2023/24 Week 18 Betting Report

Week 18 of the NHL season has been logged into the history books, as the hockey world begins setting its sights on the trade deadline and all the ensuing roster volatility. As we brace ourselves for a transfer of talent from the bottom half of the standings to the top, I’m busy retrofitting all my models to the 2021/22 and 2022/23 seasons to prepare for a strong fourth quarter and ultimately playoffs. Right now, just betting the biggest position my models are collectively betting seems to be better than any one model. We’ll see though, as building a more intelligent aggregators was a big factor compelling me to undertake this renovation. The biggest position may not always be optimal.
 
Before we go any further, it’s time for my obligatory *DISCLAIMER* it needs be noted that I’m not betting with real money. These are all fictional wagers in a spreadsheet. My mission is to engage in a mass betting campaign, picking a winner of every single game, every over/under, because it provides a complete dataset for macroeconomic analysis, which can be shared with you, shedding light on what worked and what failed. I’m also tracking the results of betting every outcome, to help me (and you) uncover previously unknown or newly emerging profit vectors. What started as a thought experiment has evolved into much more.
 
My blog has been moved to Substack this season and I’m repeatedly encouraging everyone to sign-up for a free subscription to alleviate my dependence on Twitter for traffic. I’m concerned that Elon will follow through on his threat to charge everyone for Twitter and ostensibly destroy his own company. Subscribers receive an email notification each time a new post is published, and even if Twitter stays free, the algorithm likes to hide Tweets with links so you don’t leave. Subscribers will also receive weekly picks emails that are not posted on my blog. If you like what you see in the picks below, you can have many more delivered directly to your inbox.
 
My Weekly Profit: $338
My Season Profit: $7,104
 
My Sunday Preview outlined a major project that dominated my weekend and still requires more work. Roughly 2/3 of my models were already fit to the 2 previous seasons (recording wagers in each), and once the whole team has a big historical database of picks, the applications are exciting. First and foremost, I’ll be able to fit new versions of Max Profit and the Grand Aggregator with more complex decision making, optimized to a much larger sample. The models that were not optimized to past years were fit to much smaller samples, which made them unstable and more prone to wild swings.
 
My Shorting Goalies model is one requiring some grunt work to retrofit, but it has been performing very well in its current version thus far in the 3rd quarter. Retrofitting that model required a projected probability of either goalie starting and obviously travelling backwards in time to make predictions before those games was played is off the table for at least a few more years. However, assigning a probability of either goalie starting based on what happened in the last 10 games was easy. This involved checking hundreds of permutations spread across 3,500 games, so saving it as a permanent model to use for projecting starters is probably not worth the effort in its current form. My brain is very effective at projecting starters (with some exceptions).
 
The Goalies Last 30 days retrofit is finished and the model is now online, along with Tailing History 3.0. Tails was just retrofit to the 2023/24 season 3 weeks ago, but it would have been the same amount of work to refit including past seasons, and I’m mad at myself for not thinking to do that originally. Version 2.0 was only using the previous 2 seasons (to exclude 2021), which meant that my historical database was big enough to fit this to 2021/22 and 2022/23. Anyway, Tails 2.0 was already one of the top models in the third quarter, but Tails 3.0 is even better (currently first place in my Q3 rankings). Both versions bet the same side in most games, but 3.0 has better fine tuning.
 
 
My Team of the Week: New York Rangers, +$828
 
Last week the New York Rangers were dead last in my Profitability Rankings, but they pulled themselves out of the basement going undefeated last week. Though it should be noted, it took overcoming a 2-goal deficit in the 3rd period against the Islanders Sunday that made all this possible (they were #2 in my weekly ranks heading into Sunday). That bet looked dead in the water until I Tweeted “Damned Rangers…” with a pensive emoji that I’m reasonably convinced reverse jinxed them to victory. The same trick helped the Avs overcome a deficit against the Coyotes later that same day (and Edmonton as I’m finishing this post).
 
This was one of my better weeks betting Washington games after a horrific 2nd quarter. The main contributor was going 2-0 on their overs, as their goaltending has struggled while the offense has picked up some steam. Ovechkin is starting to score again, which is putting him back on track chasing 99. Darcy Kuemper has been outperforming Charlie Lindgren, but neither have been good. They lost 6-3 to Colorado then beat Montreal 4-3. Even with the improved offense, they’ve only won 2 of their last 10 games, so I’m really only betting them to win against worse teams.
 
 
My Worst Team of the Week: Anaheim Ducks, -$700
 
The Anaheim Ducks were my biggest spoiler in the last 7 days, picking the wrong outcome on all but one of my wagers (the over in their 9-2 loss to Toronto). They won when I bet them to lose and lost badly when I bet them +1.5 goals. They lost 5-0 to Montreal, then beat Ottawa 5-1, then lost to the Leafs 9-2. Where the loss against Montreal convinced me to pick the Senators, then the win against Ottawa convinced me to take them +1.5 against Toronto and they were demolished. It was a vicious cycle. They have the ability to punch above their weight when the going is good, but in some match-ups have a week chin leading to easy knockouts.
 
