The Standardized Production Index (SPI)

After acquiring all the forward and defenseman data from 1990 to 2019 (from D-1 to D+3 seasons, if applicable), the next step was to build the thresholds or, as I call it, the ‘Standardized Production Index (SPI)’.

I won’t get into all the specifics about SPIs but the main points are:

  • Binning players by NHLe segments (SPI)
  • Looking at the age they first hit an SPI
  • Equating defensemen on the same scale as forwards

Here’s what the Standardized Production Matrix looks like for forwards drafted from 1990 to 2019…

Players who were drafted after 2012 and haven’t hit 200 games yet are removed from the above analysis as they are still developing. Players can appear multiple times here as they can move from one threshold to the next. For instance, Johnny Gaudreau would appear four times as he hit four distinct thresholds.

And here’s what the Standardized Production Matrix looks like for defensemen drafted from 1990 to 2019...

Hitting high SPIs at younger ages, for F and D, provide by far the highest probabilities of turning into a Star Producer or Average Producer or just making the NHL in general. We can see that the chances of making the NHL and making an impact, offensively, drop dramatically as lower SPIs are achieved or higher SPIs are achieved at older ages.

To understand how important these groups are, look at Star and NHLer rates of a player who hits a 2 or 3 SPI before 18 vs. one who does so between 18 and 19. These could be players born a few months apart and there are substantial differences.

The best rates for finding an impactful offensive forward or defensemen reside in two groups – an SPI of 3 Before Turning 18 and an SPI of 4 Before Turning 19. These players are often the first ones off the board, but not always. It’s not uncommon for teams to pick players with SPIs of 0, 1 or 2 early in the 1st round. Not surprisingly, 0 and 1 SPIs have the highest bust rates of the 1st round.

To really grasp how important these groups are, let’s consider an example. You’re a GM at the upcoming 2020 NHL draft and you’re due to draft 15th overall.  You see a forward, in his first eligible draft year (i.e., true draft year) with an enormously high equivalency (e.g., 4 SPI Before 19) on the board who was ranked by everybody as a 1st round talent. But you pass on him. Maybe he didn’t wow you at the combine or his skating is a bit iffy.

As GM, you do focus on junior production and think it has some relevance, so you’ve narrowed in on a few forwards that were roughly point-per-game first-year eligible players in the OHL. Let’s say they’re both 2 SPI players but one is born September 1st, 2002 (2 SPI Before 18) and one born November 1st, 2001 (2 SPI Between 18 and 19).

The probabilities change, obviously, for a 1st rounder vs. a late rounder but let’s use the overall probabilities above to paint the picture here. You’ve said no to the player with the outlier high equivalency, so you’ve just said no to roughly a 41% chance of a Star and a 76% chance of a Regular NHLer. Now you’re looking at a best case scenario of a 41% chance of making the NHL and a 8% chance of a star. You trust your gut and you choose the November 1st guy because he has better size and there’s something intangible you like more about him. 

The player you’ve now chosen has only about a 32% chance of becoming an NHLer and a 6% of becoming a star. This can improve with massive progression in the next few years but is far from a guarantee.  At this moment, you’ve effectively reduced your chance of an NHLer by more than 50% and reduced your chance of drafting an impactful scoring star by more than 85%, by passing on the 4 SPI player. 

Let’s say the player you chose takes a huge step in his D+1 year and doubles his production. Roughly speaking, he now finds himself in the bucket of players who have a 14% chance of star and 56% chance of Regular NHLer. The player has had massive progression and has improved his chances of elite offense and becoming an NHL Regular. 

However, he is still miles behind the player who was already there in the first place and is now older than that player. Hitting high SPIs and doing so early is imperative to become a star producer in the NHL. With each year that passes without hitting high SPIs, the likelihood of becoming a star and an NHLer fall dramatically.

No matter what the player you chose does from the point he is drafted, he can never obtain the same odds of elite offensive potential as the original player that was passed over.  Can he turn into a star?  Absolutely but the cards are stacked against him. This is like having the option of ninety cents or fifty cents, with the goal of turning it into a dollar, and you choose the fifty cent option.

This probably sounds far-fetched but happens nearly every draft, to varying degrees. Take 2016, for instance, when Alex DeBrincat had the 3rd highest NHLe of all 1st-year-eligible forwards in a stacked draft. He was skipped over 26 times by teams that drafted forwards, while really he should have only been skipped over by four or five players. Those players being Auston Matthews, Matthew Tkachuk, Patrik Laine, Pierre-Luc Dubois and Clayton Keller.

He dropped because he played with McDavid (in his pre-draft year not draft year) so they didn’t know if the production was real (odd, as he produced more in his DY, without McDavid), he is short and and doesn’t have the fastest foot speed for a little guy. DeBrincat, for his part, has already played two seasons in the NHL and is well on his way to being a star in the NHL, producing at a rate of 0.78 ppg. He also just had his first 40-goal campaign. Many more to come!

Let’s have a look at DeBrincat at the forwards drafted before him. Here’s look at every selection from Matthews to DeBrincat and their chances of being a Star and Making the NHL, based on a predictive model I’ve built from the SPIs (I’ll speak more to the model in a separate article).

Pay no attention to the white line running through, that’s just a way for me to filter the data by when the player was selected in the draft.

When you see it visually, it’s almost shocking. After Keller, DeBrincat has by far the highest chance of becoming a star and highest chance of becoming an NHLer (Bellows and Mascherin also have high probabilities of both but nothing like DeBrincat). There’s several players taken between Keller and DeBrincat (and Mascherin, for that matter) with high NHLer odds, as you’d expect in the 1st round and early 2nd round. However, the difference in star probability between the players chosen in between Keller and DeBrincat is quite staggering.

Not only were there numerous players with 2 SPIs (older and younger) taken, there were players that even had 0 SPIs and 1 SPIs, overagers and even a few who players who played in a league so low on the depth chart that I don’t even have an equivalency for them.

DeBrincat was skipped over for all of these players. Three years later, every team that skipped on him is regretting it. DeBrincat should have been a top 10 pick. A player like DeBrincat should never drop out of the 1st round but they do.

The high-tier groups are very subtle and appear to be mostly unnoticed by a wide margin of NHL organizations. Knowing what to look for here could drastically increase a team’s ability to draft NHLers. Not just guys who make the NHL but offensively impactful NHLers.

Leave a Reply