Unconventional Wisdom - Part 2 Point Shares

Continuing our series on statistical analyses for the Pittsburgh Penguins I decided to take a look at Point Shares (PS). Point Shares are used by Hockey Reference, a metric created by Justin Kubatko that attempts to mimic the Win Shares advanced metric used for baseball. The concept of Point Shares is that the total number of Points the team finishes the season with will be roughly equal to the team Point Shares. Then individual players get a certain amount of those Point Shares based upon their overall contribution to team success. I quite like this system, and it seems like a pretty good way to judge who should be the MVP at the end of the season, since the player with the highest PS will not only have been playing for a successful team but also would have contributed more to the team's success than his peers.

This system starts by having to calculate team data, through which he creates two new metrics Marginal Goals For and Marginal Goals Against, which is basically how your team is performing compared to the league average goals per game, although the actual formula is slightly more complex than that. But using that system they were able to create an Expected Points total that is quite accurate, with a margin of error between 4-6 Points, and that consistently applies to all NHL seasons going back to 1917-18.

From an individual player perspective they are broken down into Offensive Point Shares (OPS), Defensive Point Shares (DPS), Goalie Point Shares (GPS), and then overall PS is a sum total of those values.

OPS is based on Goals Created, which is another new metric based on the player's Goals plus half his Assists in proportion to the overall team totals. The Goals Created is then used to calculate Marginal Goals For, which is based on the players' Goals Created in relation to his total TOI and compared to the league totals.

DPS is even more complicated, because they had to create a formula that would determine how much of the team's Goals Against can be attributed to the skaters rather than the Goaltender. The complex formulae use Goals Against, Shots Against, Plus/Minus, and TOI in comparison to the league totals. GPS uses a similar system using TOI, Shots Against, and Goals Against in relation to league totals.

OPS

Like we saw with OGvT, over the past 3 seasons our current Top 6 is at the top of the chart in OPS. Crosby is in a league of his own, with Malkin being nearly as high above the next tier as Crosby is above him. Neal and Kunitz are nearly neck and neck, with a slight drop off to Dupuis and then Jokinen. Crosby has actually been decreasing over the past 3 years, although that may be a product of his overall contributions leveling off as he gets to play more often, as he is still well above everybody else despite the decrease in his individual totals. Malkin had a huge year in 2011-12, but on average his totals have remained quite consistent. Neal, on the other hand, had quite the bump in 2011-12 and maintained that performance for the past 2 seasons. Kunitz was the opposite, two years of steady performance and then a huge bump last year. Dupuis has been steadily improving every year, and Jokinen has inconsistently fluctuated up and down.

Unlike OGvT, Bennett is actually significantly below Jokinen. He is followed by D`Agostini, Jeffrey, and Sutter. At the far end, well below them is Kobasew. And then the final 3 actually have negative OPS, meaning when they are on the ice it actually detracts from the offensive output of the team, Vitale, Adams, and Glass. D`Agostini, despite his totals putting him up in Bennett territory, has steadily fallen of the past 3 seasons, to the point where he is recently one of the lowest forwards. Jeffrey had a big year in 2010-11, but then dropped a lot, but remained consistent the past 2 seasons. And then Sutter, despite being lower than them on average, has actually been on the rise the past 3 seasons. Kobasew had a slight drop in 2011-12, but overall has been quite consistent. Vitale has been dropping steadily the past 3 seasons, while Adams has fluctuated, and Glass went from two consistent seasons to an extreme drop off last year.

On D we again see Letang at the top of the list, no surprise there. Well below him is Martin, then not far behind is Bortuzzo and Niskanen, a little further down Engelland, and then Orpik. Scuderi is the lone D on the roster that actually has a negative OPS. Letang, not only is he above and beyond the most offensively gifted D, his number increased steadily the past 3 seasons. Martin was steady for 2 years and then had a huge increase last year. Bortuzzo had extremely limited TOI, so his totals are a bit suspect, but he did see marked improvement last year. Engelland and Orpik, on the other hand, increased for 2 years but then suddenly dropped off last season. Scuderi did the opposite, dropping steadily for 2 straight years but having an impressive bounce back last year.

