Eastern Edge: Buy Low and Sell High Candidates Using IPP

by Cam Metz on January 22, 2019
  • Eastern Edge
  • Eastern Edge: Buy Low and Sell High Candidates Using IPP

 

The midseason guide is still for sale, it’ll still help you get geared up for the stretch run –there is a lot of information in there that you won’t find in the ramblings or weekly columns. It’s your last chance to hone in on your team’s weaknesses.

Over the last couple of season roughly speaking an average a player in the top-200 of point producers will see around 95 goals scored while they are on the ice in all situations.  Obviously this number is a little higher in 120s for a top player, but in general the 95 goals is a great approximation for how turning an IPP into a quantified point production.  Interestingly this season we are seeing some of the top players pace for a goals scored while they are on the ice around 140.  You have to get to the top 80 to get the average just around 130 GF.  This falls in line with the general narrative of the season, there are more goals being scored this season compared to the last couple of seasons.  

I’ve written a lot about IPP and how it can dictate what we expect to see happen with a player over the rest of the season. Remember IPP is the percentage of times a player gets a point when they are on the ice and a goal is scored.  Better players tend to have higher IPPs because they are out there making plays.

It’s hard to see how IPP at a glance could impact a player on your team, when you’re talking about a couple percentage points it’s difficult to quickly translate that into total points scored for your player and how that could be changing as regression rears its head. For example if player A has an IPP of 75% but his average is usually around 67%  what does that mean for the total points score for the player? Well it depends on how many goals have been scored while they are on the ice. 

Hold that thought on IPP and what has been going on this season.

If you look at how players have produced so far this season, their “82 game pace” we’ll call it, then you can get a gauge of what they are projected to score without injuries or having a couple games less than a player on another team.  This is a great way to see how Taylor Hall is doing even though he’s been off the ice for a little while with a lower body injury.  Thankfully Frozen Tools takes care of the leg work for you – see Taylor Hall below:

 

 

So Hall is on a 92-point pace – when he is all systems go you want him in your lineup (obviously).  So you go over to the advanced stats page and you’re presented with this table:

 

 

Looking at his IPP you can see that he has a healthy 77 percent – this is a great example because you can see how consistent Hall has been the last seven seasons, he has average a 77 percent IPP.  In fact, you’ll see quite often that most players fall within 5 percent over their average over a 3 to 4 year period.   Basically if the player you are looking at is within their normal range +/-5 percent you shouldn’t expect anything wild to happen.  Obviously younger players who are turning into stars will buck this trend a little, but for the most part it will hold true.

So here comes the question then – how do we approximate a player’s projected point total due to regression in their IPP?  Well we have to set a few rules, based on the 5 percent rule – we’re really not that interested in players who are within 5 percent of their IPP average over the course of the last three years.  We only want to look at the top 150ish scores because higher point totals will necessarily imply more a games played and a larger IPP sample size for the season. 

Here is the gist of what I looked at – I generated a player’s average IPP (in all situations) over the last three seasons and compared it to this season’s IPP.  The difference gave me a percentage that if outside the 5 percent range for that player is worth looking at in greater detail.  Second, I took the player’s current pace and divided it by their IPP to get the number of goals scored while they were on the ice (GF).   We can estimate a total GF if the player played 82 games by dividing their GF/GP and multiply by 82. Multiplying by an “average” IPP we can see a new projected player pace based on this year’s higher scoring season.

Below you’ll see two tables of players in the East; one is for players likely to increase from their current pace and the other players who are going to most likely decrease.

The Delta IPP is the suggested difference from a player’s average IPP compared to this seasons IPP. Current Points is how many points the player has scored so far and their pace is based on their GP extrapolated out to 82 games.  The estimated IPP is a correction to their 18-19 IPP by the Delta IPP and their estimated points is their expected GF multiplied by their estimated IPP.  The resulting difference from their current pace compared to their estimated point pace is how many points we could see a player capture via a reversion to the mean in their IPP.

 

 

Long story short this is your list of the players you want to trade for heading into the fantasy hockey playoffs.  I’ll bet that Hedman, Crosby, Kuznetsov, and Barkov are all capable of finding a way back to their average IPP before the season ends – finding a these slight increases in points can be the difference in having a championship team. 

 

So who are you going to trade to get the players above? Well see if one of the guys below is on your team and lines up with a superstar above.

 


If you’re a Kessel owner – it sure looks like you should be trying to find a way out of this sky high IPP.

 Nyquist has been benefiting from playing with Larkin but this model’s interpretation says that selling high is advisable. 

Zibanejad has never broke the 70 percent IPP barrier before and if he’s able to do that more power to him – but it’s not the bet I’m willing to make, see if someone else will buy him at his current pace.  

 

As we highlighted last weekMax Domi has been playing over his head – it’s likely his fall will continue.

 

Hope you enjoyed this week’s article.  Hopefully this estimation doesn’t have any glaring holes – let me know if you see anything that needs tweaked.  If you’re interested in the data table send me a message on twitter and I’ll share the file with you.  Find me @DH_jcameronmetz.