Welcome to my second round preview, which is taking on a different form than originally intended since R2 will begin before R1 ends. Even waiting until late Saturday night when Leafs-Bruins wrap doesn’t leave much time between that and R2’s Sunday afternoon start. So, I’m posting this “preview” Saturday before the Toronto-Boston winner is determined, with my picks for the other series coming on social media (both get mentioned here, but there’s no lines to lock in). We aren’t getting too in-depth into past round 2 stats because we just saw a much different round one than previous playoffs (from a category standpoint) that threw curveballs at some of my models, so I’m skeptical there is replicability in that data. Tailing History bets exactly what was previously profitable and just got rocked. More on my models below.
Before we go any further, it’s time for my regularly scheduled 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.
Step one when creating this preview was reviewing my performance from previous playoffs in the 2nd round (of which I have 3 previous seasons of usable data), and reviewing my 2023 round 2 (aka R2) preview for any usable observations. My R2 performance in 2021 and 2022 was good, producing roughly $1,000 profit, split relatively evenly between favorites and underdogs. Then disaster struck in 2023. My series picks were Dallas, Toronto, New Jersey, and Edmonton, going 1 for 4 (all those teams favored to win). On the individual games, I lost nearly $2,000, and it’s not a coincidence that more than 100% of those losses came from favorites, both ML and +1.5 goals.
Plans for a round three preview were cancelled with my confidence rattled. Even reviewing my reporting on those playoffs stirred up repressed memories that have me doubting whether another round two preview might do more harm than good. Then digging a little deeper revealed that $2,000 of my losses came from betting the Toronto Maple Leafs to win. Dating back to 2021, my career playoff profit (not counting over/under) is $62 on $87,000 pretend dollars wagered. That’s probably not comforting to those of you reading this in search of wisdom, but that $62 profit is after a $4,200 loss betting Toronto to win. Take Toronto out of the equation, and my historical performance improves substantially.
Note: the 2 paragraphs above were written when Toronto was down 3-1, which was supposed to transition into the next paragraph (also written at 3-1).
To begin bragging about my ability to manipulate the butterfly effect, ostensibly repressing Leaf playoff success through my picks, feels likely to damage my intellectual credibility as a betting analyst. All I’m saying is that compelling evidence exists that I possess legitimate reverse jinxing powers. Those of you following my work longer have certainly witnessed some of those powers first hand. It’s also possible they would have choked in all those playoff games without my assistance, but that’s much less fun than believing you have magical powers to curse your least favorite team. If there is no Toronto in the 2nd round, we do not need to worry about that potential hazard.
Okay, so new compelling evidence is emerging that whatever reverse jinxing powers I may have had are weakening. The epiphany dawned on me last year vs Tampa, but then Florida’s dominance in the next round eased those concerns. Leaving those paragraphs written a week ago in this report also serves the function of transitioning into my pick for 2024. If Boston wins, I’m taking Florida in 5 games. If Toronto wins, I’m picking Leafs in 7 games. If that logic does not compute, well it’s not entirely logically based. No math went into determining those 2 choices, when checking how my models would bet game one of either opponent was an available option, deemed unnecessary (my pick for game one will be posted on social media).
For the first 4 seasons of my betting project, I mostly just published reports outlining what worked after the fact. This was actually my first season regularly giving out picks prior to puck drop, but instead of only sharing my favorite games (like most analysts), I’m frequently detailing the entire slate in a night. Please, I’d love to know if ANYONE else put out nearly as many picks as me this season that’s not behind a pay wall, and was anywhere near my rate of return. This is really hard to do. I’m playing with pretend money, and a few people have criticized me in the past saying my picks lacked legitimacy. But once I started frequently sharing and hitting bets, none of my subscribers ever challenged that validity (despite a few flagrant reverse jinx attempts).