DPS

Quite surprisingly, our defensive forwards are not at the op of the list. Crosby is at the top of the chart, followed by Kunitz and then Dupuis. Next is Malkin, Neal, and Bennett who are all roughly even. That is not the list you would expect to see, since that is basically our Top 6 forwards. Crosby has remained quite consistent over the past 3 years, while Kunitz has been a tad inconsistent, although saw a big bump last year. Dupuis was steady for two years and then also saw a big bump last year. Malkin was mostly consistent in the two years surrounding his impressive 2011-12 campaign. And Neal has been mostly consistent although decreasing ever so slightly over the past 3 seasons.

The bottom group is led by D`Agostini and Sutter, followed by Jokinen, Jeffrey, and Kobasew. Then a slight drop to Adams, and another drop to Vitale and Glass. D`Agostini improved for 2 seasons, but then had a big slump last year. And Sutter and Vitale oddly enough have been steadily decreasing the past 3 seasons. Jokinen and Glass were consistent for 2 years and then had a huge drop off last season. Jeffrey had a rough 2011-12 but started to bounce back last year. Kobasew was steady for two years and then had a big leap upwards last year. Adams has been mostly consistent but ever so slightly decreasing.

On D we still see a slightly unexpected result with Letang at the top, but then he is followed by Orpik and Martin, and then Scuderi. A bit further down we have Niskanen and Bortuzzo, and then Engelland. Letang and Niskanen improved slightly in 2011-12 and have been steady the past 2 years. Orpik had a slight drop in 2011-12 but otherwise has been quite consistent. Martin had a huge drop in 2011-12, but then an equally huge comeback last season to leapfrog past Orpik. Scuderi was consistent for 2 years and then had a huge drop off last season. Bortuzzo oddly enough despite small sample size has remained consistent for the past 2 years. And then Engelland, who had a nice bump in 2011-12 but then dropped off a bit last season.

GPS

As we saw in GGvT, Vokoun has a slight edge over Fleury, but they are nearly identical. What is interesting is that Vokoun had a big year in 2010-11 and then dropped to remain steady the past 2 seasons, while Fleury had two consistent seasons sandwiched around a poor 2011-12 season. But where they stand now they are practically dead even.

PS

The overall totals are once again what we would expect to see. Crosby is above and beyond the most important player on the team, having a high PS total than both goaltenders. After the goaltenders we have our next best forward Malkin, and then a bit further down Kunitz, followed by Neal, and then Dupuis. And lastly we have Jokinen followed very closely by Bennett. Our Top 6 + 1. As we saw with OPS, Crosby has been decreasing over the past 3 years but is still head and shoulders above everybody else on the team. Perhaps as the other players start coming into their own it means less weight for him to carry on his shoulders. Malkin again had an extremely successful 2011-12 season, but his other seasons were significantly lower, although improving. Kunitz was steady for 2 seasons and then had a breakout last year, while Neal had a huge 2011-12 season and then remained consistent for 2 years. Jokinen had a bit of drop off in 2011-12, but has remained steady the past 2 seasons.

The lower tier is led by D`Agostini, he would probably be able to manage to sneak his way into the Top 6 on a team with less talent on the Top 6 than we have. Then we have Sutter and Jeffrey, a pretty hefty leap down to Kobasew, and an equally daunting leap again down to Adams and Vitale. Glass wound up actually averaging out to a negative PS. Despite being up so high on average, D`Agostini has actually been steadily decreasing the past 3 seasons. Sutter, on the other hand, improved ever so slightly in 2011-12 and has remained consistent the past 2 seasons. Kobasew remained mostly steady for 2 years and then had a big bump last season. Adams had 2 consistent seasons sandwiched between a very decent 2011-12 campaign. Vitale has been steadily decreasing the past 3 seasons, while Glass was consistent for two years and then had a huge drop off last year.

On D the best all around defenseman is Letang, he is above and beyond better than the other skaters on the team other than Crosby and Malkin. A bit further down we have Martin, with another bit of a step down to Orpik. After that we have Niskanen, Scuderi, and Bortuzzo with practically identical numbers. And lastly Engelland, no surprise there. Letang and Niskanen, as we have seen, have been steadily improving for the past 3 seasons. Martin, on the other hand, had a disappointing slump in 2011-12 but bounced back with a huge year last season. Orpik was quite steady for 2 years and then had a slump last season. Scuderi dropped off in 2011-12, but has since remained consistent for the past 2 seasons. Bortuzzo, small sample size and all, looks like he has been steadily improving as well the past 2 years. And then Engelland, he had 2 consistent seasons sandwiched around a breakout 2011-12 campaign.