The LA Kings were my second worst team of the week, with most of the damage coming from over/under, mostly a max bet on under 6 goals when they lost 7-0 to Buffalo. There was also my Sunday over 6.5 when Cam Talbot was starting, but Cam had his first good game in a long time, his first win since Dec 23, his first time allowing 1 goal or less since Dec 7. Losing the Penguins moneyline bet with the Kings on a back-to-back and Talbot starting felt like slam dunk, but there are no sure things in hockey (well except for Colorado at home beating teams on a back-to-back, which we’ll get to later).
 

My Week 18 Results

 
*Note* “Overall Market Bets” based on betting exactly $100 on every outcome.
 
 

My decision to bet more underdogs +1.5 goals after a good performance last week was exactly the wrong thing to do, at least initially, going 0-3 Monday and Tuesday with picks shared in my reports before hitting my next 3 in a row. Still, road teams and favorites -1.5 goals were big winners (if you bet every visitor -1.5 goals (even dogs) you banked more than $1,500). If you like betting against tired teams with a rest disadvantage, you surely lost money this past week, since teams with a rest advantage went 0-7 against opponents who played yesterday. Fortunately, there were fewer opportunities than usual to short back-to-backs, otherwise your losses could have been much worse. I’m hoping that’s temporary variance and not an emerging trend, because my models love shorting btb.
 
Last week saw unders perform very well as leaguewide save percentage spiked up to .910 the week after the all-star break, which was just temporary as it was .895 in the last 7 days (which ranked as the 3rd worst in 18 weeks). Several of my algorithms had a good week, which is a little surprising when there was a shift in the scoring trend. When the total was 6.5 on the opening line, overs went 13-9. When the total was 6, unders went 8-7-4. My performance on under 6.5 was bad, but still generated $360 profit betting over/under this week, following my Small Council of algorithms in every game but one (which they were right and I was wrong, the Talbot over).
 

Edmonton overs were back with a vengeance after an extended hot streak by the unders, as their penalty killing recently transitioned from elite to awful. The best team to bet against overall was the Calgary Flames (for me as well), as their little post-all-star mini-hot streak after the Lindholm trade appears to be over. Jacob Markstrom was human in his last 2 starts, getting yanked against Detroit. The Panthers were the best team to bet on, followed by the San Jose Sharks. Granted, they’re only there because they beat Calgary 6-3 as a +320 dog, hitting the alt puckline at +665. They lost their other games, so don’t look at that and start aggressively betting the Sharks.
 
 

Team By Team Profitability Rankings

These profitability rankings are based on the sum of all my bets per team, including where the money was won or lost. Each week my new Profitability Rankings will be based on all the games in the season, not just what happened this week.
 

Me vs Myself

 
The “Me vs Myself” section outlines my competition against my betting models, in my vain attempt to prove my own decision making is superior to the models that I’ve created. Me vs my creations. If you are new here and don’t know how the rest of the models work, there’s a link for that. Typically, this is where I would share my model performance from that last week alongside the quarterly standings, however due to the ongoing remodeling, not all the current numbers are from the new versions of each model, and I’m also planning an overhaul of Max Profit and the aggregators once I’m done fitting the others to past seasons.
 
This was a bad week for my Hedge Fund that bets against line value, but I’m not worried about that because two of them received upgrades this weekend and the composite model is going to be optimized to the historical data after the retrofit is complete (Grand Aggregator too). Instead of just betting the largest position proportional to the amount, there can be specific instructions for different scenarios. So, if >60% of the money is invested in HML and 30% is VML, it might make a different pick than if it’s 60% HML and 30% H-1.5. It was on version 1.0, but once there is a large historical sample of games where all the fund members have recorded picks, then it’s easier to identify the most profitable wagers.
 
The two models leading my 3rd quarter tournament are Tailing History 3.0 and Goalies Last 30 Days 2.0, both recently retrofit. Shorting Goalies 1.0 is in 4th place behind the Grand Aggregator, but is still scheduled for a retrofit this afternoon. Shorting Travel and Game Sum 2.0 both had bad weeks and fell off the leaderboard (which I’m defining as the top 9 models who share their bets in my pick graphics. You’ll notice the order changing each time I do a new report, as I’m recalibrating constantly to have the best at the top.
 
 

Tomorrow’s Picks

 
Monday was busier than normal, leading to a quieter Tuesday, but at least there are a few intriguing match-ups (which doesn’t necessarily mean lucrative to bet, but entertaining to follow). My models had an exceptionally good Saturday as a collective, if you picked based on % they invested in each outcome. The good news is all those picks were shared late Friday night if you follow me on Twitter (as I assume most of you do). My own week was going poorly until Saturday, mostly tailing my top models. Goalies Last 30 Days 2.0 is now officially automated.
 
