, ranking as my 3rd worst goalie of the quarter.
 
 
28) Calgary Flames, (-$560):
 

The Flames disappointing season traveled on an unaltered trajectory from Q2 through Q3, though my productivity betting their opponents did take a hit. The primary source of the Flames struggles was goaltender Jacob Markstrom (.869 Q3 SV%), who declined into a timeshare with back-up Dan Vladar. Granted, Vladar wasn’t much better (.878 SV%) and I’m sure Sutter would have gladly made him the primary starter if it led to more wins. Vladar might have been porous, but he still produced a 6-3 record while Markstrom was 2-8. My performance was outstanding betting Markstrom to lose, and terrible betting Vladar to lose.
 
GM Brad Treliving was widely praised for his off-season moves and making the most of a bad situation, but it turns out he actually created a bad situation. If you are someone who bets NHL games using an expected goals model, then you have surely burned a ton of money on this dumpster fire. That seems to be the primary reason that Calgary lines have been inconsistent with their actual win-loss record. They are favored far more than they should be, and the line always moves in their direction from open to close like they are taking public money. In 60 games, they have “taken money” 45 times and their opponents 10 times. Meanwhile, they only won 45% of their games so far this season.
 
Another thing that carried over from Q2 was a weak performance by me betting Flames over/under. Their unders went 13-10 in Q2, but that reversed in Q3 thanks in part to the goalie struggles (with overs going 12-6-1). For whatever reason, my algorithms struggled with Calgary, despite one side being the clear winner. 81% of my money was invested in overs, yet somehow I only managed $89 of profit betting that outcome. Vladar overs were a better wager, I’m assuming because he got significantly better goal support.
 
 
29) Los Angeles Kings, (-$727):
 

The Los Angeles Kings played a road-heavy schedule in the third quarter, which did not hurt their performance. Their visiting win % did drop slightly from 60% in Q2 to 45% in Q3, but their home ice performance climbed from 50% to 67%, offsetting the other side. This was my second consecutive quarter losing money both when betting Kings to win and lose, which is stirring up some painful repressed memories from last season. 75% of my total money invested was on Kings to win, which was far too bullish on my part, costing me -$622.
 
Using “line value” algorithms for LA bet selection was not a winning strategy. The amount of money that I was laying on Kings to win did increase from the previous quarter, shortly after adding Pheonix Copley to my fantasy team. Copley continued to be their best goaltender, posting a .900 SV% compared to Jonathan Quick’s abysmal .841. I very nearly broke even on Copley’s starts, with most of my -$363 loss coming in Quick’s 5 starts. It’s a terrible feeling when you already bet the Kings and you find out Quick got the nod.
 
Their overs went 12-9-1 in Q1, 9-9-4 in Q2, and 12-5-0 in Q3. Their effectiveness at hitting those overs did help me significantly improve my own results on those wagers. Not once in their 17 games did my algorithm recommend an under, producing a nice return on their overs. Their goals for/against per game has barely changed from Q1 to Q2 to Q3, as they have managed to stay very consistent in that regard. I’m up $1,057 on their over/under this season, which is the only thing saving me from repeating last season’s chaos.
 
 
30) Detroit Red Wings, (-$759):
 

The Red Wings were a difficult team to figure out in Q3, mostly because they were able to reverse a previously downward trajectory and win a few more hockey games. They had big upset wins against Winnipeg, Toronto, Vegas, and Calgary, but also lost games to Columbus, Arizona, and Philadelphia. Able to beat good teams while vulnerable against bad teams is a dangerous combination that often causes me problems. When the team caught fire in the second half of Q3, I was able to abandon my short position, allowing me to profit from their wins and reclaim some of the money lost earlier in the quarter.
 
The Wings were among my worst over/under teams in Q2, and while my performance in that area did improve slightly, I still finished Q3 with an O/U loss.  There was an even 10-10 split in their over/unders, and my algorithm struggled to pick the right side. Their overs went 4-1 at the start of the quarter, then unders went 4-1, then overs went 5-1, then unders went 4-0. So basically they were flipping back-and-forth as to which side was best to bet, which is very problematic if you’re using the last 5-8 games to make your picks.
 
Ville Husso proved to be one of the best goalies to bet in Q3, finishing with a .912 SV%. The team game up 4 or more goals in 6 of his 17 starts, but only 1 goal in 5 starts. While the overall picture still looks good, there was some inconsistency for those of us yearning for predictability. The large majority of my money lost was on Ville Husso overs, as their high scoring games were often followed by low scoring games. Back-up Magnus Hellberg only started 3 games, posting a .913 SV%, his overs going 2-1.
 
 
31) Nashville Predators, (-$2,011):
 

Every season I have at least one enigma team that no matter what choice I make, it’s wrong. There have been so many instances where the Preds made me feel stupid for making whatever selection. Bad when I picked them to win, and good when picking them to lose. If my little experiment didn’t require a wager on every single game, I’d simply stop betting Predator games. That would be the optimal strategy. I faced a similar problem with Los Angeles last season, which at one point devolved into flipping a coin to make my picks, which actually worsened my rate of return.
 
The fate of this team tends to rest on the goaltending, which was a little more erratic in Q3. Saros started 14 of 18 games, posting a .905 SV% (down from .934 in Q2), winning 8 and losing 6. For context, Saros went 1-3 when I bet Nashville to win, and 7-3 when I bet them to lose. Saros is on both my fantasy teams, so I may just bet their opponent every game for the rest of the schedule attempting to reverse jinx. 78% of my stake was invested in Predator opponents, leading to a -$263 loss. Nashville unders went 11-8 in Q2 when Saros was on fire, but their overs went 10-6-2 in Q3. My algorithm recommended too many unders, but still managed a profit on both sides.
 
One unexpected thing that happened at the end of Q3 was the resignation of GM David Poile, as the team began selling off pieces that looks like the beginning of a full rebuild.  They weren’t just selling rentals, they were selling assets with years of team control remaining, like Tanner Jeannot and Mattias Ekholm. That doesn’t bode well for their Q4 outlook.
 
 
32) New York Rangers, (-$2,120):
 

The New York Rangers won 63% of their games in the 3rd quarter (up from 57% in Q2) and 66% of my stake was invested in that outcome, yet somehow I managed to lose -$664 on those wagers. Granted, -$750 of that came from a single game when they failed to beat a tired Montreal team where Sam Montembeault stopped 38 of 39 shots outduelling Igor Shesterkin. That was a bad beat, although my line value algorithms had tried to warn me that the price was far too expensive, which was ignored. That one game represented the bulk of my losses.
 
Part of the problem was that their line prices tended to be expensive, so while they won a lot of games, the payouts were low, relatively speaking. Furthermore, 8 of their 19 games featured a rested team facing an opponent who played yesterday. The tired team won 7 of those games, and those were 7 losing bets for me. The Rangers were major violators of the “law of back-to-backs” whether they had the advantage or disadvantage.
 
One of their issues was the uncharacteristically poor play from their Vezina winning goaltender. Igor’s SV% dropped from .916 in Q2 to .894 in Q3, while back-up Jaroslav Halak posted a superior .911. Shesterkin overs went 8-5 while Halak unders went 4-2. Most of my money was lost with Shesterkin in goal, nearly breaking even on Halak’s 6 starts. With 2 weeks remaining in Q3, the Rangers added Vladimir Tarasenko, and a short while later added Patrick Kane. Their line prices were already expensive prior to those moves, so it’s going to be nearly impossible to get any line value on Rangers wins in the fourth quarter.

Posted by at

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)