If you read the above justification jinxtification then pick Toronto to beat Florida and Panthers sweep, you have nobody to blame but yourself. My Leaf biases are frequently disclaimed on this blog. Now that there are hundreds (possibly even thousands) of dollars being wagered based on my recommendations, with great power comes great responsibility, one I take very seriously. My explanation for picking Toronto to beat Boston was basically the definition of the “gambler’s fallacy”, but given how many people voted for Toronto in my “which series pick would you rather bet” Twitter poll, perhaps your ability to properly conceptualize that juxtaposition jinxtaposition was lacking. I invented 2 new words today.
Past Playoff Performance
Anyway, let’s get back on track previewing the upcoming round two. A good place to start is reviewing what categories were best to bet in the previous 3 playoffs.
In the previous 3 playoffs, the best bet was home teams -1.5 goals if they were between -150 or +150 on the moneyline (that’s my definition of “mild”). Sadly, that demographic was just dummied by round one, as only the big favorites consistently delivered on the puckline (and there won’t be as many “big” favorites in the next round because only the strong survive). Granted, I’m not precisely targeting bets on the basis of how categories are performing, instead looking where my model money is distributed. Problem is, they’re all programmed on what worked previous seasons, so if what “works” flips, they’re going to be giving bad advice (more on models coming).
Looking at my model round one performance, H-1.5 mild favorites and home dogs -1.5 cost them more than $15,000 (at least prior to the last 2 R1 games). The good news is those models who got destroyed on H-1.5 were also demoted to my bench (including 2 of my best models from the 4th quarter). Once Shorting Value and Tailing History dropped out of the top 10 (entirely because of home -1.5) they were dropped from my pick graphics and their feelings no longer shared for public consumption.
Looking at my best and worst teams to bet in the previous playoffs brings back both pleasant and unpleasant memories. My wagon was hitched to the Islanders and Canadiens in 2021, my success story betting underdogs +1.5 goals in my first non-bubble playoffs logging wagers on every game. That blueprint became my strategy early in 2022, posting large losses on the same category that carried me one year earlier (home -1.5 was booming). Montreal was a big dog with the best goalie in the world stealing games. In retrospect, a big dog relying on crazy awesome goaltending to steal games is itself a risky strategy to pursue, and tough on your nerves if you’re going to watch the games.
In round 2 of the previous 2 playoffs, unders went 23-21 despite a small increase in goal scoring, but what’s a good pick varies from series to series. Florida unders went 9-0, meaning overs went 21-12 in round two games not involving the Panthers (more than half of those overs came from a Dallas or Edmonton game). There really is no blanket strategy, that applies to the entire round, it’s something that needs to be handled on a series-by-series basis. This is my first playoffs with advisors, and the Small Council did an effective job navigating round one. The deeper we get into the playoffs, the % of their inputs coming from previous playoff games continuously increases.
The other observation looking at last year’s series picks was going 1-3 betting favorites in round two. That being said, it’s entirely possible that will flip entirely and dogs will sweep the next set. There are some very close match-ups and even deliberating them in terms of who is favored on the betting line might do more harm than good. My only round one series pick based on the line price was Tampa and look how that one played out. As far as I’m concerned, there are no “line value” plays on these series lines. Granted, looking at my write-ups for last season, line price was not a factor in my decision. I picked New Jersey to beat Carolina before the line was up because the Carolina injuries were piling up.
Round One Model Results
As mentioned in an earlier post, I found a big mistake in the primary input variable for Macro Daddy and Value Voodoo midway through round one. Once that was fixed, they both immediately got hot. Once their previous mistakes were corrected, they both proved capable. Neither required upgrades before advancing to round two, since both turned a profit on home teams -1.5 goals. Game Sum 3.0 did not require any changes because it has a completely different set of commands for the next round, and only 15% of the historical picks were home -1.5 goals, all small wagers. The Oracle and Shorting Value 3.0 both lost money, but would have produced a profit if they simply bet HML instead of H-1.5. That change was made to their decision matrices, advancing to round two.
