As many of you know Tuesday night’s loss vs. the Rangers was infuriating. Another blown multi-goal lead, something that many have observed have become more of a trend since Pete Deboer took over as coach. Of course some responded in defense trotting out the usual it is standard score effects, poor goaltending and poor luck. I am doubtful of this because we have seen this now happen under multiple goalies and observation wise it seems it affects us more than other teams. That got me thinking how do the Devils compare to playing with a lead compared to the rest of the league? With the weekend hitting I took time yesterday to play around on War-on-ice (when it wasn’t down) and see.
My theory in this regard was simple. If the Devils change the way they play with a league more than the average team it would show when comparing the difference between CorsiFor% with a 2-goal lead (CF%+2) and CorsiFor% in close games that was within 1 goal (CF%Close). For this I calculated the drop in Corsi between (CF%Dif) the two situations for each team and then calculated the league average drop as well as the standard deviation to give me a milestone of what represents a noticeable difference. My comparison goes back to the 2011-12 season when Deboer first became coach. In terms of on-ice strength I used all situations since I wanted the effects of getting a PP and going on the PK to be represented. I also ran the numbers prior to last night’s game.
Now for the results. Under Deboer when the score is close the Devils have a CF% of 53.02, 5th best in the league in that situation. When they have a two or more goal lead their CF% drops to 43.82, 11th best in the league in that situations. The difference between the two situations is a 9.20 drop in percentage points, the third largest drop in the league. Now every team sees a drop on their CF% between the two situations. The average drop is 7.06 percentage points. The standard deviation is 1.57 percentage points, meaning most teams are going to be within that percentage point difference of the average. The Devils’ 9.20 percentage point drop is 1.94 percentage points from the mean, more than one standard deviation. That shows to me that the Devils have a noticeable difference in playing with the lead than the rest of the league.
After looking at that I decided to compare the last three years under Deboer to the previous three seasons prior to Deboer becoming coach. To do this I did the same thing I did for seasons 2011/2012-Pres for seasons 2008/09-2010/11. During this time period under Sutter/MacLean/Lemaire, the Devils in close game situations had a CF% of 52.59, 5th best in the league under this time period. In game with a 2+ goal lead they had a CF% of 44.81, dropping them only to 6th best in this situation. Their drop in CF% between the two situations was 7.78 (19th best). The average for this time period was 7.27 drop in percentage points with a standard deviation of 1.66. They were close enough to the average to safely conclude their drop in play during this period was due to standard score effects.
Now to compare the two numbers. For this I adjusted the Devils’ CF%Dif number from 08-11 to make it equivalent to what it would be for the average from 2011-Pres. For that I came up with a 7.55 equivalent drop. To get an understanding of the difference in how much the Devils change their play with a lead pre-Deboer and currently, I took the 9.20 number under Deboer and subtracted 7.55. That results in 1.66, which is greater than the league standard dev. for 2011-pres. That is to say the difference in how they change their play with lead pre-Deboer and under him is noticeable worse.
Now obviously when you are comparing over multiple years rosters change, but considering the team has a both a higher CF% in close situation under Deboer than pre-Deboer and a lower CF% with a 2+ goal lead under deboer than pre-deboer, I don’t think you can pin it on personnel.
Also consider with the team’s scoring woes most of this effect is mostly with only with a 2-goal lead, there are not many 3 or 4 goals leads where they potentially sit back more to skew the numbers. And that 1-goal third period leads are included in the close numbers. It may be a bigger drop off if one were to consider 1st and 2nd period close vs 3rd period with any lead.
In summary, the Devils play with a lead noticeably drops off when they have a multiple goal lead. This drop off is greater than the usual score effects the other teams face. Further, this is something that has only affected the Devils since Pete Deboer took over as coach. Prior to Pete they did not have nearly the level of drop off in their play with a lead. When people complain about how the team plays with a lead under Deboer it is not something that is imaginary. It is a real issue.
