Spurred by the recent hit by David Koci on Mike Green, I decided that the world should know how worthless Koci is to his team and the league. We all know by now that Koci is not out there to score points (he has 3 in 97 games), nor to take shots (22 in that same span, or .23 per game). In other words, he contributes more shorthanded time to his team (.39 non-fighting PIM per game) than he does goal opportunities (aka shots).
After ranting about Koci, I decided to make him historic. Developing a statistical means of determining what a player can do to help and what a player can do to hurt their team, I put Koci's ineptitude up against the worst players of the last 35 or so years. I counted goals and assists per game as a plus, shots per game as a plus, plus/minus per game as a plus or minus, and finally I took the player's non-fighting PIM per game and multiplied it by the league average power-play conversion for the years that the player was in the league (and subtracted it). I only looked at forwards, as defensemen tend to contribute positively in ways that are harder to assess, and go beyond the above categories. I also had a hard time debating whether a hit could be positive, and finding data on shot blocking (which would definitely be a positive). It's a "rookie" metric, really, but what I found was that the players that harm their team most tend to do two things in particular: they take a lot of non-fighting penalties, and take very few shots. No surprise, in a way, because of what I'm considering, but it makes sense in the context of what can help a team versus what can hurt.
For the sake of my search, I initially relegated my search to players that played at least 90 games. The results were interesting:
10th Worst: Stu Grimson, .02 Player Contributions/Game (PCG)
9th: Gary Rissling, .00 PCG
8th: Patrick Cote, -.03 PCG
7th: Kevin Kaminski, -.04 PCG
6th: Dennis Bonvie, -.06 PCG
5th: Gord Dwyer, -.10 PCG
4th: David Koci, -.21 PCG
3rd: Jay Caufield, -.26 PCG
2nd: Brantt Myhres, -.27 PCG
1st: Richard Zemlak, -.32 PCG
Yes, the immortal Richard Zemlak had the worst career of the last 35 years, and he could likely be considered one of the more useless players in league history. Over 132 games from 1986-1992, he had 2 goals, 12 assists, 39 shots, 342 non-fighting PIM (587 PIM total), and compiled a plus/minus of -26.
What was just as interesting was the additional fiddling I did with this metric. I assumed that goons would do the most harm, as they typically do not take many shots but take many penalties. What I actually found was that not all goons were created equal. Grimson is shown above, but look at these other examples:
Derek Boogaard, .09 PCG
Craig Berube, .40 PCG
Gino Odjick, .92 PCG
Daniel Carcillo, 1.47 PCG
Tiger Williams, 1.67 PCG
A goon's ability to create offense tends to offset their PIM detriment with PCG; I also found another interesting result in Stephen Peat:
Stephen Peat, .18 PCG
Despite complete ineptitude over his career (130 games, 8 goals, 2 assists, 41 shots, -16 +/-), the fact that a whopping 71 percent of his career PIM were fighting PIM's reduced the negative impact on his team. I'm not saying that he drastically helped his team, but he did not hurt his team with shorthanded situations like the others did. He also participated in one heck of a fight.
Finally, I also wanted to provide a little perspective on player value, by showing the PCG for Alex Ovechkin and Wayne Gretzky:
Wayne Gretzky, 5.62 PCG
Alexander Ovechkin, 6.77 PCG
Ovechkin's sheer volume of shots increases the likelihood that his team will score, and thus his player value is a bit higher. Like I mentioned before, this is not a perfect metric as of yet, but I like to think it nudges us closer to thinking about what are positive contributions, what are negative, and how we would assess value to those actions.
Koci is not the worst of all-time, but he's close. He also has some career left, despite my contention that he never should have had one to begin with. That being said, Zemlak carries the crown, and proud he should be.