The other 5 were eliminated in round one, and simply flipping H-1.5 for HML was not enough to redeem any of them. They will all continue to log picks, but they won’t show up at all in my pick graphics. I’ll be monitoring their results to see if anyone gets hot, as both the zig zag models have different commands after round one. The same survival of the fittest concept is applied to my original roster of models, with their round one results displayed below (not including the last 2 games).
My pre-playoff roster of models started round one strong before home pucklines crashed, which immediately sank the models who rely on puckline wagers, or were set to tail the models who had been profiting from them leading into playoffs. Home -1.5 came into the playoffs hot, which is specifically how the crash pulled down the Profit Aggregator and Max Profit. Tailing History was looking at their profitability in previous playoffs (and the previous 7 days). Shorting Value thrived on pucklines in the 4th quarter, and nearly all that $4,000 loss came from that wager. We’ll see if that continues though, because now everyone has a week’s worth of data where H-1.5 was terrible, so most should start looking at alternatives soon.
The difference between my models getting eliminated from the playoffs and the real post-season is that anyone can get back in with solid performance. They aren’t sent packing on vacation to Barbados. The over/under results were a little more encouraging, but there are two “Small Council” members down at the bottom of the standings. The Goalies Last 30 Days concept was more for regular season when back-ups start more often. If it’s the same goalie most games, the usefulness is lost. So, the original is out of the Small Council, replaced with Average Goals Last 5 Games (my old primary). While Expected Goals L30 v1 will be replaced by version 2.0 (my best in the 4th quarter). Max Profit 1.0 was replaced with the Profit Aggregator.
From the new playoff over/under models, Prime Optimus was demoted to the bench after losing $1,000. While the new expected goals model proved excellent, producing $1,000 profit. The two others lost around $400 each (average bet size near $200), but the vectors that led to those losses were adjusted to minimize damage going forward. So, I’m hoping that improves their output. That gets us through all the model updates, all of which will be reflected in the pick graphics below. The bad news is, these changes were not made prior to sharing pick graphics on Twitter, so you may notice some difference between here and there.
One important programming note, most of the picks I’ll be sharing going forward will be on social media. There just won’t be enough games each night to justify a whole blog post about them, but I’m still planning to share picks for every single game as promised, regularly shuffling best/worst performing models in/out of the starting line-up. Even the leading scorer from Q4 got benched for whiffing on too many big H-1.5 bets in games 3 and 4. This time of year it’s survival of the fittest and the weak shall perish, both for NHL teams, and my betting models.
Carolina Hurricanes at New York Rangers
Series bet: Carolina -140
Exact bet: Carolina in 7 games +500
The first confirmed 2nd round series is a little deja-view all over again, with the Carolina Hurricanes opening as a -140 favorite to defeat the higher seeded New York Rangers (that has moved to -160). That line did bring back bad memories of betting Carolina -140 last season to beat Florida. While I’m permitting myself to have the freedom to pick either side based on how round one unfolded, I already deliberated the winner of this potential series when filling out my bracket, which had the Canes advancing to the Stanley Cup final. Even though it feels like there’s a little value on NYR +120 (now +140), I’m sticking to my guns (betting Tampa round one as a value play backfired), but am going to pick Canes in 7 games at +500.
The last time Rangers bounced Carolina in the playoffs it was because Igor Shesterkin dominated the series, and Igor did post a .931 SV% against Washington. That’s my biggest concern taking the Canes, and surely creeped into my deliberations when filling out the bracket. The difference between this Canes team and the other is more weapons up front, notably Kuznetsov and Guentzel. Note that I’m making that bold proclamation before looking at the rosters from that series (pause to check) okay I’m back, and yes, Carolina made several improvements on that roster while the Ranger line-up is very similar (and now older). I’m less concerned about history repeating after quickly glancing at that history.
Not sure precisely when the game 1 line went up at Draft Kings, but I first saw it May 1st at -105 and immediately recorded a wager on Canes moneyline before even consulting my models and running any numbers (Tweeting the pick shortly afterwards). You can either interpret that blind confidence as evidence you need to pause your reading to log a Canes bet, or this mother f**king so-called expert carelessly neglected to perform the proper mathematical investigation backing up his pick. Both can be true. My pick was made in haste, but the actual preview containing the pick was not scheduled to be posted until 5-6 days later. Why don’t I stop writing and go ask my models what they think of this series…
And we’re back!
