Last night, I imagine that we all ecstatically saw the Maple Leafs take their second win of the four they need against the Tampa Bay Lightning. The 2-1 series lead is critical, according to Hockey-Reference, teams who go up 2-1 in a series while away have a 70.9% chance of winning the series, compared to 55.4% after they lose game 2 at home.
Moneypuck has actual team-based odds, based on the performance so far this season and throughout the playoffs. The Leafs currently have a 77.8% chance of defeating the Lightning and moving on to the second round. Toronto also sits atop the charts for likelihood of winning the Cup right now, at 15.9%, slightly ahead of the Avalanche, who are 2-0 up on the Predators so far.
Clearly, the odds machine at Moneypuck sees the Leafs as a team that is playing well, giving them a ~7% higher chance to win the series than historical odds would say they should have. Let’s look into the numbers in the series so far to see if the excellent regular season performance that is driving those odds is carrying into the playoffs.
Some might say that 5-0 is the only stat we need from this game, but even a dominant win like that can show chinks in the armor. If you look at just shot attempts, which is ill advised in a single game, especially with such a lopsided score, but just for context, the Leafs gave up 44 shot attempts at 5-on-5 and took just 33. That’s a 42.86 CF%, which is in the realm of very bad. But, we can apply an adjustment based on the score state of the game. It’s easier to get shots when you’re losing, from a statistics perspective, so shot attempts that the Lightning took while down 2 or 3 goals are worth less than the ones the Leafs took when the score was tied.
Applying that adjustment, using numbers from Natural Stat Trick, the Leafs had a 47.06 CF%, a significant improvement, but still not good.
Looking at the Expected goals instead, which is just shot attempts with a factor added for where those shots were taken from, the Leafs really improve. From Natural Stat Trick, again, the Leafs had a 56.73 xGF% at 5-on-5, including the score and venue adjustment.
Overall, it seems that the Leafs ran up the score, and while they took chances from good spots, they did allow the Lightning to run the game in the third period. While it didn’t hurt them in this game, we’ve certainly seen that cost them games in the past.
This one was obviously a tough loss to swallow after the high of the Game 1 win. In this one, the Leafs were pretty flat footed all throughout, and that showed in the score and the shot attempts numbers.
Score and venue adjusted, they had a 47.06 CF% and a 41.96 xGF%, per Natural Stat Trick. Interestingly, the third period told a different story than Game 1; the Lightning kept the pressure on and held 66.23% of the shot attempts despite having the 3-1 lead going in and going up 5-1 early in the 3rd. They did allow two at the very end obviously, to make the final score 5-3 and splitting the goals in the 3rd, but they were able to keep the pressure on overall, unlike the Leafs in Game 1.
This loss was pretty well deserved for Toronto, and they were left licking their wounds in the majority of the third. The consolation of coming within 2 goals likely didn’t do much consoling, nor should it.
To the Leafs’ credit, they made a statement in Game 3 last night: they weren’t going to let Game 2 defeat them, mentally. While the score for this one looked like a blowout, the Leafs scored 2 of their 5 goals with the Lightning having the goalie pulled, so with goalies in net, it was a 3-2 game.
The numbers, again from Natural Stat Trick, tell a similar story to game 1, still, with the Leafs having the disadvantage in 5-on-5 score-and-venue-adjusted CF%, at 47.04%, and again, in xGF%, they had the advantage at 53.04%.
The win in this one was great, and Jack Campbell was stellar as well, fighting off all these shot attempts by the Lightning.
It seems that when the Leafs are winning they really let the shot attempts rack up against them. A dangerous practice, to be sure.
This chart from Natural Stat Trick shows how the shot attempts looked in Game 2 and then Game 3:
This paints a slightly concerning picture for me. There’s a tale of two different Maple Leafs teams. First, the one we saw in Game 1 and 3 that got to strong leads and let the Lightning put a strong push in the third period. Then, there’s the one we saw in Game 2 that got their butts kicked a little bit, and left with a very late 2-goal consolation, but didn’t push the pace the same way the Lightning did when they were losing.
A week ago, in my last Staturday column, I compared a number of regular season stats between the two teams. The overall picture was that the Leafs should have had the advantage in this series. However, the playoffs starting woke up the 2-time Championship winning version of the Tampa Bay Lightning that we were expecting.
Overall in the series, the Lightning do have a stronghold in score and venue adjusted shot attempts, 53% to 47%, while the Leafs have a very slight advantage in expected goals, 50.4% to 49.6%. That said, a 3-game sample is too small to be reliably predictive of an outcome.
As such, I put more stock into the Moneypuck algorithm, that has the Leafs with a good chance to win the series, which again is about 78%. The two wins is obviously part of that, but their Expected Goals algorithm has the Leafs with a bit stronger a handle on the series, at 51.53%. Combining that with the 55.47% they showed in the regular season in the same model, the Moneypuck prediction explains itself.
Going forward, I have confidence the Leafs will win this thing. What about you?