Why is this a case? Obviously, this study doesn’t answer. It could be playing a more passive system with a lead and sitting back more, playing worse players with the lead or likely both. When you consider how many leads the have blown, points dropped and points given to playoff competitors, this play with a lead is probably a major contributor to the Devils missing the playoffs in the last two year.
Thank you for reading my book
My theory in this regard was simple. If the Devils change the way they play with a league more than the average team it would show when comparing the difference between CorsiFor% with a 2-goal lead (CF%+2) and CorsiFor% in close games that was within 1 goal (CF%Close). For this I calculated the drop in Corsi between (CF%Dif) the two situations for each team and then calculated the league average drop as well as the standard deviation to give me a milestone of what represents a noticeable difference. My comparison goes back to the 2011-12 season when Deboer first became coach. In terms of on-ice strength I used all situations since I wanted the effects of getting a PP and going on the PK to be represented. I also ran the numbers prior to last night’s game.
Now for the results. Under Deboer when the score is close the Devils have a CF% of 53.02, 5th best in the league in that situation. When they have a two or more goal lead their CF% drops to 43.82, 11th best in the league in that situations. The difference between the two situations is a 9.20 drop in percentage points, the third largest drop in the league. Now every team sees a drop on their CF% between the two situations. The average drop is 7.06 percentage points. The standard deviation is 1.57 percentage points, meaning most teams are going to be within that percentage point difference of the average. The Devils’ 9.20 percentage point drop is 1.94 percentage points from the mean, more than one standard deviation. That shows to me that the Devils have a noticeable difference in playing with the lead than the rest of the league.
After looking at that I decided to compare the last three years under Deboer to the previous three seasons prior to Deboer becoming coach. To do this I did the same thing I did for seasons 2011/2012-Pres for seasons 2008/09-2010/11. During this time period under Sutter/MacLean/Lemaire, the Devils in close game situations had a CF% of 52.59, 5th best in the league under this time period. In game with a 2+ goal lead they had a CF% of 44.81, dropping them only to 6th best in this situation. Their drop in CF% between the two situations was 7.78 (19th best). The average for this time period was 7.27 drop in percentage points with a standard deviation of 1.66. They were close enough to the average to safely conclude their drop in play during this period was due to standard score effects.
Now to compare the two numbers. For this I adjusted the Devils’ CF%Dif number from 08-11 to make it equivalent to what it would be for the average from 2011-Pres. For that I came up with a 7.55 equivalent drop. To get an understanding of the difference in how much the Devils change their play with a lead pre-Deboer and currently, I took the 9.20 number under Deboer and subtracted 7.55. That results in 1.66, which is greater than the league standard dev. for 2011-pres. That is to say the difference in how they change their play with lead pre-Deboer and under him is noticeable worse.
Now obviously when you are comparing over multiple years rosters change, but considering the team has a both a higher CF% in close situation under Deboer than pre-Deboer and a lower CF% with a 2+ goal lead under deboer than pre-deboer, I don’t think you can pin it on personnel.
Also consider with the team’s scoring woes most of this effect is mostly with only with a 2-goal lead, there are not many 3 or 4 goals leads where they potentially sit back more to skew the numbers. And that 1-goal third period leads are included in the close numbers. It may be a bigger drop off if one were to consider 1st and 2nd period close vs 3rd period with any lead.
In summary, the Devils play with a lead noticeably drops off when they have a multiple goal lead. This drop off is greater than the usual score effects the other teams face. Further, this is something that has only affected the Devils since Pete Deboer took over as coach. Prior to Pete they did not have nearly the level of drop off in their play with a lead. When people complain about how the team plays with a lead under Deboer it is not something that is imaginary. It is a real issue.
Why is this a case? Obviously, this study doesn’t answer. It could be playing a more passive system with a lead and sitting back more, playing worse players with the lead or likely both. When you consider how many leads the have blown, points dropped and points given to playoff competitors, this play with a lead is probably a major contributor to the Devils missing the playoffs in the last two year.
Thank you for reading my book