Looking at what the models had to say poured some cold water on my Hurricanes enthusiasm. Thankfully I only logged a minimum wager, and did check with the models prior to Tweeting the pick (which would have been irresponsible). There has since been an adjustment in which models are contributing, but the overall story did not change. Models going heavy on Rangers. The other reason there was zero hesitation making my game one pick is staying consistent with my bracket. So, I do not feel any remorse being locked into a Carolina game one bet, especially since I do feel they are the better team, despite New York having the better record. Once again, I’m taking over 5.5 goals, but with weaker confidence than the weak confidence reported in my Tweet.
Edmonton Oilers at Vancouver Canucks
Series bet: Edmonton -260
Exact bet: Edmonton in 5 games +350
The Edmonton Oilers are a giant -260 favorite to defeat the Vancouver Canucks despite being the lower seed starting on the road. Full marks to Vancouver for hanging on to beat Nashville with their 3rd string goalie, killing my 53 to 1 series parlay. If Thatcher Demko was healthy and playing at a Vezina level, maybe the Canucks would entice me to pick them as a series dog (5 PTS better in the regular season), but he’s not. There was no deliberation when logging my picks for this series which team would be me choice, but picking the number of games did cause pause. I settled on Oilers in 5 at +350 but was tempted by the juice on +500 for a sweep. Oilers in 6 was +300, so I went with 5 to get the extra +50.
That entire opening paragraph was written within 30 minutes of the Canucks defeating Nashville before even running any numbers are checking with my models about game one (pause writing to actually consult models). Okay cool, my models have 2/3 of their money on Edmonton and my top 5 models have 10 times more on that pick than Vancouver. They opened at -142 and that has moved to -135, so it looks like public money is coming in on Vancouver. For me the goaltending situation is still too risky against that mighty offense, but the defensive shell Tocchet has setup to protect the net is convincing enough that I’m laying off the puckline (likely for the whole series unless this goes off the rails). That might be hypocritical given that I’m taking over 6 goals, but let’s see how the rock-paper-scissors of it all plays out.
Vegas Golden Knights at Dallas Stars
The final second round game tomorrow is…oops…nearly forgot round one hasn’t ended yet. The Vegas Golden Knights had a big win at home on Friday night to force game seven on opening night of the next round. The Stars are a -148 favorite to win game seven at home, and my bracket pick to win the Stanley Cup (Vancouver squandered my chance at a perfect bracket), so initially there was little doubt they would be my pick in the final game. At least until checking with my models, with my best performers laying exactly 0% on the Stars. Granted, one game earlier they loved Dallas on the road and I picked Vegas, mostly as a hedge to my Stars in 6 at +450 series bet. It won, and felt smart after the fact (I also noted “do or die at home” would likely bring the best out of Vegas).
The Golden Knights were able to find another level Friday night, notably limiting the Stars high-powered offense to 23 shots. My pick was shared on Twitter not long after the game ended and I’m sticking to that choice. Among the reasons is getting +124 on the defending Stanley Cup champion that may have found their groove. I’ll be cheering for Dallas so my bracket doesn’t look foolish, meaning there is also some emotional hedging happening here. The winner will play Colorado, and if that’s Dallas, Stars are my pick in 7 games. If it’s Vegas vs Colorado, I’m leaning towards Avalanche in 7 because of home ice advantage, but it’s also entirely plausible Vegas causes them big problems. Georgiev collapsing can’t be taken off the table either.
After shuffling my models around, my bet on under 5.5 goals remains exactly the same, but the confidence coming from the other aggregators increased. I wouldn’t be shocked if this closed at 5. Goals might be really tough to come by.
That’s it for now, not enough games for parlays.
Good luck!