 
NJ @ WSH:
 

My models love New Jersey -1.5 goals (soliciting 72% of their total investment), and goalie Nico Daws has delivered 4 quality starts in a row, winning 3 of them, all by at least 2 goals. I’m going to tail on that Devils puckline at +164. There’s a chance this could be Akira Schmid or possibly Vanecek, but I think they’ll ride the hot hand against a divisional opponent battling with them for a wildcard (they’re only 4 PTS up on Washington). I’m also comfortably taking over 6.5 goals, getting a nice price at +110. That’s at least a little insurance to one of the other goalies starting.
 
 
OTT @ FLA:
 

The Florida Panthers might be the best team in the NHL and the Ottawa Senators are playing Tampa today. The Sens are scrappy, and have the talent to upset any given team any given night, but since my models are overwhelmingly on Florida 48% ML 37% PL, I’m making a max bet split between the Panthers ML -230 (60%) and the PL -105 (40%). The Sens ability to punch above their weight is why I’m not going all in PL. I’m also going to make a max bet on under 6.5 goals and am getting a nice price at +105.
 
 
DAL @ NYR:
 

The Dallas Stars played Boston today and thus it will be Scott Wedgewood in goal tomorrow. That alone was enough to put me on New York, bolstered by my models taking a 57% position on Rangers ML at -142. Dallas is still one of the best teams in the league, so I’m only making a minimum wager. There was a 3-2 vote on under 6.5 goals, but I’m very uncomfortable with that recommendation. Might be a stay away. Both these teams can score.
 
 
NYI @ PIT:
 

This game was a candidate to skip, given my models are split down the middle and these are two teams I’ve struggled to figure out. NYI ML at +120 was ultimately beat out Pens ML as the largest position 23% to 22%. My two best models like the Penguins, but the next 7 are taking the Isles. I might do better flipping a coin to decide my wager, but am going to take the Isles ML because there’s a higher payout. Under 6 goals passed by a 3-2 vote, but the two taking over are performing poorly on that bet with these teams last 30 days (12 games had less than 6 goals, while only 6 games had more.).
 
 
MIN @ WPG:
 

Both these teams play today, so this will be Marc-Andre Fleury vs Laurent Brossoit. The largest model position is the Minnesota ML at 36%, but 64% of the total money is on Winnipeg (when I build a smarter version of the Grand Aggregator this week (hopefully), situations like this will be accounted for). I’m just putting a minimum wager on the Wild ML +136 because Fleury has been excellent in the last 6 weeks, much better than Brossoit (.952 SV% last 30 days vs .902). Under 6 goals only passed by a 3-2 vote, but the two dissenting voices are getting destroyed betting over 6.4 with these teams last 30 days. These teams combined to play 17 games in the last 30 days, 12 went under 6, only 3 went over 6, with 2 pushes. Give me the under, as both back-up goalies are above average.
 
VAN @ COL:
 

This puts me in a difficult position as both teams are high on my list of favorites to pick, but the fact that Vancouver just played Minnesota (who has some big bodies who like to throw their weight around) and are flying up to Denver tonight is a big selling point. The Canucks are a very good team, but I’m going to take the Avs ML at -148, with a large wager considering the Canucks are first place in the NHL. Why? Colorado is 13-1 with a back-to-back advantage on home ice since October 2019. Heading up to altitude when you’re already gassed is bad for recovery. Under 6.5 goals passed by a 3-2 vote, but those taking the under are losing money on that wager with these teams last 30 days. Canucks just lost 10-7, and that data was not used in the computation above. Maybe stay away from that one.
 
 
NSH @ VEG:
 

While I’ll confess to a crisis of confidence in the Nashville Predators lately, I love betting against home teams on back-to-backs who flew home last night. Maybe that’s less of an advantage if the Preds spent the night in Vegas enjoying the nightlife (if I have any subscribers who live in Vegas and plugged into the high-end party scene, I’d love tips when specific teams are in the process of catching the Vegas flu). My models are 56% on the Nashville moneyline +124, so that’s my pick as well. Though if this turns out to be Kevin Lankinen against Adin Hill, my confidence declines. But that would also make me feel better about over 5.5 goals. This should have been a 5-0 vote, but Game Sum did not have a sample of games with a 5.5 total involving either team last 30 days (it was programmed with ties bet under). That might also mean that the line should not be 5.5. I expect this to reach 6 by puck drop (I actually just checked and it already moved to 6).
 
 
CBJ @ LA:
 

My initial thought looking at this game was the Kings getting home from a 6-game road trip, and are among the worst teams in the league in this situation since October 2019. Yet Shorting Travel is abstaining because it doesn’t bet road longshots. No sense me tailing that angle when that specific model won’t even follow. I’ll put a small wager on Kings ML -258, even though that feels like a crazy price for a team that had lost 14 of 17 games before the all-star break. At least they are 4-1 since the break under the new coach. Also give me under 6.5 goals, which only passed with a 3-2 vote.